Yang-Lee and Fisher zeros generalized on some far-from-equilibrium systems

Combustion and NO x emission characteristics of a retrofitted down-fired 660MW e utility boiler at different loadsZhengqi Li ⇑,Guangkui Liu,Qunyi Zhu,Zhichao Chen,Feng RenSchool of Energy Science and Engineering,Harbin Institute of Technology,92,West Dazhi Street,Harbin 150001,Chinaa r t i c l e i n f o Article history:Received 7September 2010Received in revised form 23November 2010Accepted 20January 2011Available online 12February 2011Keywords:Down-fired boiler RetrofitCarbon content in fly ash Thermal efficiencya b s t r a c tIndustrial experiments were performed for a retrofitted 660MW e full-scale down-fired boiler.Measure-ments of ignition of the primary air/fuel mixture flow,the gas temperature distribution of the furnace and the gas components in the furnace were conducted at loads of 660,550and 330MW e .With decreasing load,the gas temperature decreases and the ignition position of the primary coal/air flow becomes farther along the axis of the fuel-rich pipe in the burner region under the arches.The furnace temperature also decreases with decreasing load,as does the difference between the temperatures in the burning region and the lower position of the burnout region.With decreasing load,the exhaust gas temperature decreases from 129.8°C to 114.3°C,while NO x emissions decrease from 2448to 1610mg/m 3.All three loads result in low carbon content in fly ash and great boiler thermal efficiency higher than 92%.Com-pared with the case of 660MW e before retrofit,the exhaust gas temperature decreased from 136to 129.8°C,the carbon content in the fly ash decreased from 9.55%to 2.43%and the boiler efficiency increased from 84.54%to 93.66%.Ó2011Elsevier Ltd.All rights reserved.1.IntroductionReserves of anthracite and lean coal are abundant and globally distributed.The use of anthracite and lean coal,which have low-volatility contents,presents difficulties in ignition and burnout.More and more research has been conducted around the world to solve these problems.At present,down-fired combustion is widely applied in the power industry to consume low-volatility coals,and its main merit is that it achieves a high degree of burn-out by prolonging the residence time of pulverized-coal in the fur-nace.However,a practical down-fired boiler operation still suffers from the problems of high carbon content in the fly ash and poor flame stabilization at low load without oil support firing.Precise understanding of the behavior of char particles in pul-verized-coal combustion systems is critical in determining funda-mental processes that occur during heterogeneous bustion data obtained for full-scale equipment can give the combustion and NO x emission characteristics of real combustors,in particular the turbulent flow of industrial coal flames.As a re-sult,studies using full-scale equipment are highly desirable and a necessity.In the surrounds of wall-fired boilers,measurements are made of the local mean concentrations of O 2,CO,CO 2,and NO x ,gas temperatures,and char burnout through several observ-ing doors in the utility boilers [1–5].Experiments have also been performed for tangentially fired boilers [6–9].However,for down-fired boilers adopting Foster Wheeler technology,there has only been the work of Li et al.,who measured the gas temperature,gas species concentrations in the furnace and carbon content in the fly ash in a 300MW e boiler before and after retrofit [10–14].In the present study,in situ experiments were carried out for a 660MW e down-fired pulverized-coal boiler after retrofit;this unit has the maximum capacity of any currently used worldwide [15].Measurements of ignition of the primary air/fuel mixture flow,the gas temperature distribution of the furnace and the gas compo-nents in the furnace were made for this full-scale boiler at the rated,middle and half loads.The collected data were used to deter-mine the combustion and NO x emissions characteristics of the boi-ler at different loads.The results obtained from these experiments help solve similar problems and benefit the design and operation of 600MW e and 1000MW e down-fired boilers,and the data can be used to support theoretical and numerical calculations.2.The utility boilerThe investigated 660MW e boiler,having the largest capacity of any down-fired boiler in the world,was made by Foster Wheeler (FW)Corp.Fig.1a is a schematic diagram of the furnace.Arches di-vide the furnace into two:the lower furnace below the arches and the upper furnace above the arches.Originally,36cyclone burners were arranged on the arches to produce a W-shaped flame.The fuel-rich flow streaming from the cyclone nozzle is near the0306-2619/$-see front matter Ó2011Elsevier Ltd.All rights reserved.doi:10.1016/j.apenergy.2011.01.048Corresponding author.Tel.:+8645186418854;fax:+8645186412528.E-mail address:green@ (Z.Li).water-cooled wall while the fuel-leanflow jetting from the fuel-lean pipe faces the furnace center.Under the arches,there are three tiers of secondary air denoted D,E and F,all of which are fed into the furnace horizontally.Details of the specific structure of this kind of boiler can be found in the literature[10–13].During practical operation,the boiler suffers from great resis-tance and serious abrasion of the cyclone and high carbon content infly ash up to7–10%.Fuel-richflows form after the separation of the coal/airflows in the cyclones.The swirl intensities of fuel-rich flows decrease at the exits under the effect of the adjustable vane, which provides great resistance.The fuel-richflows with residual swirl have little rigidity and abrade the nozzles of the cyclones [16].The secondary air under the arches enters the furnace horizon-tally and mixes prematurely with the fuel-richflow.This reduces the down-forwardflow and space for combustion and decreases the temperature in the lower furnace,which causes a series of problems including the delayed ignition of pulverized-coal and high carbon content in thefly ash[13,17].There is fuel-leanflow with low momentum and shallow penetration depth near the furnace center, which readily shortens theflame and reduces upflow to the burnout region quickly.This is also a reason for the poor burnout degree of coal particles and high carbon content in thefly ash[18].In2009,Li et al.retrofitted the combustion system with many items.Fig.1b is a schematic diagram of the retrofitted combustion system.A fuel louver concentrator replaced the original cyclone as it has much lower resistance than the cyclone.The fuel/airflowfirst passes through the louver concentrator and separates into fuel-rich flow and fuel-leanflow.The fuel-richflow further divides into two and is injected into the furnace through the original primary air ducts facing the furnace center while the fuel-leanflow as vent air is injected into the furnace near the wall.The vent air ducts are con-trolled by valves.The horizontal secondary air is modified by the F-tier secondary air with an angle of declination of20°.In this retrofitted boiler,measurements were conducted at loads of660,550and330MW e with a constant opening of the vent air valves.3.Data acquisition methods and experimental conditionsThe formal in situ experiments were carried out in the2026tph down-fired pulverized-coal boiler to investigate certain aspects of the combustion process and NO x formation in the furnace.During the experiments,soot blowing and sewer bleeding were not per-mitted.The coal used in the experiments was a mixture of anthra-cite and lean coal.Sample characteristics of coals before and after retrofit are presented in Table1.The following parameters were measured:(1)The gas temper-ature of the furnace was measured with a leucoscope through observing doors in the front,rear and side walls.The layout of the observing doors is shown in Fig.2.The measurement error is 50°C.(2)The gas temperature of the furnace was measured with a nickel chromium–nickel silicon thermocouple through observing doors1,2and3.As shown in Fig.2,observing door1is in the air-flow zone of the tier D and E slots,observing door2is in the airflow zone of the tier F slots,and observing door3is above the arches. The end of the bare thermocouple was exposed in the furnace,so the temperature measured should be higher than the local gas temperature because of highflame radiation.However,because of radiation between the bare thermocouple and the water-cooled wall,and the close proximity of the two,the temperature mea-sured should be lower than the local gas temperature.The calcula-tions have indicated that in the region of highest temperature the ‘‘true’’temperature do not exceed the measured one by more than 8%[3,19].The thermocouples used were thoroughly checked be-fore leaving the factory,giving high confidence in the temperature measurement results.To minimize errors due to ash deposition,Table1Characteristics of the coal used in the experiments before and after retrofit.Quantity Before retrofit After retrofitProximate analysis,wt.