J3.4 Ensemble Kalman filter assimilation of Doppler radar data
with a compressible nonhydrostatic model
Mingjing Tong and Ming Xue*
School of Meteorology and Center for Analysis and Prediction of Storms
University of Oklahoma, Norman, OK
1. Introduction
*Since its first introduction by Evensen (1994), the ensemble Kalman filter (EnKF) technique for data assimilation has received much attention. A number of studies have been done to exploit its applications and performances. The EnKF was designed to sim-plify the computation of the flow-dependent error sta-tistics without the use of approximate closure scheme as extended Kalman filter does. Rather than solving the equation for the time evolution of the probability density function of the model state, EnKF applies the Monte Carlo method to estimate the forecast error statistics. A large ensemble of model states are inte-grated forward in time using the dynamic equations, the moments of the probability density function are then calculated from this ensemble of model states for different times (Evensen 2003).
In the recent decades, various techniques have been developed for analyzing and retrieving atmos-pheric state at the convective scale from Doppler ra-dar data. These methods range from purely kinematic to expensive 4D variational method that employs a nonhydrostatic prediction model and its adjoint (e.g., Gal-Chen 1978; Sun et al. 1991; Qiu and Xu 1992; Shapiro et al. 1995; Sun and Crook 1997; Gao et al. 1999; Wu et al. 2000; Weygandt et al. 2002). Most of the latter work deals with retrieval and assimilation of radial velocity and/or reflectivity data from single Dop-pler radar. For the purpose of initializing NWP models, the 4DVAR method (e.g., Sun and Crook 1997; Gao et al. 1998) promises to provide an initial condition that is consistent with the prediction model and is able to effectively use multiple volume scans from radar. However, due to the need for an adjoint that should include detailed physics parameterizations and the high computational cost, 4DVAR assimilations of Doppler radar data have been limited to using rela-tively simple model configurations.
Compared to the 4DVAR method, the EnKF scheme is more flexible and much easier to set up. Under the right assumptions, its solution is equivalent to the optimal 4DVAR analysis. In fact, EnKF has recently been applied to the assimilation of simulated Doppler radar data for a convective storm (Snyder and Zhang 2003; Zhang et al. 2003) and of real radar data by Dowell et al. (2003). All three studies used the same anelastic cloud model of Sun and Crook (1997). Simple warm-rain microphysics scheme is used in these as well as afore-quoted 4DVAR studies with Wu * Corresponding author address:
Dr. Ming Xue, SOM, SEC 1310, 100 E. Boyd, Norman OK 73019,mxue@https://www.doczj.com/doc/7614318057.html,. et al (2000) being one exception. In the latter, the ice mi-crophysics scheme used was simplified and microphysical variables, among others, were analyzed from dual-polarization radar data. In the study, water and ice phase microphysical variables are first derived from the polariza-tion reflectivity data before being assimilated into the model.
In this study, we report on the development of an EnKF system based on a general-purpose compressible nonhydrostatic model, and the application of the system to the assimilation of simulated single Doppler radar radial velocity and/or reflectivity data. A sophisticated ice micro-physics scheme is employed. The performance of the EnKF scheme in 'recovering' complete model structures, including wind, temperature, pressure fields and all water and ice categories are examined. The impact of radial velocity and reflectivity data as well as their spatial cover-age on the analysis are investigated. Section 2 describes the EnKF assimilation system and design of OSS (Observ-ing System Simulation) experiments, and section 3 pre-sent and discussed the experiment results. A concluding section is given at the end.
2. Assimilation System and Experimental Design
a) The prediction model and truth simulation
In this study, we test our EnKF assimilation system using simulated data from a classic May 20, 1977 Del City, Oklahoma supercell storm case (Ray et al. 1981). Such simulation experiments are commonly referred to as Ob-serving System Simulation Experiments (OSSE, see, e.g., Lord et al. 1997). The forecast model used is the Ad-vanced Regional Prediction System (ARPS; Xue et al. 2000), In this study, the ARPS is used in a 3D cloud model mode and the prognostic variables include three velocity components ,,,
u v w potential temperature θ, pressure p, and six categories of water substances (water vapor spe-cific humidity q v,cloud water mixing ratio q c, rainwater mix-ing ratio q r, cloud ice mixing ratio q i, snow mixing ratio q s and hail mixing ratio q h). The microphysical processes are parameterized using the modified three-category ice scheme of Lin et al. (1983) and its implementation follows Tao and Simpson (1993).
For all experiments unless otherwise noted, the physi-cal domain is 646416
km km km
××. The model grid com-prises of 353535
××grid points, with grid intervals of 2 km in both x and y directions and of 0.5 km in the vertical. The truth simulation or nature run was initialized from a modi-fied real sounding plus a 4K ellipsoidal thermal bubble centered at48
x km
=, 16
y km
= and 1.5
z km
=, and with radius of 10km in x and y and 1.5km in vertical direc-tions. Open boundary conditions are used at the lateral
boundaries. A radiation condition is also used at the top boundary. Free-slip conditions are applied to the bottom boundary. The length of simulation is 75 min-utes. A constant wind of 3u = ms -1 and 14v = ms -1 is subtracted from the original sounding to keep the primary storm cell near the center of model grid. De-spite a relatively coarse resolution, the evolution of the simulated storms is very similar to those docu-mented in Xue et al. (2001).
b) Simulation of radar observations
The simulated observations are assumed to be available on the grid points. The radial velocity is cal-culated from cos sin cos cos sin r V u v w αβαβα=++ , (1) where αis the elevation angle and βthe azimuth angle of radar beams, and u, v and w are velocities from the simulation.
The logarithmic reflectivity is estimated from equations as follows:
10log()dBZ Z =, (2) r s h Z Z Z Z =++, (3) where Z r , Z s , Z h are contributions from rain, snow and hail. The rain component of the reflectivity is calcu-lated from
1.75
1.75
1.7560.75720()
(8.010)r r r C q Z ρπρ=× , (4) where C is a constant, 31000/r kg m ρ=is the density of rainwater, ρis the density of air. If the temperature is less than zero centigrade then the snow component of reflectivity is
0.25
1.75
2
1.7560.757200.176()0.93(310)s
s s i C q Z ρρπρ×=
×, (5)
otherwise ,
1.75
1.75
1.7560.75720()
(310)s s s
q C
Z ρπρ=
×, (6)
where 3100/s kg m ρ=is the density of snow and
3917/i kg m ρ=is the density of ice. The hail compo-nent is calculated from
0.95
1.6625
1.7540.75 1.75
720()(4.010)h h h
C Z q ρπρ=×??????, (7) where 3
913/h kg m ρ=is the density of hail.
The radar is located at the southwest corner of the computational domain. For the data sampling and data assimilation, we assume the observation opera-tor to be perfect. As with most 4DVAR and EnKF studies, the prediction model is also assumed perfect, i.e., no model error is explicitly taken into account.
c) The EnKF data assimilation procedure
Our EnKF implementation is based on the algorithm described by Evensen (1994), Burgers et al. (1998) and Houtekamer et al. (1998). The analysis equation is
1[]()a f f T f o f i i i i x x P H HP H R y Hx Τ?=++?, (8) where i represents the i th ensemble member and f i x is the first guess obtained from the i th ensemble forecast. f P is the forecast error covariance and R is the observation error covariance. Perturbed sets of observations are used to update each ensemble member and o i y is the i th per-turbed observation. H is the observation operator, which converts the model states to the observation parameters. The forecast error covariances are calculated from
1
()()1
N
f
f
f f f i
i i
P H x x Hx Hx N Τ
Τ
=
???∑ ,
(9)
and
1
()()1
N
f
f f f f i i i
HP H Hx Hx Hx Hx N Τ
Τ
=
???∑,
(10)
where the overbar denotes the ensemble mean and N is the number of ensemble members.