%(as received)Volatile10.729.29Ash30.8431.51Moisture0.54 2.48Fixed carbon57.9056.72Net heating value(kJ/kg)23,53121,250Ultimate analysis,wt.%(as received)Carbon59.7059.48Hydrogen 2.95 2.52Oxygen 3.81 3.53Nitrogen0.820.83Sulfur 1.340.79Z.Li et al./Applied Energy88(2011)2400–24062401the thermocouples were frequently retracted from the furnace and, where necessary,any deposits were carefully removed.Moreover, the probes were replaced if there was any thermal distortion ob-served.(3)The primary air/flow temperature distribution was measured with the same thermocouple inserted parallel to the axis of the fuel-rich pipe,as shown in Fig.1.(4)Gas compositions were sampled using a water-cooled stainless steel probe2.5m in length for analysis of the local mean O2,CO and NO x concentrations.As shown in Fig.3,the probe comprised a centrally located10mm (inner diameter)tube,through which quenched samples were evacuated,surrounded by a tube for probe cooling.The probe was cleaned frequently by blowing high-pressure air through it to maintain a constant suction rate.The water-cooled probe was inserted into the furnace through observing doors1–3(see Fig.2).The gases withdrawn were analyzed online using a Testo 350M system.Theflue gas after the air heater was also analyzed online.Calibrations with standard mixtures including zero concen-trations were performed before each measurement session.The measurement error for O2and CO2concentrations was1%,while that for CO and NO x concentrations was50ppm.(5)Unburnt car-bon infly ash was determined by collectingfly ash using a particle-sampling device with constant suction speed.Values of the main operating parameters for the three operating loads are listed in Table2.4.Results and discussionFig.4shows the gas temperature distribution of the fuel-burn-ing zone along the cyclone axes of the burners;zero points were set at the tips of the burner nozzles in the furnace.With decreasing load,the gas temperature decreases and the ignition position of the primary coal/airflow becomes farther along the axis of the fuel-rich pipe in the burner region under the arches,especially as the load decreases from550to330MW e.For the rated load,the gas temperature rises rapidly as the measurement point extends to lower positions,exceeding1000°C at a position400mm from the end of the fuel-rich nozzle and exceeding1200°C at 1400mm.For the load of550MW e,the gas temperature is a little lower than that for the rated load in the burner region,exceeding 1000°C at a position800mm from the fuel-rich nozzle.For the half load,the gas temperature rises slowly and the temperature gradi-ent is lower than in the other two cases,barely reaching1000°C at 2400mm.The mass ratio of coal/air decreases as the load decreases,indi-cating that the concentration of pulverized-coal falls in the primary air and the momentum of coal/airflow decreases,resulting in aTable2Boiler operating conditions and measurement results.Quantity Before retrofit After retrofit660MW e660MW e550MW e330MW eTotalflux of the primary air(kg/s)121127110.574Temperature of the primaryair(°C)130124125114Totalflux of the secondaryair(kg/s)657620511396Temperature of thesecondary air(°C)398405397362 Coal feeding rate(ton/h)268.6282.7244.6147.7 O2at the furnace exit(dryvolume%)3.35 2.58 2.02 5.31O2influe gases(dry volume%)4.76 3.99 3.047.27NO x influe gas(mg/m3at6%O2dry)1181244821631610Carbon infly ash(%)9.55 2.43 6.02 3.24 Exhaust gas temperature(°C)136129.8119.3114.3Thermal efficiency of the boiler(%)84.5493.6692.5992.862402Z.Li et al./Applied Energy88(2011)2400–2406shallower penetration depth and shorter residence time of the pri-mary coal/airflow in the lower furnace.For these reasons,the gas temperature measured along the axis direction of the fuel-rich pipe decreases and the ignition of the primary coal/airflow in the bur-ner region under the arches is farther along the axis.The mass ratio of coal/air is defined to indicate the coal concen-tration in the primary air.As the load falls from660to330MW e, the mass ratio of coal/air drops from0.618to0.554,which indi-cates an obvious decrease in the coal concentration;thus,the heat and time needed for coal ignition increase.This is one of the rea-sons for the low gas temperature in the burner region and far igni-tion position of the coal particles at half load.For the load of 550MW e,the mass ratio of coal/air is0.615,which is little different to that for the rated load.Furthermore,the quantity of primary air and coal feed rate both decrease as the load decreases,as does the momentum of coal/airflow,leading to shallow penetration of the pulverized-coal in the lower furnace.The short residence times then lower the overall temperature of the burner zone to the ex-tent that the fuel-richflow cannot obtain sufficient heat from the recirculating up-flowing gases(see Fig.1).Fig.5presents furnace gas temperature variations as measured using the thermocouple inserted through observing doors1–3.On the whole,the gas temperature measured through observing doors 1–3decreases with decreasing load.In the case of the temperature distribution measured through observing door1,the gas tempera-turefirst increases and then decreases because some of the high-temperature gas recirculates into the near-wall zone.As the measurement points move deeper into the furnace,the measured temperature begins to fall as the thermocouple enters the fuel-rich flow.At the rated load,the gas temperature is steadily around 1000°C at points further than1400mm,while for the load of 550MW e,the measured temperature is a little lower.At the same position but for a load of330MW e,the temperature falls below 700°C gradually,indicating that the fuel-richflow has not ignited. This temperature sequence is the same as that for the temperature gradients of the fuel-richflow mentioned above,indicating that ignition conditions worsen as the load decreases.In the cases of measuring the temperature through observing doors2and3,once the temperature reached1250°C,no further measurements were taken at deeper positions for the simple rea-son of protecting the thermocouple from burnout.From the rela-tively limited temperature distribution,Fig.5shows that in the zone near observing door2,the temperature rises rapidly at loads of660and550MW e,and at400mm,temperatures already exceed 1250°C.This indicates that coal burns more intensely in the lower furnace in both cases.At the load of330MW e,the temperature rises to nearly1200°C at the position1800mm from the side wall, increasing more slowly than in the other two cases.As illustrated above for the gas temperature in the burner region,the momentum of coal/airflow at half load is low,which results in a shallow pen-etration depth and short residence time of the primary coal/air flow and low furnace temperature in the lower furnace.Further-more,the F-tier air accounts for a large proportion of the secondary air entering the furnace under the arches;therefore,much of the air does not take part in combustion immediately,which decreases the temperature in the region near observing door2.In the zones near observing door3above the arches,the temperature increases to more than1100°C at a load of330MW e,which is just a little lower than the temperature near observing door2.This can be ex-plained by the delay in the fuel-richflow ignition causing much of the fuel to burn in the upper furnace.Fig.6presents furnace temperature variations measured through observing doors for the three loads.The furnace tempera-tures shown in Fig.6are averages of values measured through observing doors at the same level.In the three cases,gas tempera-ture peaks are all in the lower furnace,and the distances from the peak positions to the exit of the furnace are relatively large.Resi-dence times for coal burning in the higher-temperature zone are thus longer,which favors fuel burnout.The difference between temperatures in the burning region and the lower position of the burnout region decreases as the load decreases,and is just35°C at a load of330MW e.This is because the momentum of coal/air flow decreases,which results in a shallower penetration depth and shorter residence time of the primary coal/airflow in the lower furnace;the ignition position becomes farther from the fuel-rich nozzle,and more pulverized-coal combusts completely in the upper furnace.Moreover,the temperature in the furnace decreases with the decreasing load.The reason for this is that the total heatZ.Li et al./Applied Energy88(2011)2400–24062403provided to the furnace also falls with decreasing load,decreasing the furnace temperature.Fig.7shows the variations in gas species concentrations in the zones near observing doors1–3.Gas species concentration mea-surements begin400mm from the side wall to avoid the effect of an air leak.In all three cases,the O2concentration sequence is C(door1)>C(door2)>C(door3),which describes well theflow of coal as illustrated in Fig.1b.Fuel-richflowsfirst inject down-ward into the furnace and then reverse their direction upwards in the down-fired boiler.The coal/airflows pass sequentially through the observation doors1,2,and3and O2is consumed con-tinuously along the path of theflame.Thus,in each case,the O2 concentration in the zone near observation door1was the highest, that in the zone near observation door2was intermediate and that near observation door3was the lowest.In the airflowflow zone of the D and E tiers and in the zone near the wall,a great quantity of hot gas lowers the measured O2con-centration.When the measurement points are deeper and closer to the fuel-richflow,the measured value increases.O2concentra-tions near observing door1increase with decreasing load,espe-cially from550to330MW e.For loads of660and550MW e,at the position1800mm from the wall,O2concentrations are nearly 14%,suggesting that the fuel-richflow is beginning to react and a certain amount of O2is being consumed.For the load of 330MW e,however,O2concentrations are higher than18%,indi-cating that the fuel-richflow has not ignited,which agrees with the conclusion drawn from the temperature analysis.In the near-wall zone of observing doors2and3,O2concentrations at half load are higher than in the other two cases,even as high as8%in the near-wall zone of observing door3.At all three loads,CO concentrations strictly rule contrary to O2 concentrations.In the zones near observing doors1–3,the higher the O2concentration,the lower the CO concentration.In the air-flowflow zone of the D and E tiers,CO concentrations decrease with decreasing load because of the delayed ignition of the fuel/ airflow.In the near-wall zone of observing doors2and3,CO con-centrations in the case of the half load are lower than in the other two cases because of the highly oxidizing atmosphere in the furnace.2404Z.Li et al./Applied Energy88(2011)2400–2406In the zone near port1at loads of660and550MW e,volatiles are released and form a certain amount of NO x.However,biased combustion restricts the formation of NO x,and hence,NO x concen-tration here is only645mg/m3and630mg/m3respectively at a position1800mm from the wall.However,for the load of 330MW e,the NO x concentration is as high as3706mg/m3at the same position,because the high O2concentrations form the strong oxidative atmosphere near the observing door1.Moreover,the fuel/airflows with weak rigidity mix with the fuel gas quickly in the furnace;these are the reason for the high NO x concentration. Compared with those measured through observing door1,NO x concentrations for660and550MW e operation increase rapidly in the airflow zone of observing door2because fuel–NO forms con-stantly with further combustion of the pulverized-coal and ther-mal NO increasingly forms under the high-temperature condition in the lower furnace.The NO x concentration at a load of 550MW e is lower than that for the rated load because of the re-duced fuel–NO formation with a lower coal feed rate and the low-er-furnace temperature in this zone.In the case of330MW e,NO x concentrations decrease sharply,because the reducing atmosphere become higher with a great amount of O2consuming.In the zone near observing door3,the NO x concentration decreases with decreasing load.Table2lists results obtained for theflue gas after the air hea-ter.It is obvious that as the load decreases,the exhaust gas tem-perature decreases from129.8°C to114.3°C,while NO x emissions decrease from2448to1610mg/m3.Overall,the NO x content in theflue gas at the three loads was at high levels, exceeding the proposed Chinese emission standard for anthracite (1100mg/Nm3).To decrease these emissions,further technolo-gies,such as over-fire air[15,20],selective catalytic reduction and selective non-catalytic reduction[21,22]should be taken into account.All three loads result in low carbon content in thefly ash and great boiler thermal efficiency higher than92%.The O2 concentrations at the furnace exit when operating at660and 550MW e were2.58%and2.02%,respectively,with the small dif-ference between the two caused by thefluctuation of air supply during operation.The carbon in thefly ash at a load of550MW e was6.02%,which is higher than that at660MW e because the lower-furnace temperature was lower by as much as50°C than the temperature at the higher operating load.To enhance heat absorption in the reheater and to ensure sufficient temperature of the reheat steam,the method of increasing the excess air coef-ficient is commonly adopted when operated at330MW e.Thus, the O2concentration at the furnace exit at330MW e was5.31%, much higher than that at660or550MW e.Despite the high O2 concentration,the quantity of air supplied,including both pri-mary and secondary air,was low for the330MW e case,being only63%of the quantity supplied at660MW e and76%of the quantity supplied at660MW e,as shown in Table2.Therefore, the mean velocity of the fuel gas was lower and the residence time of pulverized-coal was longer in the furnace.These factors explain the high degree of burnout of pulverized-coal and low carbon content in thefly ash.Table2also shows the main operating parameters and mea-surement results at a load of660MW e before the combustion sys-tem retrofit.After retrofit,the exhaust gas temperature decreased from136to129.8°C,the carbon content in thefly ash decreased from9.55%to 2.43%and the boiler efficiency increased from 84.54%to93.66%.However,the NO x content in theflue gas (O2=6%)was higher than before the retrofit,because premature ignition of the pulverized-coal prolonged the residence time of coal particles in the lower furnace with high temperature.Furthermore, the O2concentration in theflue gas decreased from4.76%to3.99%, which increased the NO x concentration after conversion of O2at 6%.5.ConclusionIndustrial experiments were performed in a retrofitted 660MW e full-scale down-fired boiler.Measurements of ignition of the primary air/fuel mixtureflow,gas temperature distribution of the furnace and the gas components in the furnace were con-ducted at loads of660,550and330MW e.The results were as follows.With decreasing load,the gas temperature decreases and the ignition position of the primary coal/airflow becomes farther along the axis of the fuel-rich pipe in the burner region under the arches, especially as the load decreases from550to330MW e.With decreasing load,the gas temperature measured through observing doors1–3decreases,the ignition conditions worsen. Moreover,the combustion intensity of pulverized-coal reduces in the lower furnace.The furnace temperature also decreases with decreasing load,as does the difference between temperatures in the burning region and in the lower position of the burnout region.O2concentrations near observing door1increase with decreas-ing load,while O2concentrations at half load are higher than in the other two cases in the near-wall zone of observing doors2and3. CO concentrations strictly rule contrary to O2concentrations.The higher the O2concentration,the lower the CO concentration near observing doors1–3.NO x concentrations(O2=6%)near observing door1were low at loads of660and550MW e than at the load of 330MW pared with those measured through observing door 1,NO x concentrations for660and550MW e operation increase rapidly in the airflow zone of observing door2while NO x concen-trations at the load of330MW e decrease.In the zone near observ-ing door3,the NO x concentration decreases with decreasing load.With decreasing load,the exhaust gas temperature decreases from129.8°C to114.3°C,while NO x emissions decrease from 2448to1610mg/m3.All three loads result in low carbon content in thefly ash and great boiler thermal efficiency higher than92%.Compared with the case of660MW e before retrofit,the exhaust gas temperature decreased from136to129.8°C,the carbon con-tent in thefly ash decreased from9.55%to2.43%and the boiler efficiency increased from84.54%to93.66%.AcknowledgmentThis work was sponsored by the Hi-Tech Research and Development Program of China(863Program)(Contract No. 2006AA05Z321).References[1]Li ZQ,Jing JP,Liu GK,Chen ZC,Liu CL.Measurement of gas species,temper-atures,char burnout,and wall heatfluxes in a200-MWe lignite-fired boiler at different loads.Appl Energy2010;87:1217–30.