We start the initial ensemble forecast at the 20 min-utes of the model simulated storm. To initialize the en-semble members, Gaussian noises with zero mean are added to the horizontally homogeneous initial guess that is based on the environmental sounding. The standard de-viation of the random noises is 3 ms -1
for u , v and w and 3K for potential temperature. The pressure and moisture fields are not perturbed. The first analysis is performed at 25 minutes. One hundred ensemble members are used for the assimilation experiments.
The observations are assimilated every 5 minutes. The observation errors are assumed to be uncorrelated; therefore, observations can be and are analyzed sequen-tially one at a time. We limit the influence region of each observation to a rectangular region with half width of 2 grid intervals in both horizontal and vertical directions, a pro-cedure known as covariance localization. Observations are perturbed by adding Gaussian noises, with the stan-dard deviations being 1 ms -1
for radial velocity and 5 dBZ for reflectivity.
Table 1 gives a list of ten experiments reported in this paper. First the assimilation scheme is tested by only as-similating the simulated radial velocity data with full data coverage or data covering regions with significant reflectiv-ity only. The impact of using reflectivity data is evaluated among other experiments.
3. The Assimilation Experiments
a) Assimilations using radial velocity data only
In experiment VrFull we assume that the radial veloc-ity data cover the entire computational domain. In experi-ment VrCloudy the same data are available only in cloudy regions where reflectivity is greater than 10dBZ. Although in real cases, the radial velocity data are generally un-available or are unreliable outside the cloudy regions, we would like to see how data coverage impacts the quality of EnKF assimilation.
Table 1. List of Data Assimilation Experiments
Experiment Observation: Radial velocity (Vr) and/or Reflectivity (Z) Update pres-sure
Update q i , q s , q h Update u , v , w , q v , q c , q i when assimilating
reflectivity
VrFull Vr yes yes VrFLD The same as VrFull, except for a 2.5 times as large a domain in the y direction VrCloudy Vr (Z > 10 dBZ) yes yes VrCnop Vr (Z > 10 dBZ) no yes ZCloudy Z (Z > 10 dBZ) no yes yes VrZCa Vr & Z (Z > 10 dBZ) no yes yes VrZCb Vr & Z (Z > 10 dBZ) no yes no VrZCc Vr & Z (Z > 10 dBZ) no yes yes, start from 4th cycle VrCZfull Vr (Z > 10dBZ) & Z no yes yes, start from 4th cycle VrCnoice Vr (Z > 10 dBZ) no no
T=40 min
T=60 min T=80 min
T=100 min
T r u e
V r F u l l
V r C l o u d y
(km)
(km)
(km)
(i)(j)(k)(l)
Figure 1. Vertical velocity (m s -1; shaded), horizontal wind vectors (m s -1
), and perturbation poten-tial temperature 'θ(K; contour) at z =6 km for truth simulation (a)~(d), experiments VrFull (e)~(h) and VrCloudy (i)~(l) at T=40, 60, 80 and 100 min.
T=40 min
T=60 min T=80 min T=100 min
T r u e
V r F u l l
V r C l o u d
y
(km)
(km)
(km)
(b)(c)(d)
(f)(g)(h)
(i)(j)(k)(l)
Figure 2. The vertical cross-sections of wind vectors (m s -1
), perturbation potential temperature (K; shaded), which pass through the maximum updraft, for the truth simulation (a)~(d), experi-ment VrFull (e)~(h) and experiment VrCloudy (i)~(l) at T=40, 60, 80 and 100 min.
As one can see by comparing with the truth fields, the wind, temperature, and microphysical variable fields can be accurately retrieved by experiment
VrFull (Figs. 1, 2, 10). At 80 minutes (the end of 12th
assimilation cycle), the maximum updraft is 40.5 ms -1,
which is very close to the 40.7 ms -1
of truth simulation (Fig. 2 though the specific maximum values are not shown). The strength of the low level cold pool, repre-sented by the minimum perturbation potential tem-perature at the first scalar level (250m) above ground, is -5.42 K for the simulation and -5.0 K for experiment VrFull (not shown). The overall structure and evolu-tion of the assimilated storm is very close to the true storm at this stage, indicating that the assimilation of radial velocity data using the EnKF scheme is suc-cessful.
Compared to VrFull, VrCloudy is also able to es-tablish the basic structures of the model storm well. The evolution of the model storm, including the cell splitting between 40 and 60 minutes and further split-ting after 80 minutes are all well reproduced (Fig. 1). However, the strength of the updraft, the edge and strength of the low level cold pool (not shown) and also the distribution of the six water substances are not as accurate as in VrFull (Figs. 1, 2 and 10). Since we added perturbations to the initial guess every-where and covariance localization was made, the envi-ronmental state in the clear air regions can not be updated by the analysis directly when data are limited to the cloudy regions. The model fields are therefore somewhat noisy. In real cases, other data representative of the environment can be used to improve the environmental analysis.
We use the rms error of the mean of ensemble analy-ses and forecasts to judge the quality of the mean analysis. The rms errors are averaged over those grid points where the reflectivity is greater than 10 dBZ. The errors for VrFull and VrCloudy are plotted in Fig. 3. The rms errors of ve-locities, temperature and hydrometeor variables are seen to decrease rapidly in the first four assimilation cycles (over 20 minutes). The rms errors are much smaller when the data covers the entire domain.
The EnKF data assimilation scheme works well for all, except pressure fields. Figure 3(e) shows that for the case when pressure field is analyzed (updated by the analysis), the pressure field becomes less accurate than the back-ground forecast after each analysis. In fact, in this case and for almost all cycles, it is the model forecast that re-duces the error in pressure. When data coverage is limited to the cloudy regions, the situation becomes even worse (the red curve in Fig. 3 (e)). The model pressure field (not shown here) indicates excessive acoustic oscillations in the analysis and forecast solutions, especially at
0.02.0
4.06.08.0
2040
6080100
0.02.0
4.06.08.0
2040
6080100
0.02.0
4.06.08.0
2040
6080
100
0.01.0
2.03.0
4.0
2040
6080
100
020
40
6080
100
1202040
6080100
0.00.20.40.60.81.020
40
6080100
0.00.2
0.4
0.60.81.0
1.220
40
6080100
0.00.4
0.8
1.2
1.6
20
40
6080100
0.0
0.2
0.4
0.6
0.8
20
40
6080100
0.00.2
0.4
0.60.8
1.0
1.220
40
6080100
))
(g/kg)(g/kg)(g/kg)(g/kg)(g/kg)time (min)
time (min)
time (min time (min)
time (min)
time (min)
time (min)
time (min time (min)
time (min)
a)
b)
c)
d)
e)
f)
g)
h)
i)
j)
qc
qr
qv, qi
qs
qh
Figure 3. The rms errors of the mean of ensemble forecasts and analyses, averaged over points at
which the reflectivity is greater than 10dBZ for: a) u (ms -1), b) v (ms -1), c) w (ms -1
) and d) perturbation potential temperature 'θ(K), e) perturbation pressure 'p (Pa), f) q c (g kg -1) g) q r (g kg -1), h) q v (g kg -1;
thick curve), q i (g kg -1), i) q s (g kg -1), j) q h (g kg -1
), for experiment VrCloudy (red), experiment VrFull (black) and experiment VrFLD (only the rms error of 'p is shown, which is the blue curve in (e)).
the lower levels. The problem results from the com-pressible nature of the forecast model. Because the background error correlations associated with the acoustic modes cannot be reliably estimated (mainly because of its high frequency), it is hard for the analy-sis to exactly satisfy the mass continuity equation. When the continuity equation is not satisfied, acoustic modes get excited. The problem is more severe during the earlier assimilation cycles and/or when data coverage is incomplete. The acoustic wave amplitudes actually decreases in the subsequent model forecast, due to dynamic adjustment and built-in damping of acoustic modes in the model. When the data coverage is complete, it takes only a few cycles for the pressure field to adjust and become close to the truth, except in the lower layer. The situation is worse when data coverage is incomplete because in the clear air regions, the model variables are not up-dated by the analysis and large imbalances occur at the edge of cloudy regions, large amplitude acoustic waves are excited there. As a result, the forecast er-ror covariance computed from the ensemble states becomes incorrect and the analysis update does not improve the background forecast of pressure. In the cloud-scale EnKF assimilation studies that use an anelastic model, pressure is diagnosed from the wind field. When wind field does not satisfy the anelastic
mass continuity equation, similar problem may occur
although this issue has not been discussed in the literature.
The open lateral boundary condition used in our simu-lations is another possible source of mass continuity error. To examine the boundary effect, another experiment VrFLD is performed that is the same as VrFull, except the domain 2.5 times as long in the y direction. As the lateral
boundaries are further removed from the storms, the pres-sure analysis and forecast errors are significantly reduced (as shown by the blue curve in Fig 3 b). Since the analysis update to pressure generally hurts the model solution, the update to pressure should not be performed. This is so in our later experiments, and in gen-eral, the model is able to establish a pressure field that is
consistent with the wind and other fields. b) Impact of assimilating reflectivity data Reflectivity is a measurement that is provided by all types of weather radar. In this section, we examine if re-flectivity data alone is sufficient for the model to reproduce the true storm, and we also study its value when used in combination with radial velocity data (which is unavailable
from non-Doppler radars). We note that the observation operator for reflectivity is nonlinear and there exist more uncertainties in the operator with reflectivity data than with radial velocity. In our model with ice microphysics, the rainwater, snow and hail mixing ratios, q r , q s , and q h , are
directly related to the reflectivity (Eqs. 3-6). The experi-ments therefore further test the performance of EnKF scheme in the case of nonlinear observation operator. In
the first experiment ZCloudy, only simulated reflectivity greater than 10dBZ is assimilated.
0.02.04.06.08.0
2040
60801000.02.04.06.08.0
2040
60801000.02.04.06.08.0
2040
6080
1000.01.0
2.03.0
4.0
2040
6080
100020
40
6080
100
1202040
6080100
0.00.20.40.60.81.020
40
6080100
0.00.2
0.4
0.60.81.0
1.220
40
6080100
0.00.4
0.8
1.2
1.6
20
40
6080
100
0.00.2
0.4
0.6
0.8
20
40
6080100
0.00.2
0.4
0.60.8
1.0
1.220
40
6080100
))(g/kg)(g/kg)(g/kg)(g/kg)(g/kg)time (min)
time (min)
time (min time (min)
time (min)
time (min)
time (min)
time (min
time (min)
time (min)
a)b)c)d)e)f)g)h)i)
j)qc
qr
qv, qi
qs
qh
Figure 4. The rms errors of the mean of ensemble forecasts and analyses, averaged over points at which
the reflectivity is greater than 10dBZ for: a) u (ms -1), b) v (ms -1), c) w (ms -1
) and d) perturbation potential
temperature 'θ(K), e) perturbation pressure 'p (Pa), f) q c (g kg -1) g) q r (g kg -1), h) q v (g kg -1
; thick curve), q i (g kg -1), i) q s (g kg -1), j) q h (g kg -1),for experiments ZCloudy (red) and VrCnop (black).
0.02.0
4.0
6.0
8.0
2040
6080100
0.02.0
4.0
6.0
8.0
2040
6080100
0.02.0
4.0
6.0
8.0
2040
6080
100
0.0
1.0
2.0
3.0
4.0
2040
6080100
020
40
6080
100
1202040
6080100
0.00.20.40.60.81.020
40
6080100
0.00.20.4
0.60.81.0
1.220
40
6080100
0.00.4
0.8
1.2
1.620
40
6080
100
0.0
0.2
0.40.6
0.820
40
6080100
0.00.20.4
0.60.81.0
1.220
40
6080100
))(g/kg)(g/kg)(g/kg)(g/kg)(g/kg)time (min)
time (min)
time (min time (min)
time (min)
(m/s)(m/s)(m/s)(K )(Pa)time (min)
time (min)
time (min
time (min)
time (min)
a)b)
c)
d)
e)
f)
g)
h)
i)
j)
u
v
w
pt'
p'
qc
qr
qv, qi
qs
qh
Figure 5. As Fig. 4 for experiments VrCnop (black) and VrZCa (red).
As can be seen in Fig. 4 (red curves), the analysis starts to reduce the rms errors for q r and q h first (start-ing from the second cycle) and then for w , 'θ, q c , q v , q s and q i . Significant reduction in u and v errors did not start until after six to seven cycles. We compare this experiment with VrCnop, which is the same as VrCloudy, except that the pressure field is not updated. The accuracy of the retrieved snow and hail fields is
better while the accuracy of retrieved rain and ice fields is comparable. For water vapor field, the rms error is smaller than that of VrCnop in the later part of the analy-sis period. But the velocity and temperature fields are not as good as those of VrCnop. The result is reason-able, because the relation between reflectivity and the wind and temperature fields is indirect. The information of reflectivity can be used to effectively correct errors in
0.02.0
4.06.08.0
2040
6080100
0.02.0
4.06.08.0
2040
6080100
0.02.0
4.06.08.0
2040
6080
100
0.0
1.0
2.03.0
4.0
2040
6080100
020
40
6080
100
1202040
6080100
0.00.20.40.60.81.020
40
6080100
0.00.20.4
0.60.81.0
1.220
40
6080100
0.00.4
0.8
1.2
1.6
20
40
6080100
0.0
0.2
0.4
0.6
0.8
20
40
6080100
0.00.2
0.4
0.60.8
1.0
1.220
40
6080100
))
(g/kg)(g/kg)(g/kg)(g/kg)(g/kg)time (min)
time (min)
time (min time (min)
time (min)
time (min)
time (min)
time (min time (min)
time (min)
a)b)c)d)e)
f)
g)
h)
i)
j)
qc
qr
qv, qi
qs
qh
Figure 6. As Fig. 4 but for experiments VrZCa (red) and VrZCb (black).
the wind and temperature fields only after the EnKF system produces correct forecast error covariances between the variables.
In the next set of experiments, we combine the radial velocity and the reflectivity data into the assimi-lation process. In experiment VrZCa, both the radial velocity and reflectivity in cloudy regions are assimi-lated (red curve in Fig. 5). Compared to VrCnop, we can see that when additional reflectivity data are as-similated the analysis errors of q s , q h , q r and q i are smaller than those in VrCnop. Whereas for most of the other control variables, in the first part of assimila-tion period (before 65 min), including reflectivity data hurts the model solution. In the later part of the period, the analysis update of all variables, except pressure, is better or is as good as in VrCnop. Our explanation is that at the early stage of the assimilation period, background error covariances between reflectivity and the fields not directly related to reflectivity were not reliable. Updating these variables based on reflectivity data and the unreliable covariances therefore hurts the analysis. As the model state gets closer to truth, the covariances estimated are improved, leading to direct positive impact on other fields.