[2]Li ZQ,Jing JP,Chen ZC,Ren F,Xu B,Wei HD,et bustion characteristicsand NO x emissions of enhanced ignition-dual register and central-fuel-rich swirl burners in a300-MWe wall-fired pulverized coal utility bust Sci Technol2008;180:1370–94.[3]Costa M,Azevedo JLT,Carvalho bustion characteristics of a front-wall-fired pulverized coal300MW e utility bust Sci Technol 1997;129:277–93.[4]Costa M,Silva P,Azevedo JLT.Measurements of gas species,temperature,andchar burnout in a low-NO x pulverized coal-fired utility bust Sci Technol2003;175:271–89.[5]Costa M,Azevedo JLT.Experimental characterization of an industrialpulverized coal-fired furnace under deep staging bust Sci Technol2007;179:1923–35.[6]Kouprianova VI,Tanetsakunvatanab V.Optimization of excess air for theimprovement of environmental performance of a150MW boilerfired with Thai lignite.Appl Energy2003;74:445–53.[7]Li S,Xu TM,Hui SE,Wei XL.NO x emission and thermal efficiency of a300MW eutility boiler retrofitted by air staging.Appl Energy2009;86:1797–803.Z.Li et al./Applied Energy88(2011)2400–24062405[8]Toshikazu T,Hirofumi O,Dernjatinc P,Savolainenc K.Reducing the minimumload and NO x emissions for lignite-fired boiler by applying a stable-flame concept.Appl Energy2003;74:415–24.[9]Li ZQ,Yang LB,Qiu PH,Sun R,Chen LZ,Sun SZ.Experimental study of thecombustion efficiency and formation of NO x in industrial pulverized coal combustion.Int J Energy Res2004;28:511–20.[10]Li ZQ,Ren F,Zhang J,Zhang XH,Chen ZC,Chen LZ.Influence of vent air valveopening on combustion characteristics of a down-fired pulverized-coal 300MW e utility boiler.Fuel2007;86:245–62.[11]Ren F,Li ZQ,Jing JP,Zhang XH,Chen ZC,Zhang JW.Influence of the adjustablevane position on theflow and combustion characteristics of a down-fired pulverized-coal300MW e utility boiler.Fuel Process Technol2008;89: 1297–305.[12]Li ZQ,Ren F,Chen ZC,Wang JJ,Chen Z,Zhang JW.Influence of oil atomized aironflow and combustion characteristics in a300MW e down-fired –Pac J Chem Eng2010;5:488–96.[13]Li ZQ,Ren F,Chen ZC,Chen Z,Wang JJ.Influence of declivitous secondary air oncombustion characteristics of a down-fired300-MWe utility boiler.Fuel 2010;89:410–6.[14]Li ZQ,Ren F,Chen ZC,Liu GK,Xu ZX.Improved NO x emissions and combustioncharacteristics for a retrofitted down-fired300-MWe utility boiler.Environ Sci Technol2010;44:3926–31.[15]Garcia-Mallol JA,Steitz T,Chu CY,Jiang PZ.Ultra-low NO x advanced FW archfiring:central power station applications.In:2nd US China NO x and SO2 control workshop,Dalian;2005.[16]Zhang J.Experimental study and numerical simulation on separation char-acteristics of cyclone separator of downfired boiler.Harbin:School of Energy Science and Engineering,Harbin Institute of Technology;2006[in Chinese]. [17]Ren F,Li ZQ,Chen ZC,Wang JJ,Chen Z.Influence of the down-draft secondaryair on the furnace aerodynamic characteristics of a down-fired boiler.Energy Fuels2009;23:2437–43.[18]Ren F,Li ZQ,Chen ZC,Xu ZX,Yang GH.Experimental investigations into gas/particleflows in a down-fired boiler:influence of the vent air ratio.Energy Fuels2010;24:1592–602.[19]De DS.Measurement offlame temperature with a multi-elementthermocouple.J Inst Energy1981;54:113–6.[20]Li ZQ,Ren F,Liu GK,Shen SP,Chen ZC.Influence of the overfire air ratio on theNO x emission and combustion characteristics of a down-fired300MW e utility boiler.Environ Sci Technol2010;44:6510–6.[21]Sounak R,Hegde MS,Giridhar M.Catalysis for NO x abatement.Appl Energy2010;86:2283–97.[22]Liang ZY,Ma XQ,Lin H,Tang YT.The energy consumption and environmentalimpacts of SCR technology in China.Appl Energy2010.doi:10.1016/ j.apenergy.2010.10.010.2406Z.Li et al./Applied Energy88(2011)2400–2406。

Systems&Control Letters54(2005)429–434/locate/sysconleSuboptimal control for nonlinear systems:a successiveapproximation approachଁGong-You Tang∗College of Information Science and Engineering,Ocean University of China,Qingdao266071,ChinaReceived11June2002;received in revised form15July2004;accepted24September2004Available online10November2004AbstractThis paper presents a successive approximation approach(SAA)designing optimal controllers for a class of nonlinear systems with a quadratic performance index.By using the SAA,the nonlinear optimal control problem is transformed into a sequence of nonhomogeneous linear two-point boundary value(TPBV)problems.The optimal control law obtained consists of an accurate linear feedback term and a nonlinear compensation term which is the limit of an adjoint vector sequence.By using thefinite-step iteration of the nonlinear compensation sequence,we can obtain a suboptimal control law.Simulation examples are employed to test the validity of the SAA.©2004Elsevier B.V.All rights reserved.Keywords:Nonlinear systems;Optimal control;Suboptimal control;SAA1.IntroductionThe investigation of the optimal control is of im-portance in modern control theory.The theory andthe application of optimal control for linear time-invariant systems have been developed perfectly. However,as for nonlinear systems,synthesis prob-lems that are solved by classic control theory lead todifficult computations.People have studied optimalଁResearch supported by the National Natural Science Founda-tion of China(60074001)and the Natural Science Foundation of Shandong Province(Y2000G02).∗Tel.:+865325901980;fax:+865325901225.E-mail address:gtang@(G.-Y.Tang).0167-6911/$-see front matter©2004Elsevier B.V.All rights reserved. doi:10.1016/j.sysconle.2004.09.012control of nonlinear systems for decades.Becerra and Roberts[2]gave out the optimal parameters ap-proximation for the optimal control of the discrete nonlinear systems.Stojanovic[7]focused on the op-timal attenuation control of ellipse nonlinear systems. Sokhin[6]studied the optimal control for a nonlinear system described by afluctuate equation.Hager[3] considered an optimal control method based on the multiplicator machine.Teo and Jennings[10]applied an optimal control approach to nonlinear systems with inequality constraint.Alt[1]analyzed the stability of nonlinear optimal control system with constraint. But all the results mentioned above are still limited to the descriptive ones.These results are limited in real application.430G.-Y.Tang/Systems&Control Letters54(2005)429–434 For the convenient implementation,many subopti-mal control methods have risen which do not pursuethe optimal control performance indexes.Nagurka andYen[4]developed the method of suboptimal controlfor nonlinear system based on the Fourier series.New-man and Souccar[5]addressed the suboptimal controland the robust analysis for second-order nonlinear sys-tems.Xi and Geng[11]presented a predictable controlmethod for suboptimal control of nonlinear systems.Tang,Qu,and Gao[8]proposed a sensitivity approachfor suboptimal control of nonlinear systems.In this paper,a successive approximation approach(SAA)of suboptimal control for a nonlinear system isproposed.Wefirst treat the nonlinear term in the sys-tem as an additional disturbance.Then using the SAAof differential equation theory,we transform the non-linear two-point boundary value(TPBV)problem intoa sequence of nonhomogeneous linear TPBV prob-lems.The optimal control law obtained consists of anaccurate linear feedback term and a nonlinear com-pensation term which is the limit of an adjoint vectorsequence.By using thefinite-stepiteration of the non-linear compensation sequence,we can obtain a sub-optimal control law.Finally,simulation examples areemployed to test the validity of the SAA.2.Problem statementConsider nonlinear systems described by˙x(t)=Ax(t)+Bu(t)+f(x),t>t0,x(t0)=x0,(1)where x∈R n,u∈R r are the state vector and thecontrol vector,respectively.A and B are real constantmatrices of appropriate dimensions,x0is the initialvector,f:C1(R n)→U⊂R n,f(0)≡0whichsatisfies the Lipschitz conditions on R n.The quadraticcost functional of system(1)isJ=12x T(t f)Q f x(t f)+tft0[x T(t)Qx(t)+u T(t)Ru(t)]d t,(2)where Q f,Q∈R n×n are positive-semidefinite ma-trices,R∈R r×r is a positive-definite matrix.The op-timal control problem is tofind a control law u∗(t)which minimizes the quadratic cost functional(2)sub-ject to the dynamic equality constraint(1).According to the optimal control theory and the necessary condi-tions for the optimality,we can obtain the following nonlinear TPBV problem−˙ (t)=Qx(t)+A T (t)+f x (t),˙x(t)=Ax(t)−BR−1B T (t)+f(x),t∈t T=(t0,t f],(t f)=Q f x(t f),x(t0)=x0,(3) and the optimal control law can be written asu∗(t)=−R−1B T (t).