We note that in Dowell et al. (2003) that uses warm rain microphysics, only q r is updated when as-similating reflectivity observations. We performed a corresponding experiment, VrZCb, in which only q r , q s , and q h were updated when assimilating reflectivity (black curve in Fig. 6). With respect to the wind and temperature fields, this does lead to better results before the sixth cycle. However, after the sixth cycle, the accuracy of the wind and temperature fields is not as good as the case that we do update all variables (except for pressure). As discussed earlier, in the
EnKF system, there is clearly a delay before reliable co-variances between reflectivity and the wind and thermody-namic fields are established. After the initial delay, the reflectivity becomes directly beneficial in retrieving the wind and thermodynamic fields and the assimilation of it leads to overall better analysis. For this reason, we do not believe it appropriate to exclude wind and temperature from the analysis update based on reflectivity data.
To see if we can further improve the analysis, in VrZCc we apply the update due to reflectivity to q r , q s , and q h only before the fourth cycle. After the fourth cycle, all control variables (except for pressure) are updated based on reflectivity. Figure 7 shows that by doing so, the analy-ses for all control variables are improved and so is the forecast of pressure field.
In the previous experiments, only reflectivity larger than 10 dBZ is assimilated. In reality, zero reflectivity out-side the cloudy regions also contains valid information. One can assume that the assimilate reflectivity data cover the entire analysis domain. In our next OSS experiment, named VrCZfull, such assumption is made, but radial ve-locity data are still only available in cloudy regions (Z > 10 dBZ). As was done in experiment VrZCc, the velocities (u, v, w ), perturbation potential temperature 'θ, q v , q c , and q i are updated starting from the fourth cycle when assimilat-ing reflectivity. Figure 8 shows that with the complete re-flectivity coverage, not only the microphysical fields are much improved, but also the wind and temperature fields. The complete coverage of reflectivity data can remove spurious disturbance that can otherwise develop in the data void regions. Experiment VrCZfull produces the best result among those that does not assume complete cover-age of V r data.
0.02.0
4.06.08.0
2040
6080100
0.02.0
4.06.08.0
2040
6080100
0.02.0
4.06.08.0
2040
6080
100
0.0
1.0
2.03.0
4.0
2040
6080100
020
40
6080
100
1202040
6080100
0.00.20.40.60.81.020
40
6080100
0.00.20.4
0.60.81.0
1.220
40
6080100
0.00.4
0.8
1.2
1.6
20
40
6080
100
0.0
0.2
0.4
0.6
0.8
20
40
6080100
0.00.2
0.4
0.60.8
1.0
1.220
40
6080100
))(g/kg)(g/kg)(g/kg)(g/kg)(g/kg)time (min)
time (min)
time (min time (min)
time (min)
time (min)
time (min)
time (min
time (min)
time (min)
a)b)c)d)e)
f)
g)
h)
i)
j)
qc
qr
qv, qi
qs
qh
Figure 7. As Fig. 4 but for experiments VrZCb (black) and VrZCc (red).
0.02.0
4.0
6.0
8.0
2040
6080100
0.02.0
4.0
6.0
8.0
2040
6080100
0.02.0
4.0
6.0
8.0
2040
6080100
0.0
1.0
2.0
3.0
4.0
2040
6080100
020
40
6080
100
1202040
6080100
0.00.2
0.40.60.81.020
40
6080100
0.00.2
0.4
0.60.81.0
1.220
40
6080100
0.00.4
0.8
1.2
1.620
40
6080
100
0.0
0.2
0.40.6
0.820
40
6080100
0.00.2
0.4
0.60.81.0
1.220
40
6080100
))(g/kg)(g/kg)(g/kg)(g/kg)(g/kg)time (min)
time (min)
time (min time (min)
time (min)
(m/s)(m/s)(m/s)(K )(Pa)time (min)
time (min)
time (min
time (min)
time (min)
a)b)c)d)e)
f)g)h)i)j)u
v
w
pt'
p'
qc
qr
qv, qi
qs
qh
Figure 8. As Fig. 4 but for experiments VrZCc (red) and VrCZfull (black).
0.02.0
4.06.08.0
2040
6080100
0.02.0
4.06.08.0
2040
6080100
0.02.0
4.06.08.0
2040
6080100
0.0
1.0
2.03.0
4.0
2040
6080100
020
40
6080
100
1202040
6080100
0.00.2
0.40.60.81.020
40
6080100
0.00.2
0.4
0.60.81.0
1.220
40
6080100
0.00.4
0.8
1.2
1.6
20
40
6080100
0.0
0.2
0.4
0.6
0.8
20
40
6080100
0.00.2
0.4
0.60.8
1.0
1.220
40
6080100
))
(g/kg)(g/kg)(g/kg)(g/kg)(g/kg)time (min)
time (min)
time (min time (min)
time (min)
time (min)
time (min)
time (min time (min)
time (min)
a)b)c)d)e)
f)g)h)
i)j)
qc
qr
qv, qi
qs
qh
Figure 9. As Fig. 4 but for experiments VrCnop (black) and VrCnoice (red).
c) Retrieval of microphysical fields
The microphysics retrieval is an important aspect of convective-scale data assimilation. Relatively few previous studies have focused on this problem. Most of these studies used only simple microphysical parameterization and the ice phase is usually ex-cluded. The recent attempt of Wu et al (2000) uses a 4DVAR data assimilation system to assimilate dual-polarization radar data into a model of deep convec-tive cloud with both liquid and ice phase microphysics. In their study, the microphysics scheme is simplified and consists of only three categories: rain, hail and cloud liquid-ice. The reflectivity and differential reflec-tivity data were converted to rain and hail mixing ratios
first, rather than being directly assimilated. The differ-ential reflectivity data were necessary for such a con-version. In our study, the original detailed ice micro-physics parameterization is used and only regular reflectivity measurement is assumed available. Our problem is more difficult here because more water and ice species have to be determined and no dual polari-zation information is available. Figure 10 shows the distribution of the five categories of hydrometeor for the true run and for assimilation experiments VrFull, VrCnop, VrCnoice, ZCloudy and VrCZfull. The figure shows that the EnKF data assimilation system is able to establish detailed microphysical structures that have very good fidelity. The quality of actual analysis does depend on the usage and availability of data, as seen earlier by the error plots. To better understand the way the EnKF scheme works when retrieving the microphysical fields, we performed another experiment, named VrCnoice, in which q i , q s , and q h (as well as pressure) are not updated by the analysis and only radial velocity in cloudy regions are assimilated. For these three variables, the difference between VrCnoice
and VrCnop is relatively small before the 6th
cycle, but be-comes significant after that time (Fig. 9). For the early pe-riod, the relatively small difference reflects relatively weak
or unreliable link (through background error covariance)
between the observation (V r ) and these three variables. The link apparently becomes stronger and more effective in correcting errors in these fields at the later stage. On the other hand, despite of the lack of direct correction to q i , q s , and q h by V r , the errors in the former are still reduced in time in general. Such reductions are achieved through
model dynamics – when other model fields are improved,
fields that are not directly updated have to adjust and be-come consistent with these other fields. The more accurate hydrometeor fields, in turn, help improve the overall model state. The link through model dynamics is more important at the early period of assimilation when background error
covariances are less reliable. Despite the effectiveness of radial velocity assimilation, the reflectivity is most important for retrieving the micro-physical fields because it has the most direct link to the hydrometeors. Experiment ZCloudy helps us understand this point. With only reflectivity, we cannot obtain as good an analysis for wind and thermodynamic fields, but the microphysical fields can be retrieved accurately (Fig. 4 and Fig. 10).
(6-a)
(6-b)
(6-c)
(6-d)
(6-e)
(km)(km)(km)(km)(km)(k m )
(k m )
(k m )
(k m )
(k m )
(k m )
qc qr qi qs qh
T r u e
V r F u l l
V r C n o p
Z C l o u d y
V r C n o i c e
V r C Z f u l l
Figure 10. The cross-sections of q c , q r , q i , q s and q h (g kg -1
) fields that pass through the maximum up-draft for the truth simulation (1-a)~(1-e), and experiments VrFull (2-a)~(2-e), VrCnop (3-a)~(3-e),
VrCnoice (4-a)~(4-e), ZCloudy (5-a)~(5-e) and VrCZfull (6-a)~(6-e) at T=65 min.
4. Concluding and Discussion
In this study we applied the ensemble Kalman fil-ter technique to the assimilation of simulated radar radial velocity and reflectivity data using a com-pressible model with a complex microphysics scheme. The method is shown to have great potentials for the assimilation of such data. Through flow-dependent forecast error covariance estimation from the ensem-ble states, not only the wind and thermodynamic
fields can be retrieved accurately, all five categories of hydrometeor can also be retrieved successfully. Reliable covariances between the observations and variables not directly related to them can be obtained after a few assimi-lation cycles even when they are started from initial guess made of an environmental sounding plus random pertur-bations. After the initial number of cycles, useful observa-tional information can be spread to the indirectly related variables through reliable forecast error covariances. Up-dating indirectly related variables after the first few cycles when assimilating reflectivity data produces the best
analysis. Using reflectivity information in clear air re-gions is also beneficial. When using a compressible model, acoustic wave mode can be excited by errors in the wind and pressure analyses and it is recom-mended that pressure be excluded from the analysis update. This choice is shown to work the best.
Of course, caution should be applied when inter-preting OSSE results. Both forecast model and for-ward observation operators are assumed perfect, which happens to work well for OSSE data. Despite a recent success in applying EnKF to real radar data (which still used a model in idealized settings), much work is still needed in moving us in the direction of real case and real data.
Acknowledgement
This work was supported by NSF grant ATM0129892. The second author was also supported by NSF ATM9909007 and an DOT-FAA grant. The authors benefited from discussions with Drs. Fuqing Zhang and David Dowell.
Reference
Burgers, G., P. J. van Leeuwen, and G. Evensen, 1998: Analysis scheme in the ensemble Kalman filter. Mon. Wea. Rev., 126, 1719-1724. Dowell, D., and F. Zhang, 2003: Wind and Thermody-namic Retrievals in the 17 May 1981 Arcadia, Oklahoma Supercell: Ensemble Kalman Filter Experiments. Submitted to Mon. Wea. Rev. Evensen, G., 1994: Sequential data assimilation with
a nonlinear quasi-geostrophic model using
Monte Carlo methods to forecast error statistics.
J. Geophys. Res., 99 (C5), 10 143-10 162. Evensen, G., 2003: The ensemble Kalman filter: Theoretical formulation and practical implementa-
tion. Submitted to Ocear Dynamics.
Gal-Chen, T., 1978: A method for the initialization of the anelastic equations: Implications for matching models with observations. Mon. Wea. Rev., 106,
587-606.
Gao, J., M. Xue, Z. Wang, and K. K. Droegemeier, 1998: The initial condition and explicit prediction of convection using ARPS adjoint and other re-
trievals methods with WSR-88D data. 12th Conf.
Num. Wea. Pred., Phoenix AZ, Amer. Meteor.
Soc., 176-178.
Gao, J.-D., M. Xue, A. Shapiro, and K. K. Droege-meier, 1999: A variational method for the analy-
sis of three-dimensional wind fields from two Doppler radars. Mon. Wea. Rev., 127, 2128-2142. Houtekamer, P. L., and H. L. Mitchell, 1998: Data assimilation using an ensemble Kalman filter technique. Mon. Wea. Rev., 126, 796-811. Lord, S. J., E. Kalnay, R. Daley, G. D. Emmitt, and R. At-las, 1997: Using OSSEs in the design of the future generation of integrated observing systems. Preprint volume, 1st Symposium on Integrated Observation Systems, Long Beach, CA, Amer. Meteor. Soc.
Qiu, C.-J. and Q. Xu, 1992: A simple adjoint method of wind analysis for single-Doppler data. J. Atmos. Oce-
anic Technol., 9, 588-598.
Ray, P. S., B. Johnson, K. W. Johnson, J. S. Bradberry, J.
J. Stephens, K. K. Wagner, R. B. Wilhelmson, and J.
B. Klemp, 1981: The morphology of severe tornadic
storms on 20 May 1977. J. Atmos. Sci., 38, 1643-
1663.
Shapiro, A., S. Ellis, and J. Shaw, 1995: Single-Doppler radar retrievals with Phoenix II data: Clear air and mi-
croburst wind retrievals in the planetary boundary layer. J. Atmos. Sci., 52, 1265-1287.
Snyder, C., and F. Zhang, 2003: Assimilation of Simulated Doppler Radar Observations with an Ensemble Kal-
man Filter. Mon. Wea. Rev., 131, 1663-1677.
Sun, J., D. W. Flicker, and D. K. Lilly, 1991: Recovery of three-dimensional wind and temperature fields from simulated single-Doppler radar data. J. Atmos. Sci., 48, 876-890.
Sun, J. and N. A. Crook, 1997: Dynamical and micro-physical retrieval from Doppler radar observations us-
ing a cloud model and its adjoint. Part I: Model devel-
opment and simulated data experiments. J. Atmos.
Sci., 54, 1642-1661.
Weygandt, S. S., A. Shapiro, and K. K. Droegemeier, 2002: Retrieval of Model Initial Fields from Single-Doppler Observations of a Supercell Thunderstorm. Part I: Single-Doppler Velocity Retrieval. Mon. Wea. Rev., 130, 433-453.
Wu, B., J Verlinde., and J. Sun, 2000: Dynamical and Microphysical retrievals from Doppler Radar observa-
tions of a deep convective cloud. J. Atmos. Sci., 57, 262-283.
Xue, M., K. K. Droegemeier, and V. Wong, 2000: The Ad-vanced Regional Prediction System (ARPS) - A mul-
tiscale nonhydrostatic atmospheric simulation and prediction tool. Part I: Model dynamics and verifica-
tion. Meteor. Atmos. Physics, 75, 161-193.
Xue, M., K. K. Droegemeier, V. Wong, A. Shapiro, K.
Brewster, F. Carr, D. Weber, Y. Liu, and D. Wang, 2001: The Advanced Regional Prediction System (ARPS) - A multi-scale nonhydrostatic atmospheric simulation and prediction tool. Part II: Model physics and applications. Meteor. Atmos. Phys., 76, 143-166. Zhang, F., C. Snyder, and J. Sun, 2003: Impacts of initial estimate and observations on the convective-scale data assimilation with an ensemble Kalman filter.
Submitted to Mon. Wea. Rev.