(4) Unfortunately,the analytical solution of this nonlinear TPBV problem in(3)is difficult to be solved gener-ally.Therefore,it is necessary tofind the approximate approaches for solving the optimal problem of nonlin-ear systems.In this paper we will propose a SAA. 3.PreliminariesConsider the time-varying nonlinear system˙x(t)=A(t)x(t)+f(x),t∈t T,x(t0)=x0,(5) where x∈R n is the state vector,A∈R n×n,x0is the initial state vector,f:C1(R n)→U,f(0)≡0which satisfies the Lipschitz conditions on R n.Define vector function sequence{x(k)(t)}asx(0)(t)= (t,t0)x0,t∈t T,x(k)(t)= (t,t0)x0+tt0(t,r)f(x(k−1)(r))d r, t∈t T,x(k)(t0)=x0,k=1,2,...,(6) where is the state transfer matrix with respect to the matrix A(t).Lemma1.Sequence(6)uniformly converge to the so-lution of system(5).Proof.Consider{x(k)(t)}as a sequence of C Nt0,t f, from(6)x(1)(t)−x(0)(t)=tt0(t,r)f(x(0)(r))d r,t∈t T.(7)G.-Y.Tang /Systems &Control Letters 54(2005)429–434431Because f satisfies the Lipschitz conditions on R n ,it followsM =sup t ∈R T(t,t 0) ,= z 0 ,sup t ∈R T f (z,t) z ,z ∈U,sup t ∈R Tf (v,t)−f (w,t) v −w ,v,w ∈U,(8)where M , , ,and are some positive constants, · denotes any appropriate vector or matrix norm.Noting that (t 0,t 0) = I =1,hence M 1.From (7)and (8)we obtainx (1)(t)−x (0)(t) Mtt 0x (0)(r) d r M 2(t −t 0),t ∈t T .(9)From (6)one gets x(2)(t)−x(1)(t)=tt 0(t,r)f (x (1)(r))−f (x (0)(r)) d r,t ∈t T(10)and thenx (2)(t)−x (1)(t)M t t 0 f (x (1)(r))−f (x (0)(r)) d rM tt 0x (1)(r)−x (0)(r) d r12!M 3(t −t 0)2,t ∈t T .(11)By the mathematics induction,we obtain x(k)(t)−x(k −1)(t)k −1Mk +1(t−t 0)kk !,t ∈t T ,k =1,2, (12)According to trigonometry inequality,for any jx (k +j)(t)−x (k)(t)k +ji =k +1i −1 Mi +1(t −t 0)i i !k M k +2(t−t 0)k +1exp ( M(t −t 0)),t ∈t T ,k =0,1,2,....(13)Thus {x (k)(t)}is a Cauchy sequence in C N [t 0,t f ].This sequence is uniformly convergent [9].For j is random,the limit of this sequence is the solution of system (5).The proof is complete. 4.SAA designing processConstruct the set of TPBV problems(0)(t)=f (x (0))=0,t ∈t T ,−˙(k)(t)=Qx (k)(t)+A T (k)(t)+(f x (k −1)(x (k −1)) x (k −1)(t),˙x (k)(t)=Ax (k)(t)−BR −1B T (k)(t)+f (x (k −1)),t ∈t T ,(k)(t f )=Q f x(t f ),x (k)(t 0)=x 0,k =1,2,...(14)and the corresponding optimal control sequence u (k)(t)=−R −1B T (k)(t).(15)For the k th optimal problem,optimal state trajectoryand optimal control law are x (k)(t)and u (k)(t),respec-tively.We give out the following theorem.Theorem 1.Assume that {x (k)(t)}and {u (k)(t)}are the solution sequence of (14).Then {x (k)(t)}and {u (k)(t)}uniformly converge to the optimal state tra-jectory x ∗(t)and optimal control law u ∗(t)for system (1)with the quadratic cost functional (2),respectively .Proof.Let(k)(t)=P (t)x (k)(t)+g (k)(t),t ∈t T ,k =1,2,...,(16)where P ∈R n ×n is unknown positive-semidefinite function matrix.g (k)∈R n is the k th adjoint vector.Calculating the derivative to the both side of the (16),we get˙ (k)(t)=˙P (t)x (k)(t)+P (t)˙x (k)(t)+˙g (k)(t)= ˙P(t)+P (t)A −P (t)SP (t)x (k)(t)−P (t)Sg (k)(t)+P (t)f (x (k −1))+˙g (k)(t),(17)432G.-Y.Tang /Systems &Control Letters 54(2005)429–434where S =BR −1B T .Substituting (16)into the second equation of (14)and comparing with (17),one can ob-tain the following Riccati matrix differential equation:−˙P(t)=P (t)A +A TP (t)−P (t)SP (t)+Q,P (t f )=Q f (18)and adjoint vector differential equations˙g (k)(t)=−[(A −SP (t))]T g (k)(t)−P (t)f (x (k −1))−f x (k −1)(x (k −1)) (k −1),g(k)(t f )=0,k =1,2,3, (19)We can get the unique positive-semidefinite matrix solution P (t)from (18).Note that P (t),f (x (k −1)),f x (k −1)(x (k −1))and (k −1)(t)are known functions in (19).So (19)is a nonhomogeneous linear vector dif-ferential equation for some certain k .Then we can obtain g (k)(t)by using reversing integration:g (0)(t)=0,g (k)(t)=t ft(t,r)[P (r)f (x(k −1)(r))+f x (k −1)(x (k −1)(r)) (k −1)(r)]d r,k =1,2,3,...,(20)where is the state transfer matrix with respect to the matrix (SP −A)T .Substituting (16)into (15),we obtain the k th optimal control lawu (k)(t)=−R −1B T [P (t)x (k)(t)+g (k)(t)].(21)Substituting (16)into the third equation of (14),we can get the k th optimal closed-loop system ˙x (k)(t)=[A −SP (t)]x (k)(t)−Sg (k)(t)+f (x (k −1)),x (k)(0)=x 0.(22)According to Lemma 1,the solution sequences{g (k)(t)},{x (k)(t)}of (20)and (22)are uniformlyconvergent.The control sequence {u (k)(t)}is only related to {x (k)(t)},{g (k)(t)},so it is also uniformly convergent.Define g(t)and u ∗(t)as the limits of sequences {g (k)(t)}and {u (k)(t)},respectively.Ac-cording to Lemma 1,the limit of sequence {x (k)(t)}is the optimal state trajectory x ∗(t)of the optimalcontrol problem (1)with the quadratic cost functional(2).Then we get the optimal control law u ∗(t)=−R −1B T [P (t)x(t)+lim k →∞g (k)(t)].(23)The proof is complete.In fact,we cannot calculate the optimal control lawin (23).We may find a suboptimal control law in prac-tical applications by replacing ∞with N in (23)u N (t)=−R −1B T [P x(t)+g (N)(t)].(24)Remark 1.In (24)x(t)is an accurate solution in case of k →∞,only g (N)is an approximation by substi-tuting a finite-stepiteration of g (N)for g (∞).So the suboptimal control law u N (t)in (24)is closer to the optimal control law than u (N)(t)in (21).Algorithm 1.Suboptimal control law of system (1)Step 1:Solve the positive-semidefinite matrix P (t)from Riccati matrix differential equation (18).Let x 0(t)=g 0(t)=0,J 0=0and k =1.Step 2:Obtain the k th adjoint vector g (k)(t)from (20).Step 3:Letting N =k ,calculate u N (t)from Eq.(24).Step 4:Find J N fromJ N =12x T(t f )Q f x(t f )+ t f[x T (t)Qx(t)+u T N (t)Ru N (t)]d t .(25)Step 5:If |(J N −J N −1)/J N |< ,then stopand out-put u N (t).Step 6:Calculate x (k)(t)from (22).Step 7:Letting k =k +1,go to step2.Remark 2.If t f →∞,the quadratic cost functional (2)is rewritten asJ =12∞0[x T (t)Qx(t)+u T (t)Ru(t)]d t.(26)The algorithm presented also can be used.In this sit-uation,Riccati matrix differential equation (18)is re-duced as the following Riccati matrix equation:A T P +P A −P BR −1B T P +Q =0.(27)G.-Y.Tang/Systems&Control Letters54(2005)429–434433Fig.1.Simulation curve of the system when k=1,2,3,and4.5.A simulation exampleConsider the nonlinear system described by˙x1˙x2=01−11x1x2+x1x2x22+1u,x1(0) x2(0)=−0.8.(28)The quadratic cost functional is chosenJ=12 10(x21+x22+u2)d t.(29)Simulation results are presented in Fig.1.The cost functional values at the different iteration times are listed in Table1.From Fig.1,it can be seen that the more iterates we take,the better are the approx-imations to both the state and the control functions. This leads to a better approximation of the optimal cost shown in Table1.If we choose =0.01,then the relative error of the cost functional values satisfies |(J4−J3)/J4|< .It indicates the4th suboptimal con-trol law u4is very close to the optimal control law u∗.Table1Cost functional values at the different iteration timesIteration time k1234 Cost functional J8.5557 6.9407 6.5392 6.52576.ConclusionsA successive approximation algorithm has been used to generate the suboptimal solutions of nonlinear systems.It consists of solving a Riccati equation and a vector linear differential equation at each step.The results of the simulation show the validity of the ap-proach mentioned.The proposed method is promising and easy to implement.References[1]W.Alt,Stability of solutions to control constrained nonlinearoptimal control problems,Appl.Math.Optim.21(1)(1990) 53–68.[2]V.M.Becerra,P.D.Roberts,Dynamic integrated systemoptimization and parameter estimation for discrete time optimal control of nonlinear systems,Internat.J.Control63(2)(1996)257–281.434G.-Y.Tang/Systems&Control Letters54(2005)429–434[3]W.W.Hager,Multiplier methods for nonlinear optimal control,SIAM J.Numer.Anal.27(4)(1990)1061–1080.[4]M.L.Nagurka,V.Yen,Fourier-based optimal control ofnonlinear dynamic systems,Trans.ASME J.Dyn.Syst.Meas.Control112(1)(1990)17–26.[5]W.S.Newman,K.Souccar,Robust,near time-optimal controlof nonlinear second-order systems:theory and experiments, Trans.ASME J.Dyn.Syst.Meas.Control113(3)(1991) 363–370.[6]A.S.Sokhin,Some optimal-control problems for a nonlinearcontrol system described by a wave equation,Differential Equations17(3)(1981)346–353.[7]S.Stojanovic,Optimal damping control and nonlinear ellipticsystems,SIAM J.Control Optim.29(3)(1991)594–608.[8]G.-Y.Tang,H.-P.Qu,Y.-M.Gao,Sensitivity approach ofsuboptimal control for a class of nonlinear systems,J.Ocean Univ.Qingdao32(4)(2002)615–620(in Chinese).[9]A.E.Taylor,y,Introduction to Functional Analysis,second ed.,Wiley,New York,1980.[10]K.L.Teo,L.S.Jennings,Nonlinear optimal control problemswith continuous state inequality constraints,J.Optim.Theory Appl.63(1)(1989)1–22.[11]Y.Xi,X.Geng,The suboptimality analysis of predictivecontrol for continuous nonlinear systems,Acta Automat.Sinica25(5)(1999)673–676.。