调试指南 第一部分：设备安装位置 一、建议录播主机、跟踪主机、音频处理器都安装在机柜，有利于设备散热，讲台到黑板墙距离为1.2米，第一排学生课桌到黑板墙距离为2.5米。 二、老师云台摄像机和老师全景摄像机（如上图位置），居中安装在教室后面墙上（壁装），若教室过长（超过12米），可在距离黑板约10米的位置吊装,摄像机离地约2.3米。 三、学生云台摄像机安装在黑板上沿齐平，居中安装，学生全景摄像机安装在黑板或显示大屏的两侧均可，摄像机离地面高度为2米。具体情况可依据现场情况而定。 四、三个跟踪辅助摄像机的位置分别为：老师辅助：安装在里黑板墙面约4.5米的天花顶上，横向居中安装
板书辅助：安装在黑板的左上角或右上角天花顶上，离黑板边框约1米，离黑板墙面约30厘米，如下图所示。 学生辅助：安装在学生云台摄像机正上方的天花顶上。备注：第一排学生要求离黑板垂直距离为2.5米。 五、吊麦安装：位置可参照第一张图片中位置 第一排（学生吊麦）：离墙约1.8米，左右对称安装，分别离教室左右中轴线1.5-2.5米位置，视教室宽度而定，斜向下指向讲台。 第二排（老师吊麦）：离墙约2.5米，左右对称，斜向下指向学生区域 第三排（学生吊麦）：分别与第二排两支吊麦在同一纵线上，距离第二排约3米
第二部分：录播参数配置 一、打开IE浏览器，输入录播IP（默认：169.254.178.178，帐号密码均为：admin），登录至导播界面。(注意：将电脑IP设置成录播主机同一网段) 二、打开“软件下载”菜单，下载安装FBVLC播放器、PC辅助软件（解压出三个软件：鼠标检测MouseDetector_v1.0，片头片尾、知识点索引PPTCapture_v1.0，安装设置在教师上课电脑，第三个 A V AJoystick v1.0安装在管理电脑与导播控制键盘对接）。 三、检查“视频1”至“视频5”是否有图像，并调整出正确的对应 画面：视频1：老师云台，视频2：学生云台，视频3：老师枪机，视频4：学生枪机，视频5：老师上课电脑VGA或HDMI 信号，视频6：片头片尾。通过鼠标点击跟踪检查云台摄像机的控制线是否有效。 四、打开“系统设置”菜单，设置如下参数： 1、网络参数设置：设置录播主机IP地址 2、云台参数设置：跟踪连接参数→选择“连接到跟踪主机” 3、直播参数设置：环出模式→选择“1080P60” 网络直播参数→传输模式→选择“TCP模式”补充说明：(1) TCP是面向连接的传输控制协议，而UDP提供了无连接的数据报服务; (2) TCP具有高可靠性，确保传输数据的正确性，不出现丢失或乱序;UDP在传输数据前不建立连接，不对数据报进行检查与修改，无须等待对方的应答，所以会出现分组丢失、重复、
软件开发过程详解 软件开发过程即软件设计思路和方法的一般过程，包括设计软件的功能和实现的算法和方法、软件的总体结构设计和模块设计、编程和调试、程序联调和测试以及编写、提交程序。 生产一个最终能满足需求且达到工程目标的软件产品所需要的步骤。软件开发过程覆盖了需求、设计、实现、确认以及维护等活动。需求活动包括问题分析和需求分析。问题分析获取需求定义，又称软件需求规约。需求分析生成功能规约。设计活动一般包括概要设计和详细设计。概要设计建立整个软件系统结构，包括子系统、模块以及相关层次的说明、每一模块的接口定义。详细设计产生程序员可用的模块说明，包括每一模块中数据结构说明及加工描述。实现活动把设计结果转换为可执行的程序代码。确认活动贯穿于整个开发过程，实现完成后的确认，保证最终产品满足用户的要求。维护活动包括使用过程中的扩充、修改与完善。 1.需求分析 1.1 需求分析的特点和任务 需求分析是软件开发的第一步。获取需求的一个必不可少的结果是对项目中描述的客户需求的普遍理解。一旦理解了需求，分析者、开发者和客户就能探索出描述这些需求的多种解决方案。参与需求获取者只有在他们理解了问题之后才能开始设计系统，否则，对需求定义的任何改进，设计上都必须大量的返工。把需求获取集中在用户任务上—而不是集中在用户接口上—有助于防止开发组由于草率处理设计问题而造成的失误。有几种原因使需求分析变得困难：（1）客户说不清楚需求；（2）需求自身经常变动；（3）分析人员或客户理解有误。 需求获取、分析、编写需求规格说明和验证并不遵循线性的顺序，这些活动是相互隔开、增量和反复的。当你和客户合作时，你就将会问一些问题，并且取得他们所提供的信息（需求获取）。同时，你将处理这些信息以理解它们，并把它们分成不同的类别，还要把客户需求同可能的软件需求相联系（分析）。然后，你可以使客户信息结构化，并编写成文档和示意图（说明）。下一步，就可以让客户代表评审文档并纠正存在的错误（验证）。这四个过程贯穿着需求分析的整个阶段。需求获取可能是软件开发中最困难、最关键、最易出错及最需要交流的方面。需求获取只有通过有效的客户—开发者的合作才能成功。分析者必须建立一个对问题进行彻底探讨的环境，而这些问题与产品有关。为了方便清晰地进行交流，就要列出重要的小组，而不是假想所有的参与者都持有相同的看法。对需求问题的全面考察需要一种技术，利用这种技术不但考虑了问题的功能需求方面，还可讨论项目的非功能需求。确定用户已经理解：对于某些功能的讨论并不意味着即将在产品中实现它。对于想到的需求必须集中处理并设定优先级，以避免一个不能带来任何益处的无限大的项目。 1.2.需求分析的一般方法
计算机系统软件和应用软件安装调试 设置BIOS 开启计算机或重新启动计算机后，在屏幕显示“Waiting……”时，按下“Del”键进入CMOS的设置界面。如果按得太晚，计算机将会启动系统，这时只有重新启动计算机了。我们可在开机后立刻按住Delete键直到进入CMOS（不同的品牌计算机进入Bios设置的键不同，比如我的lenovo就是按F1）。进入后，我们可以用方向键移动光标选择CMOS设置界面上的选项，然后按Enter 进入副选单，用ESC键来返回副菜单，用PAGE UP和PAGE DOWN键来选择具体选项，F10键保留并退出BIOS设置。 (1) 硬盘初始化 硬盘初始化主要包括低级格式化、分区、高级格式化这三个步骤，然后才能利用硬盘存储数据。 ①低级格式化一般在硬盘出厂时已经完成
②分区的主要步骤：建立主分区》建立扩展分区》建立逻辑分区》 激活主分区》格式化所有分区。 主分区主要包含操作系统启动所必须的文件和数据；扩展分区，我们可以根据需要及操作系统的磁盘管理能力而设置；至于逻辑分区，扩展分区不能直接使用，要将其分成一个或多个逻辑驱动的区域，也叫逻辑驱动器，才可以为操作系统识别和使用。 ③在对硬盘执行了分区的操作后，还需要对硬盘进行格式化操作， 才可以正常使用硬盘，这里的格式化操作是指高级格式化，高级格式化使用Format命令完成，在DOS系统或Windows操作系统中都有该命令。 (2) 安装中文Microsoft Windows7操作系统 可用硬盘或光盘安装，这里介绍光盘安装。 ①放入光盘，重启笔记本电脑，在进入BIOS的瞬间，快速按下 F12键（lenovo按F12，其他计算机一般按F1或Delete）。选 择光驱启动项，然后回车，即可进入光盘启动。 ②开始引导，如果光盘没问题，接下来就开始装系统了。 ③“要安装的语言”选择“中文(简体)”，“时间和货币格式”选择“中 文(简体，中国)”，“键盘和输入方法”选择“中文(简体)-美式键 盘”，）点击“下一步”。 ④选择版本，按照当前安装光盘提示为准，直接点击“下一步”即 可。
软件系统安装与实施合同书精 装版 Effectively restrain the parties’ actions and ensure that the legitimate rights and interests of the state, collectives and individuals are not harmed ( 合同范本 ) 甲方：______________________ 乙方：______________________ 日期：_______年_____月_____日 编号：MZ-HT-071834
软件系统安装与实施合同书精装版 甲方：_________公司（系统使用企业） 住所：_________ 法定代表人：_________ 乙方：_________公司（系统开发公司） 住所：_________ 法定代表人：_________ 甲、乙双方在遵照国家有关法律 、法规 的规定，本着友好、协作的精神，经共同协商，就甲、乙双方共同在甲方生产现场实施“_________监控系统”事宜达成如下条款：第一条双方责任的约定 为改善甲方生产管理手段，甲、乙双方合作实施_________监控
系统_________，并分别承担下列相应条款规定的责任。 1．甲方责任 （1）甲方负责提供系统安装调试所需的有关资料和指定配合乙方工作的人员，配合乙方在规定的期限内，完成上述软件系统所需的生产工艺信息数据库的建立和参与_________系统的连接工作。 （2）甲方须提供项目实施所需的尽量详尽的_________格式文件，以便于进行_________系统_________的数据交换工作。 （3）未经乙方事先同意，甲方不得将该软件或其中任何一部分提供给第三方使用。 （4）甲方负责上述软件系统在本单位使用过程中的版权保密工作。 2．乙方责任 （1）乙方向甲方提供“_________监控系统_________”软件一套，其中包括服务器版调度排产子系统_________个；工艺准备子系统_________个；远程控制数据采集端_________个。远程数据采集器在本系统定为计算机，并负责在乙方指定工作现场的安装与实施。
titlesec&titletoc中文文档 张海军编译 makeday1984@https://www.doczj.com/doc/7614318057.html, 2009年10月 目录 1简介,1 2titlesec基本功能,2 2.1.格式,2.—2.2.间隔, 3.—2.3.工具,3. 3titlesec用法进阶,3 3.1.标题格式,3.—3.2.标题间距, 4.—3.3.与间隔相关的工具, 5.—3.4.标题 填充,5.—3.5.页面类型,6.—3.6.断行,6. 4titletoc部分,6 4.1.titletoc快速上手,6. 1简介 The titlesec and titletoc宏包是用来改变L A T E X中默认标题和目录样式的，可以提供当前L A T E X中没有的功能。Piet van Oostrum写的fancyhdr宏包、Rowland McDonnell的sectsty宏包以及Peter Wilson的tocloft宏包用法更容易些；如果希望用法简单的朋友，可以考虑使用它们。 要想正确使用titlesec宏包，首先要明白L A T E X中标题的构成，一个完整的标题是由标签+间隔+标题内容构成的。比如： 1.这是一个标题，此标题中 1.就是这个标题的标签，这是一个标签是此标题的内容，它们之间的间距就是间隔了。 1
2titlesec基本功能 改变标题样式最容易的方法就是用几向个命令和一系列选项。如果你感觉用这种方法已经能满足你的需求，就不要读除本节之外的其它章节了1。 2.1格式 格式里用三组选项来控制字体的簇、大小以及对齐方法。没有必要设置每一个选项，因为有些选项已经有默认值了。 rm s f t t md b f up i t s l s c 用来控制字体的族和形状2，默认是bf,详情见表1。 项目意义备注（相当于） rm roman字体\textrm{...} sf sans serif字体\textsf{...} tt typewriter字体\texttt{...} md mdseries(中等粗体)\textmd{...} bf bfseries(粗体)\textbf{...} up直立字体\textup{...} it italic字体\textit{...} sl slanted字体\textsl{...} sc小号大写字母\textsc{...} 表1:字体族、形状选项 bf和md属于控制字体形状，其余均是切换字体族的。 b i g medium s m a l l t i n y（大、中、小、很小） 用来标题字体的大小，默认是big。 1这句话是宏包作者说的，不过我感觉大多情况下，是不能满足需要的，特别是中文排版，英文 可能会好些！ 2L A T E X中的字体有5种属性：编码、族、形状、系列和尺寸。 2
系统安装调试及项目验收方案
目录 1.1系统安装调试及验收 (3) 1.1.1设备运输保障 (3) 1.1.2安装调试环境 (6) 1.1.3系统安装 (10) 1.1.4系统测试 (12) 1.1.5验收 (39)
1.1系统安装调试及验收 1.1.1设备运输保障 1.1.1.1人员和保障措施 我公司将对本次送货异常重视，以运输单位牵头，合力精心布置、科学分工，运输单位组成：生产作业组，技术监督组，安全保障组。具体保障措施如下： 1、建立运输总负责人制，严格保证设备在运输过程中不损坏、不丢失零部件。 2、遇大风、暴雨、大雪时，不准进行运输作业。 3、参与本次运输工作的所有车辆器具，使用前应详细检查、维修好，由质检员、安全员共同 鉴定合格后方可使用。 4、在运输装卸过程中，积极与用户沟通、征求意见互相协助，使运送工作安全、合格、用户 满意。 1.1.1.2包装及标记 1.1.1. 2.1包装 提供的所有设备和材料都具备适应内陆运输和多次搬运、装卸的坚固包装，包装内含有减振、防冲击的措施，保证在运输、装卸过程中完好无损。 若包装无法防止运输、装卸过程中垂直、水平加速度引起的设备损坏，我方会在设备的设计结构上予以解决。 包装将按设备特点，按需要分别加上防潮、防霉、防锈、防腐蚀的保护措施，以保证货物在没有任何损坏或腐蚀的情况下安全运抵合同设备安装地点。 包装所用的材料及包装物结构具有较强的可复原性，以保证货物在现场开箱后能方便地按原包装复原。 将尽量考虑安装现场的环境，采用防潮防冻包装。在包装货物时，按贷物类别进行装箱。 备品备件将在包装箱外加以注明，分批或一次性发货。 专用工具也会分别包装。各种设备的松散零部件将采用好的包装方式，装入尺寸适当的箱内。 1.1.1. 2.2标记 按规定对货物进行包装。对包装箱内和捆内的各散装部件都将标记在系统装配图中的部
嵌入式软件开发流程 一、嵌入式软件开发流程 1.1 嵌入式系统开发概述 由嵌入式系统本身的特性所影响，嵌入式系统开发与通用系统的开发有很大的区别。嵌入式系统的开发主要分为系统总体开发、嵌入式硬件开发和嵌入式软件开发3大部分，其总体流程图如图1.1所示。 图1.1 嵌入式系统开发流程图 在系统总体开发中，由于嵌入式系统与硬件依赖非常紧密，往往某些需求只能通过特定的硬件才能实现，因此需要进行处理器选型，以更好地满足产品的需求。另外，对于有些硬件和软件都可以实现的功能，就需要在成本和性能上做出抉择。往往通过硬件实现会增加产品的成本，但能大大提高产品的性能和可靠性。 再次，开发环境的选择对于嵌入式系统的开发也有很大的影响。这里的开发环境包括嵌入式操作系统的选择以及开发工具的选择等。比如，对开发成本和进度限制较大的产品可以选择嵌入式Linux，对实时性要求非常高的产品可以选择Vxworks等。