Thermodynamics a nd S ta1s1cal M echanicsSpring 2012Lecture 1Introduc1on t o T hermodynamicsDimi C ulcer微尺度11－002dimi@Office h ours: b y a ppointmentMy E nglish• e.g. = f or e xample • i.e. = t hat m eans • certain ~ 某 • assignment = h omework • mean = a verage • You c an s peak i n C hinese (slowly) • You c an s end e mail i n C hinese • Assignments a re i n E nglish • You c an a nswer i n C hinese• the U K = B ritain • the U S = A merica • TA = t eaching a ssistant • 帮教 • phenomenon, p henomena• momentum, m omentaSome w ords• fundamental a dj . 基础的,必要的; n . 基本法则• empirical 以观察或实验为依据的• law 定律• principle 原理• to v iolate 违反,违背• equa1on 方程• func1on 函数• variable a dj. 可变的; n . 变量• constant 常数• thermometer 温度计• macroscopic 宏观的• microscopic 微观的• arbitrary 任意的 • aVrac1ve 吸引的 • repulsive 排斥的 • Infinitesimal 极微小的 • concept 概念 • parameter 参数 • maVer 物质 • characterize 表征,描绘 • propor1onal 成比例的 • phenomenological 唯象的• concrete 实体的 • abstract 抽象的 • en1ty 实体 • deriva1on 衍生，派生State o f m aVer a nd p hase• High-­‐school p hysics u nderstanding• There a re 3 s tates o f m a)er• Solid, l iquid a nd g as• More g enerally, a s tate i s c alled a p hase• Solid p hase• Liquid p hase• Gas p hase o r g aseous p hase• Phase i s a m ore g eneral, m ore u seful c oncept • For e xample F e b ecomes f erromagne1c b elow T c• We s peak o f a n o rdinary p hase a nd a f erromagne1c p hase • Right n ow t his i s j ust v ocabulary• Later y ou w ill l earn t he p hysical d ifferenceWhat i s t hermodynamics?• Thermodynamics i s a m acroscopic t heory • It d escribes l arge, m any-­‐par1cle s ystems• Thermodynamics h as t wo t asks• Define a ppropriate p hysical q uan11es i n o rder t ocharacterize m acroscopic p roper1es o f m aVer• Relate t hese q uan11es b y a s et o f e qua1ons, w hich a reuniversally v alid (i.e. d o n ot d epend o n s pecific s ystem) • Thermodynamics i s p henomenological• It t akes i ts c oncepts d irectly f rom e xperiments• Thermodynamics i s a ll c lassical p hysics• No q uantum m echanics• We d o n ot t alk a bout a toms, m oleculesThermodynamics & S ta1s1cal M echanics• Sta1s1cal m echanics i s a m icroscopic t heory • We c are a bout t he d ynamics o f i ndividual p ar1cles• Sta1s1cal m echanics h as t wo b ranches• Classical s ta1s1cal m echanics• Quantum s ta1s1cal m echanics• We w ill s tudy b oth i n t his c ourse• Sta1s1cal m echanics h elps u s t o l ink t he p hysical laws o f t he m icroscopic w orld w ith t hose o f t he macroscopic w orld• One p ar1cle – m echanics• Many p ar1cles – s ta1s1csApplica1ons o f t hermodynamics • Thermodynamics h as m any a pplica1ons• Chemical r eac1ons• Phase t ransi1ons» e.g. s olid t o l iquid, l iquid t o g as» Can b e m ore c omplicated• Solid s tate» Magne1sm» Superconduc1vity a nd s uperfluidity (e.g. l iquid H e)» Charge a nd h eat t ransport• Stars, e.g. t he f orma1on o f s tars, C handrasekhar l imit• The a tmosphere, w eather, c limate• Biology e.g. l ife• All t hese s ystems o bey c ommon & g eneral l awsOverview o f t hermodynamics • In t he first p art o f t he c ourse w e w ill s tudy • The l aws o f t hermodynamics» 0, 1, 2, 3» These a re v ery g eneral• Phase t ransi1ons» Solid t o l iquid t o g as» Paramagne1c t o f erromagne1c» These a re c alled c ri1cal p henomena• Non-­‐equilibrium t hermodynamics» Heat t ransport• Kine1c t heory o f g ases» Explains p, V, T b y c onsidering d ynamics o f p ar1clesImportant c oncepts i n t hermodynamics • Thermodynamic s ystem• Dynamical s ystem w ith m any d egrees o f f reedom• Any m acroscopic s ystem i s a t hermodynamic s ystem • Environment = s urroundings = e verything e lse • Thermodynamic p arameters• Measurable m acroscopic q uan11es a ssociated w ith s ystem• Pressure, v olume, t emperature• Thermodynamic s tate• Specified b y a s et o f a ll v alues o f a ll t he t hermodynamicparameters n ecessary t o d escribe t he s ystemIsolated, c losed, a nd o pen s ystems • Isolated s ystem (idealized)• Does n ot i nteract w ith t he e nvironment i n A NY w ay• Separated b y a w all (par11on) w hich d oes n ot a llowexchange o f h eat o r p ar1cles (maVer)• Total e nergy, v olume a nd p ar1cle n umber a re c onserved • Closed s ystem• Can e xchange e nergy w ith t he e nvironment• Cannot e xchange p ar1cles w ith t he e nvironment• Par1cle n umber i s c onserved• Total e nergy i s n ot c onserved• If c losed s ystem i n e quilibrium w ith e nvironment» Total e nergy h as a verage v alue r elated t o t emperature o fenvironment» Can u se t emperature t o c haracterize s tate o f s ystemIsolated, c losed, a nd o pen s ystems • Open s ystem• Can e xchange e nergy a nd p ar1cles w ith t he e nvironment• Total e nergy n ot c onserved• Par1cle n umber n ot c onserved• If o pen s ystem i n e quilibrium w ith e nvironment» Total e nergy h as a verage v alue r elated t o t emperature o fenvironment» Par1cle n umber h as a verage v alue r elated t o t emperature o fenvironment» Can u se t emperature t o c haracterize s tate o f s ystem» Can a lso u se c hemical p oten1al t o c haracterize t he s ystem» We w ill d efine t he c hemical p oten1al l ater• Right n ow w e a re u sing t he w ord t emperature • Later w e w ill d efine i t p roperlyHomogeneous s ystems• Homogeneous s ystem• The p roper1es o f t he s ystem a re t he s ame f or a ny p art • Heterogeneous s ystem• The p roper1es o f t he s ystem c hange d iscon1nuously a tcertain s urfaces• A h eterogeneous s ystem h as h omogeneous p arts • Different p arts o f t he s ystem a re i n d ifferent p hases• Phase b oundary• One s ystem, d ifferent p hases• e.g. p ot c ontaining w ater, a ir a nd s team• There a re t wo p hases – l iquid p hase a nd g aseous p hase• The b oundary b etween t hem i s c alled p hase b oundaryThermodynamic e quilibrium• Think o f a n i solated s ystem• The s ystem i s l ej s tanding f or a l ong 1me• It c omes t o a final s tate w hich d oes n ot c hange • Thermodynamic p arameters a re c onstant• The s ystem i s i n t hermodynamic e quilibrium i f i ts thermodynamic s tate d oes n ot c hange w ith 1me • A t hermodynamic s tate i s a n e quilibrium s tate • Thermodynamic e quilibrium• Also c alled t hermal e quilibrium• Ojen j ust e quilibriumThermal e quilibrium o f t wo s ystems • Generally c onsider t wo i solated s ystems A a nd B • Bring A a nd B i nto c ontact w ith e ach o ther• Ajer a l ong 1me t he t otal s ystem A+B r eaches e quilibrium• Then A a nd B a re i n e quilibrium w ith e ach o ther• A a nd B a re a lso (separately) i n t hermal e quilibrium • Think o f a c losed s ystem• A c losed s ystem e xchanges h eat w ith t he e nvironment• We c an t hink o f t he e nvironment a s a h eat b ath• e.g. t ea o n t able• Ajer a l ong 1me t emperature o f t ea = r oom t emperature• Tea i s i n t hermodynamic e quilibrium w ith r oom• Macroscopic v ariables w hich h ave a d efinite value f or e ach e ach s tate o f t hermal e quilibrium • State v ariables a re d efined o nly i n e quilibrium • Can b e m easured o nly i n e quilibrium• Usually w e o nly n eed a f ew (e.g. 3-­‐4) s tate variables t o s pecify t he s tate o f a s ystem• Microscopic q uan11es a re N OT s tate v ariables • posi1on• velocity• momentum• Heat a nd w ork a re a lso N OT s tate v ariables• State v ariables c an b e • Pressure p• Volume V• Temperature T• Energy E (or U)• Entropy S• Par1cle n umber N• Chemical p oten1al μ• Total c harge Q• Total d ipole m oment P• Magne1za1on M• Refrac1ve i ndex ε• Viscosity o f a fluid• Chemical c omposi1onIntensive a nd e xtensive v ariables • Extensive q uan1ty• Scales w ith s ystem s ize• Propor1onal t o t he a mount o f m aVer i n t he s ystem• They a re a ddi1ve• Volume, e nergy, p ar1cle n umber, e ntropy• Intensive q uan1ty• Independent o f t he s ize o f t he s ystem• Independent o f t he a mount o f m aVer i n t he s ystem• Pressure, t emperature, d ensity• Also r efrac1ve i ndex a nd o thers• Can b e d efined l ocally (i.e. f unc1on o f r)• But m ost o f t he 1me w e a ssume t hem t o b e u niformEqua1on o f s tate• Func1onal r ela1onship b etween s tate v ariables • It d escribes a s ystem i n e quilibrium• For e xample i f p arameters a re p, V a nd T• There i s s ome f unc1on f(p, V, T) s uch t hatf(p,V,T)=0• Reduces n umber o f i ndependent p arameters f rom 3 t o 2• f i s g iven w hen t he s ystem i s s pecified = s tate f unc1on• State o f s ystem i s a p oint i n p, V, T s pace• Equa1on o f s tate d efines a s urface i n t his s pace• Any p oint o n t he s urface i s a s tate i n e quilibrium• Other p arameters – o ther s paces, p oints, s urface• Experimentally• All g ases b ehave i n a u niversal w ay w hen t hey a resufficiently d ilute = w hen t he d ensity i s l ow e nough • Ideal g as• Idealiza1on o f t his d ilute l imit• Par1cles a re p oint-­‐like (i.e. h ave z ero s ize)• No i nterac1ons b etween p ar1cles• An i deal g as i s a n i dealized s ystem• Parameters f or i deal g as• p, V, T, N• Boyle’s l aw (1664) – a t c onstant t emperaturepV= constant = pVpV =nRT • Ideal g as e qua1on o f s tate• Defines i deal g as t emperature s cale• Like h igh-­‐school p hysics• Ideal g as – H e a t v ery l ow d ensity• Measure p V/Nk o f i deal g as a t T a t w hich w ater b oils • Also m easure a t T a t w hich w ater f reezes• Draw a s traight l ine c onnec1ng t hem• Divide i nto 100 u nits – t his i s t he K elvin s cale• How t o u se• Bring o bject i n c ontact w ith i deal g as• Measure p V o f i deal g as• Read o ff t he t emperatureMore a bout t hermometers• Remember• Measuring t emperature i s r elated t o t he e qua1on o f s tate • To d efine a s cale• Choose T0 = 273.15K i n h onour o f K elvin• p0, V0 a re g iven i n y our t extbook• Celsius T C• Water f reezes = 0• Water b oils = 100• Fahrenheit T F• T F = 1.8 T C + 32• T F = T C at -­‐40 d egrees• Degrees C elsius a nd d egrees F ahrenheit• But K elvin, n ot d egrees K elvinZeroth l aw o f t hermodynamics • If• System A i s i n e quilibrium w ith s ystem B• System B i s i n e quilibrium w ith s ystem C• Then• System A i s i n e quilibrium w ith s ystem C• Two s ystems A a nd B i n e quilibrium• Func1ons o f s tate f A a nd f B• See b lackboard f or d eriva1on• Systems i n t hermal e quilibrium• They h ave a c ommon i ntensive q uan1ty – t emperature• Systems n ot i n e quilibrium h ave d ifferent t emperatures • See b lackboard f or d eriva1on• First t ake t he i deal g as e qua1on o f s tate• Real g as – p hysical c onsidera1ons• First, p roper v olume• V i s n ot a c onstant b ecause V → 0 a s T → 0• Change V t o V – N b, w here b i s s ome p arameter• Second, i nterac1ons b etween p ar1cles• Interac1on i s m ainly a Vrac1ve• Consider a g lobe w ith a d ensity N/V• Inside t he g lobe t he a verage f orce b etween p ar1cles i s 0• But n ear t he w alls t his i s n ot t rue• Par1cles o n s urface f eel n et f orce t owards i nside• So p f or r eal g as i s s maller t han f or i deal g as• p ideal = p real + p0Exact a nd p ar1al d ifferen1als• State f unc1on = f unc1on o f m any v ariables • We m ake u se o f e xact a nd p ar1al d ifferen1als• For e xample i n e quilibriumf(p,V,T)=0• This m eans w e c an r egard p=p(V,T), V=V(p,T) a nd T=T(p,V) • Differen1al a lso c alled d eriva1ve• Total d eriva1ve, t otal d ifferen1al• See b lackboard f or d eriva1onTypes o f p rocesses• Infinitesimal p rocess• Difference b etween i ni1al a nd final s tate i s i nfinitesimal • Quasi-­‐sta1c p rocess (ideal p rocess)• System a nd s urroundings m aintain t hermal e quilibrium• Change m ust b e s low e nough – S LOW p rocess• Quasi-­‐sta1c p rocesses a re r eversible – a p rocess i sreversible i f i t r etraces i ts h istory i n 1me w hen t he e xternalcondi1ons r etrace t heir h istory i n 1me• Isothermal• Constant t emperature• Adiaba1c• No t hermal c ontact w ith t he s urroundings• However w ork i s d one o n t he s urroundings o r b y t hesurroundingsThermocouple 热电偶• Seebeck effect – t hermoelectric effect• Temperature d ifference g enerates v oltage i n c onductor• Join s econd c onductor t o m easure t he v oltage• Second c onductor h as o pposite effect• But i f m aterial i s d ifferent n et c urrent flows ∝ T• Used a s a t emperature s ensor• Useful b ecause• Ac1ve o ver w ide r ange o f t emperatures – 000s o f d egrees• Powers i tself• However• Not v ery a ccurate – n ot s ensi1ve b y m ore t han 1 d egree• So y ou c an u se i t w hen y ou d o n ot n eed m uch p recision• e.g. g as t urbine, d iesel e ngine, k iln (=oven), i ndustrySummary• Isolated, o pen a nd c losed s ystems• Thermodynamic e quilibrium• Intensive a nd e xtensive v ariables• State v ariables a nd e qua1on o f s tate • Ideal g as – n o i nterac1ons• Van d e W aals g as – w eak i nterac1ons • Temperature – t hermodynamic d efini1on • Thermometer• Assignments w ill b e g iven n ext w eek。

序号论文题目作者1Mandatory Clearing of Derivatives and Systemic Risk of Bank Holding CompaniesShaofang Li and Matej Marinč2Heterogeneous preferenc es and risk sharing at households level in China Jennifer T. Lai, Isabel K. M. Yan, XingjianYi3Financial Globalization and the International Transmission of Interest RateShocks: The Federal Reserve and ChinaXiaoli Wan4Information Asymmetry in Quasi Public Goods Crowdfunding-A Case fromChina's EV Charging PileMarket Yan Li, Qi Zhang, Ge Wang, Siyuan Chen5Microstructure Noise,Fat-Tailed Distribution of Stock Returns, and Idiosyncratic Volatility Puzzle Shouyu Yao, Zhenming Fang, Xin Cui, Chunfeng Wang6Mandatory CSR Disclosure, monitoring and investment efficiency: Evidencefrom a quasi-natural experiment in ChinaLi Liu, Gary Gang Tian7Linguistic Distance and M ergers and Acquisitions:Evidence from China Lu Li, Yang Duan, Yuqian He, Kam C. Chan8Failure and Rescue in C entral Clearing Counterparty Design Zhenyu Cui, Chihoon Lee, 刘彦初, Kai Wang9Margin Trading and Asset Pricing-Evidence from tShujing Wang, K.C. John Weihe Chinese Stock Market1 0Asset Pricing With ReturnExtrapolationLawrence J. Jin, Pengfei Sui1 1Familiarity Bias? The Role of Directors with Country-specific Experience inCross-Border Mergers and AcquisitionsHaoyuan Ding, Yichuan Hu, Chang Li1 2汇率制度改革与资本账户开放路径选择——基于开放经济DSGE 模型的分析彭红枫, 肖祖沔, 祝小全1 3How do bank policies affect the equity risk and cost of capitalMike Qinghao Mao，K.C. John Wei1 4Optimal Capital Structure,Investment and Debt-Equity SwapEi Yang, Shaoxuan Zheng1 5Sales, Monetary Policy, and Durable GoodsWenbin Wu1 6The crude oil-stock market link and its determinants: Evidence from emerging economiesXiaoqian Wen, Hua Cheng, Yihao Zhang1 7Dependence structure and risk spillovers betweenoil and stock markets from BRICS countriesYonghong Jiang(姜永宏), He Nie(聂禾), Bin Mo(莫斌)1 8Asymmetric Variance Premium, Skewness Premium, and the Cross-Sectionof Stock ReturnsTao Huang, Junye Li, Fei Wu1 9Top managerial power and stock price efficiency:Evidence from ChinaMeifen Qian, Ping-Wen Sun, Bin Yu2 0Leverage Risk and Intermediary Asset Pricing: Evidence in ChinaXu Feng, Yajun Xiao2 1Responding to financial crisis Bank credit expansion with Chinese characteristicsLongyao Zhang，Enjiang Cheng，Sara Hsu2 2External Source of Political Connection: FinancialAdvisor and Chinese AcquisitionXiaoGang BI, Danni WANG2 3Switching Due DiligenceAuditors in Chinese Mergers and AcquisitionsJie Tang, Xiaogang Bi2 4Information environment,systematic volatility and stock return synchronicityJing Wang, Steven X. Wei, Wayne Yu2 5A tail of two worlds: Stock crash, market context,and expected returnsS. Ghon Rhee, Feng Wu2 6Illiquidity Shocks and Asymmetric Stock Market Reactions around the WorldIs Underreaction or Illiquidity Spiral the CulpritTe-Feng Chen，K.C. John Wei2 7Venture Capital, Innovation and Post-IPO Long-runPerformanceXueyong Zhang, Yeqing Zhang2 8Household Debt, Aggregation and Asset Pricing PuzzlesGaosheng Ju, Qi Li2 9Competition, Syndication,and Entry in the VentureCapital MarketSuting Hong3Underwriter Reputation a QiuyueZhang, XueyongZhang0nd Post-IPO Price Perfor mance: New Evidence from IPO Fraud of ChineseListed Firms3 1Optimal Reinsurance andInvestment Problem withDefault Risk and Bounded MemoryChao Deng, Wenlong Bian, Baiyi Wu3 2Party Committee Secretary serving on corporate boards and firm bribery channelsHamish D. Anderson, Jing Liao, Jingjing Yang, Martin Young3 3Index tracking model, downside risk and non-parametric kernel estimationJinbo Huang, Yong Li, Haixiang Yao3 4Financial Stability with Fire-salesRyuichiro Izumi, Yang Li3 5Correlation Ambiguity, Listing Choice and MarketMicrostructureJunyong He（何俊勇）, Helen Hui Huang,Shunming Zhang, Wei Zhu3 6Macroeconomic Conditions and Stimuli of InvestmentYingxian Tan, Xin Xia, Jinqiang Yang3 7Financial Frictions and Trade DynamicsPaul Bergin, Ling Feng, Ching-Yi Lin3 8The Determinants of Dynamic Linkage between Stock Prices and Foreign Exchange Rates: Evidencefrom ChinaR.W. Lin3 9Horizon-unbiased Investment with Ambiguous VolatilityQian Lin, Xianming Sun4 0Time-Varying Ambiguity and Bank Lending吴伟劭, Sandy Suardi4 1Business Cycle and theValue of Cash_International EvidenceJiaxing You，Juanjuan Huang，Ling Lin，Min Xiao4 2Monetary Regulation andLiquidity Distribution Mechanism between the RealEconomy and Capital MarketMingli Xu, Hao Wang, Jisheng Yang4 3Understanding AH Premium in China Stock MarketRenbin Zhang, Tongbin Zhang4 4How Do Labor Unions Affect Payout PolicyBeibei Shen4 5Determinants of CoCo Bond Issuance InternationalEvidenceChenyu Shan，Dragon Yongjun Tang，Meng Xie4 6Dissecting the Long-termPerformance of the Chinese Stock MarketFranklin Allen，Jun “QJ” Qian，Chenyu Shan，Julie Lei Zhu4 7Credit Default Swaps andDebt OverhangTak-Yuen Wong, Jin Yu4 8LeBow College of BusinessBang Nam Jeon, Ji Wu4 9A Conditional Valuation of Corporate Pension PlansJun Cai, Miao Luo, Alan J. Marcus5 0Interest rate liberalizationand bank pricing in credit marketYonghui Chen, Xuezhong He, Yan Zeng5 1Rational Learning and Trading Behavior in Limit Order MarketsXue-Zhong He,Shen Lin5 2Expropriation and the Sensitivity of Investment toZheng WangCash Flow-A Natural Exp eriment from China's Split -Share Structure Reform5 3Investor attention and commonality in asset pricing anomaliesLei Jiang (姜磊) , Jinyu Liu (刘津宇) , LinPeng (彭琳) , Baolian Wang (王宝链)5 4Does Privatization Reform Alleviate Ownership Discrimination? Evidence from the Split-share Structure Reform in ChinaJinyu Liu (刘津宇) , Zhengwei Wang (王正位) , Wuxiang Zhu (朱武祥)5 5Leverage-Induced Fire Sales and Stock Market CrashesJiangze Bian, Zhiguo He, Kelly Shue, HaoZhou5 6Solving the High-dimensional Markowitz Optimization Problem: When Sparse Regression Meets Random Matrix TheoryMengmeng Ao, Yingying Li, Xinghua Zheng5 7Disclosure quality, price efficiency, and expected returnsKung-Cheng Ho, Shih-Cheng Lee, Ping-Wen Sun5 8有限理性对开放式基金“业绩-流量关系”的影响伍燕然, 王凯5 9Product Market Structureand Nepotism in Bank Loan MarketsHua Cheng, Yan Dong, Xue Li6 0是谁影响了股价崩盘风险：有形信息VS 无形信息——基于投资者行为视角的证据史永东杨瑞杰6 1动态银行网络系统的系统性风险研究范宏，杨明6 2信贷约束, 人口红利和经济增长薛熠, 张悦, 马倚虹6 3社会资本, 非正式保险与家庭的股市参与吴卫星, 祁震6 4我国债券评级行业竞争变化对评级质量影响的实证研究黄晓薇, 安小雪6 5一种动态信用风险传染模型及其应用陈典发, 邓军, 冯建芬6 6媒体, 公司治理与上市公司财务欺诈行为孙艳梅, 汪昌云6 7模糊厌恶下的动态最优投资策略张金清, 金泽宇, Yunbi An6 8中国特色的市场微观结构模型——基于准入门槛和政府调控的多市场均衡分析刘红忠, 毛杰6 9央行沟通, 政策不确定性与通胀预期王少林7 0理财产品膨胀, 利率市场化与银行风险承担问题研究项后军, 闫玉7 1Uncovering the myth of the housing price in Chinese metropolises: Allowing for behavioral heterogeneity among investors张玲, 边文龙, 张浩7 2Anticipating critical transitions of Chinese housingmarketsQun Zhang（张群）, Didier Sornette, HaoZhang7 3房价上涨, 家庭债务与居民消费周利7 4族群分配政策会影响公司价值吗？——来自马来西亚土著股权配额制的经验证据范祚军, 唐菁菁, 曾海舰7Macroeconomic Trade Eff Makram El-Shagia and Lin Zhanga5ects of Vehicle Currencie s: Evidence from 19th Century China7 6高管薪酬粘性仅仅是奖优不惩劣吗？——基于非委托代理视角王修华, 谷溪7 7P2P 网贷平台的声誉依赖与信息中介的制度约束戴晓凤, 雷宇7 8P2P平台只做信息中介可行吗？刘轶, 王于栋7 9创伤经历, 风险偏好与家庭资产选择——基于全国基线调查微观数据的证据乔海曙, 粟亚8 0市场竞争, 银行市场势力与流动性创造效率——来自中国银行业的经验证据李明辉, 黄叶苨8 1关联博弈, 声誉积累与新创企业的破茧成蝶米运生, 周明明, 石晓敏, 黄斯韬8 2基于改进粒子群模糊神经网络的信用评估研究熊志斌8 3Model-free estimation of tail risk and moments using option pricesCarole Bernard, Chengli Zheng8 4劳动调整成本, 经济增长与货币政策的就业效应杨柳8 5流动性创造, 银行稳定性与经济增长：来自中国商业银行的经验证据宋琴, 郑振龙8 6金融发展对中国企业资本结构的影响——基于资本供应视角郑承利, 秦妮8 7最终控制人投资组合与风险分担：转嫁还是共担？——来自民营上市公司的王红建, 汤泰劼, 刘梓微经验证据8 8微博, 自愿信息披露与上市公司盈余信息质量胡军, 王甄, 胡援成8 9我国机构投资者的投资行为理性吗？——基于主成分分析法的研究吕江林, 李兴9 0股票信息披露与投资者异质信念定价模型刘维奇, 李林波9 1Pollution and Performance: Do Investors Make Worse Trades on Hazy Days?俞红海, 黄杰鲲, 许年行9 2影子银行, 行政监督与国家审计治理——基于国务院107号文与审计公告的准自然实验王家华, 曹源芳9 3中国金融市场联动特征与体系性风险识别何枫, 田利辉9 4法律保护, 社会信任与商业银行全要素生产率李双建, 李俊青, 刘凯丰9 5金融素养存在性别差异么？——基于婚姻分工,认知水平和风险态度的实证分析廖理, 初众, 张伟强9 6异质波动率之谜是否已被完全解释？陈湘鹏, 周皓, 金涛9 7系统性金融风险与SRISK的适用性陈湘鹏, 周皓, 王正位, 金涛9 8Fuel the Engine: Bank Credit and Firm Innovation戚树森9 9Evaluating Asset PricingFactorsYu Ren, Yue Qiu, Tian Xie1为什么人民币越来越重周颖刚, 王艺明, 程欣0 0要？-基于网络分析方法的汇率证据1 0 1考虑时变风险溢酬的市场有效性检验：基于上证50ETF 期权市场陈蓉, 黄帅1 0 2宏观经济政策与股市系统性风险——宏微观混合β估测方法的提出与检验邓可斌, 关子桓, 陈彬1 0 3企业的合作文化促进创新了吗？潘越, 潘健平, 马奕涵1 0 4银行竞争背景下定向降准政策的“普惠”效应——基于主板和新三板的三农, 小微企业数据的分析郭晔, 徐菲, 舒中桥1 0 5Do Social Connections Mitigate Hold-up? Evidencefrom Relation-Specific Investment and Innovation in Vertical RelationshipsSudipto Dasgupta, Kuo Zhang, Chenqi Zhu1 0 6“好”的不确定性, “坏”的不确定性与股票市场定价——基于中国A 股高频数据的研究陈国进, 丁杰, 赵向琴1 0 7金融发展, 法律保护与危机传染——中国特征与国际比较游家兴, 张哲远18IPO信息披露的可读性分析黄方亮, 王彤彤1 0 9When Are Stocks Less Volatile in the Long Run?Eric Jondeau, QunzhiZhang, Xiaoneng Zhu1 1房价波动, 楼市调控与宏观金融风险——基于未定权曹廷求, 高睿0益分析法的研究1 1 1P2P 网络借贷市场中高收入真的低风险吗？胡金焱, 张笑1 1 2文化认同与民间金融：基于方言视角的经验研究张博, 胡金焱1 1 3互联网：经济活动的空间“离心力” ——来自中国家庭创业的证据曹廷求, 王可1 1 4P2P 网络借贷与大众创业——基于“人人贷”微观数据的经验研究胡金焱, 李建文1 1 5经济周期视角下的晋升压力与地方政府债务融资行为司海平, 刘小鸽, 魏建1 1 6基于微观复杂网络与宏观金融工程的风险传染研究沈沛龙, 李志楠1 1 7股票流动性, 投资者关注与公司创新张信东, 原东良1 1 8Local government debt scale estimating and riskevaluating based on localgovernmental balance sheetBaijie Wang, Sihan Zhang1 1 9“水至清则无鱼”适用于影子银行监管吗？——基于国内银行理财产品业务监管套利和透明度的研究刘莉亚, 黄叶苨, 周边1 2 0钱荒的根源：经济基本面变化，还是货币政策冲击？陈华, 郑晓亚, 赵自然1 2 1投资者“听话听音”能获得超额收益吗？-----从中国上市公司投资者电话交流会中高管声音情绪视角出发刘莉亚, 陈瑞华, 闵敏, 朱小能1 2 2“与众不同者”的价值——基于基金独特性的视角徐龙炳, 顾力绘1 2 3Optimal consumption withstochastic income and time-inconsistent preferences王远平, 牛英杰, 杨金强1 2 4卖方分析师研报有价值吗？——来自评级修正的证据安郁强, 陈选娟1 2 5碳排放权交易与企业研发-来自中国的证据马文杰, 杨奕, 赵晓菊1 2 6金融冲击, 房地产和其它实体投资, 宏观经济波动冯玲, 葛璐澜, 李志远, 柳永明1 2 7从“一带一路”货币汇率参照篮子中看推进人民币国际化的要素丁剑平, 方琛琳, 叶伟, 张冲1 2 8我国企业年金集合计划的多重委托代理问题及其影响研究初立萍, 陈选娟, 曾韵1 2 9定向增发折价：风险投资的作用及其机制李曜, 宋贺, 龙玉1 3 0基于投入占用产出技术的部门系统重要性研究覃筱, 祁伊1 3 1Margin Trading and PriceEfficiency—InformationContent or Price-Adjustm吕大永, 吴文锋ent Speed1 3 2同业交流与创新表现：基于期刊举办学术会议视角曹志奇, 吴文锋, 舒海兵1 3 3Economic Policy Uncertainty and Stock Price Crash Risk金雪军, 陈紫晴, 杨晓兰1 3 4融资融券提高了股价信息含量吗？白俊, 宫晓云, 孟庆玺1 3 5金融资产收益率的人口结构悖论：来自递延储蓄和替代效应视角的解释易祯, 朱超1 3 6高管限薪与公司业绩——来自2015年“限薪令”自然实验的证据尹志超, 蒋佳伶, 冯瑞河, 栗媛1 3 7开放经济条件下我国货币政策选择的理论与经验分析刘尧成, 庄雅淳1 3 8Why do financially constrained suppliers provide trade credit in China? ——An extended redistribution view田钢, 于博1 3 9“股市”还是“赌市” ：基于特质性偏度的投资者博彩偏好定价研究王春峰, 姚守宇, 房振明, 崔欣1 4 0Dynamic effects of Analyst Forecasts on Information Asymmetry comparingwith Company’s Announcements in Chinese Security Market李洋, 向健凯, 王春峰, 房振明1 4雾霾天气与投资者决策-来自互联网论坛的证据张维, 孟祥桐, 冯绪11 4 2利差扭曲与最优金融稳定政策何国华, 李洁, 刘岩1 4 3零利率下限, 汇率传递与货币政策王胜, 周上尧1 4 4银行规模, 综合化经营与我国银行业的系统性风险张天顶1 4 5高管薪酬与资本结构动态调整罗琦, 谢辰, 应惟伟1 4 6股东持股, 技术溢出与企业创新崔静波, 郑旸1 4 7新型流动性指标与流动性溢价杨兴哲, 周翔翼1 4 8超额准备金, 货币政策传导机制与调控方式转型——基于银行信贷市场的分析王晓芳, 郑斌1 4 9投资者羊群行为总是不利于市场运行吗？——来自中国艺术品市场的证据石阳, 刘瑞明1 5 0经济政策不确定性与企业贷款成本宋全云, 吴雨, 钱龙1 5 1Maturity Mismatch and Incentives: Evidence fromBank Issued Wealth Management Products in ChinaJinjin Liu, Hongyan Fang, Ronghua Luo,Senyang Zhao1存贷比约束, 货币政策与银王擎, 李元52行信贷1 5 3When Interim CEOs areNamed Formal CEO: A Try-out Succession何潇潇1 5 4尖刀上的舞蹈——民间借贷债务催收涉黑化倾向的田野调查与经济解释丁骋骋1 5 5股市政策对股票市场的影响——基于投资者社会互动的视角杨晓兰, 王伟超, 高媚1 5 6关联关系, 融资约束缓解与投资动态调整—来自担保贷款的融资与风险权衡张小茜1 5 7A Least Squares Regression Realized CovariationEstimation under MMSIngmarNolte, MichalisVasios, ValeriVoev,QiXu1 5 8遗产继承会扩大财富差距吗？——来自中国健康与养老追踪调查的证据韦宏耀, 钟涨宝1 5 9“新常态”下省际人民币汇率传递效应的非对称性和非线性及其影响因素研究---基于多维DCC-GARCH模型曹伟, 钱水土, 万谍1 6 0Do Non-Discretionary Allocation and Uniform-Pricing Rule Really Matter for Information Production in IPOsBo Liu, Xinru Ma1 6 1“险资举牌”现象与分散股权时代我国上市公司的治理郑志刚, 石丽娜, 黄继承, 郭杰16企业过度负债的同群效应李志生, 孔东民, 苏诚, 李好21 6 3会计准则具有去杠杆的作用吗？——来自上市公司资本结构决策的微观证据晏超1 6 4我国金融部门宏观金融风险研究——基于重构资产负债表和CCA 模型的分析白小滢, 王乔乔1 6 5经济发展水平对跨境资本流动的门限效应——兼评中国“非理性”对外投资周先平, 谭本艳, 向古月1 6 6An Experimental Study on a Novel Artificial Financial Market Model曹宏铎, 李旲, 欧阳辉1 6 7社会保险和大众创业：“创业枷锁”还是“创业催化剂”？王永钦, 戴芸, 包特1 6 8利率市场化, 银行利率与货币政策价格型调控陆军, 黄嘉, 童玉芬1 6 9房地产金融化：界定, 表现及其经济效应李建军, 韩珣1 7 0The role of foreign and domestic venture capital ininnovation evidence fromChinaJiangjing Que，Xueyong Zhang1 7 1The Dynamic Co-movement between the Exchange Rates of European Countries and Chinese Renminbi黄乃静, 汪寿阳1 7 2经济政策不确定性与企业投资：实物期权渠道还是金融摩擦渠道？谭小芬, 张文婧1 7 3Liquidity Risk and Corporate Risk-takingJing-zhi Huang, Huayi Tang, Yuan Wang,Rui Zhong1 7 4CDS Trading and StockPrice Crash RiskJinyu Liu, Jeffrey Ng, Dragon Yongjun Tang, Rui Zhong1 7 5Measuring Financial RiskContagion among International Stock Markets: AEMD-Copula-CoVaR ApproachDa Wang, Qiang Zhang, Changqing Luo1 7 6去杠杆, 转杠杆与货币政策传导周俊仰, 汪勇1 7 7Household Financial Constraint and Local BankingStructureYiyi Bai, Zhisheng Li1 7 8人民币汇率与境内外股票市场波动溢出效应研究——基于三元BEKK-GARCH模型吴丽华, 江蓝微, 曾连彬1 7 9The long term impact of the 1959-1961 famine inChina on stock holdingsof survivorsJennifer T. Lai, Xingjian Yi, Cong Zhou1 8 0金融杠杆与资产价格泡沫：理论机制及其非对称效应刘晓星, 石广平1 8 1银行业竞争融资约束与企业创新来自银监会与企业专利申请数据的证据李子健, 李春涛1 8 2VPIN, volatility and the 2015 stock market crash in China Evidence from the SSE 50 index derivative marketsYuqin Huang, Xingguo Luo, Shihua Qin，Libin Tao1 8 3Baidu News InformationFlow and Return Volatility-Evidence for the SIAH沈德华, 李晓, 张维1 8 4官员晋升、人力资本错配与企业劳动投资效率孔东民, 项君怡, 刘莎莎1 8 5货币政策、银行所有制与影子银行王擎, 李川, 盛夏1 8 6契约条款与债券融资成本的相互选择史永东, 王彤彤, 田渊博1 8 7货币政策、同业业务与银行流动性创造郭晔, 程玉伟, 黄振1 8 8The Role of Analysts: AnExamination of the Idiosyncratic Volatility Anomaly in the Chinese Stock MarketMing Gu, George J. Jiang, and Bu Xu1 8 9Cross-Ownership in Inter-Corporate Loans in China戚树森。