1.2 嵌入式软件开发概述 嵌入式软件开发总体流程为图4.15中“软件设计实现”部分所示，它同通用计算机软件开发一样，分为需求分析、软件概要设计、软件详细设计、软件实现和软件测试。其中嵌入式软件需求分析与硬件的需求分析合二为一，故没有分开画出。 由于在嵌入式软件开发的工具非常多，为了更好地帮助读者选择开发工具，下面首先对嵌入式软件开发过程中所使用的工具做一简单归纳。 嵌入式软件的开发工具根据不同的开发过程而划分，比如在需求分析阶段，可以选择IBM的Rational Rose等软件，而在程序开发阶段可以采用CodeWarrior（下面要介绍的ADS 的一个工具）等，在调试阶段所用的Multi-ICE等。同时，不同的嵌入式操作系统往往会有配套的开发工具，比如Vxworks有集成开发环境Tornado，WindowsCE的集成开发环境WindowsCE Platform等。此外，不同的处理器可能还有对应的开发工具，比如ARM的常用集成开发工具ADS、IAR和RealView等。在这里，大多数软件都有比较高的使用费用，但也可以大大加快产品的开发进度，用户可以根据需求自行选择。图4.16是嵌入式开发的不同阶段的常用软件。 图1.2 嵌入式开发不同阶段的常用软件 嵌入式系统的软件开发与通常软件开发的区别主要在于软件实现部分，其中又可以分为编译和调试两部分，下面分别对这两部分进行讲解。 1．交叉编译 嵌入式软件开发所采用的编译为交叉编译。所谓交叉编译就是在一个平台上生成可以在另一个平台上执行的代码。在第3章中已经提到，编译的最主要的工作就在将程序转化成运行该程序的CPU所能识别的机器代码，由于不同的体系结构有不同的指令系统。因此，不同的CPU需要有相应的编译器，而交叉编译就如同翻译一样，把相同的程序代码翻译成不同CPU的对应可执行二进制文件。要注意的是，编译器本身也是程序，也要在与之对应的某一个CPU平台上运行。嵌入式系统交叉编译环境如图4.17所示。
本公司工地验收测试前及设备出厂前进行整机全面测试，包括软件，硬件及附属设备，并整理提出表明测试结果的设备出厂测试报告。 对设备进行通电运行试验，并对全通信网络进行统调。经过对全网络仔细严格的网络测试后，其性能及稳定性可达到移交时，须将测试记录和最新版软件以及有关文件，资料，备用备件、工具、仪表等移交。 系统在现场安装后如果当地电话局有关部门认为需要对程控交换机进行测试时，本公司须派出人员在现场配合测试。在试运行验收须作出以下测试： ◆故障率观察，各项性能和功能测试，诊断维护控制功能测试，传输指标测试， 其他必要的 测试，检查各电话分机之安装及其标称功能，检查后备电源工作状况，检查各配线设备之安装及回路指示，检查备份控制机层之工作状态 ◆安装 所有铜电缆应在线槽或管道内布线。包单位应在地下一层总配线室、电缆联接箱、光纤密 封盒、电缆进口、垂直电缆、网络联接块以及其它地方按照工业惯例标明导体的电极(末端及环)。承包单位应提供正确地完成安装工作所需的任何专用安装设备或工具。这将包括终接电缆设施、铜/光纤电缆测试和接线设施、通讯设施、电缆转盘的支撑架、或其它的安装电缆所需的工具，在没有合适垫具的情况下﹐承包单位不得卷动或贮存电缆转盘。承包单位不应在电力线路旁安装任何电缆﹐或与其它电气器具共享同一根线管、线槽或套管。 ◆电磁干扰的分离 带有不同种类信号或不同电压的设备如集装在一个共享的容器内﹐应按供
货商的有关 要求﹐有效地与任何其它一类的设备屏蔽以避免电磁干扰。 ◆电缆的安排 根据图纸或规格说明书的要求提供一切必需的电缆插锁、插座、接线耳等等﹐并按照电缆 的种类与入口方各将它们固定于安装板和安装带上。 整齐地安装系统内所有线路的电缆导体﹐并按正规间距将所有半导体固定以防止 电缆在运行情况下可能的损坏(例如热膨胀、震动等)或引起其它线路短路。 ◆电缆的终接 不要让已绝缘的导体触及未绝缘的有电部件或锐利的边缘。 每个终端只能连接一条电缆导体。在一个终端连接两条或更多导体将不会被接受。 ◆接地 A、承包单位应负责在所有新安装的电缆框架上提供一个被批准的接地点﹐并保证能正确 与任何现有设施连接。承包单位还须再现正确地将所有有关的电缆、包合体、柜、服务箱和框架连接起来以保证接地的延续性。所有接地应由铜线或铜带组成﹐应由一个被批准的大厦接地点供应﹐并应与主要的电气接地点连接。 B、应根据BS7671给所有门、盖板等等提供线路保护导体。接地铜导体的最小断面积为2.5
ctex宏包说明 https://www.doczj.com/doc/7614318057.html,? 版本号：v1.02c修改日期：2011/03/11 摘要 ctex宏包提供了一个统一的中文L A T E X文档框架，底层支持CCT、CJK和xeCJK 三种中文L A T E X系统。ctex宏包提供了编写中文L A T E X文档常用的一些宏定义和命令。 ctex宏包需要CCT系统或者CJK宏包或者xeCJK宏包的支持。主要文件包括ctexart.cls、ctexrep.cls、ctexbook.cls和ctex.sty、ctexcap.sty。 ctex宏包由https://www.doczj.com/doc/7614318057.html,制作并负责维护。 目录 1简介2 2使用帮助3 2.1使用CJK或xeCJK (3) 2.2使用CCT (3) 2.3选项 (4) 2.3.1只能用于文档类的选项 (4) 2.3.2只能用于文档类和ctexcap.sty的选项 (4) 2.3.3中文编码选项 (4) 2.3.4中文字库选项 (5) 2.3.5CCT引擎选项 (5) 2.3.6排版风格选项 (5) 2.3.7宏包兼容选项 (6) 2.3.8缺省选项 (6) 2.4基本命令 (6) 2.4.1字体设置 (6) 2.4.2字号、字距、字宽和缩进 (7) ?https://www.doczj.com/doc/7614318057.html, 1
1简介2 2.4.3中文数字转换 (7) 2.5高级设置 (8) 2.5.1章节标题设置 (9) 2.5.2部分修改标题格式 (12) 2.5.3附录标题设置 (12) 2.5.4其他标题设置 (13) 2.5.5其他设置 (13) 2.6配置文件 (14) 3版本更新15 4开发人员17 1简介 这个宏包的部分原始代码来自于由王磊编写cjkbook.cls文档类，还有一小部分原始代码来自于吴凌云编写的GB.cap文件。原来的这些工作都是零零碎碎编写的，没有认真、系统的设计，也没有用户文档，非常不利于维护和改进。2003年，吴凌云用doc和docstrip工具重新编写了整个文档，并增加了许多新的功能。2007年，oseen和王越在ctex宏包基础上增加了对UTF-8编码的支持，开发出了ctexutf8宏包。2009年5月，我们在Google Code建立了ctex-kit项目1，对ctex宏包及相关宏包和脚本进行了整合，并加入了对XeT E X的支持。该项目由https://www.doczj.com/doc/7614318057.html,社区的开发者共同维护，新版本号为v0.9。在开发新版本时，考虑到合作开发和调试的方便，我们不再使用doc和docstrip工具，改为直接编写宏包文件。 最初Knuth设计开发T E X的时候没有考虑到支持多国语言，特别是多字节的中日韩语言。这使得T E X以至后来的L A T E X对中文的支持一直不是很好。即使在CJK解决了中文字符处理的问题以后，中文用户使用L A T E X仍然要面对许多困难。最常见的就是中文化的标题。由于中文习惯和西方语言的不同，使得很难直接使用原有的标题结构来表示中文标题。因此需要对标准L A T E X宏包做较大的修改。此外，还有诸如中文字号的对应关系等等。ctex宏包正是尝试着解决这些问题。中间很多地方用到了在https://www.doczj.com/doc/7614318057.html,论坛上的讨论结果，在此对参与讨论的朋友们表示感谢。 ctex宏包由五个主要文件构成：ctexart.cls、ctexrep.cls、ctexbook.cls和ctex.sty、ctexcap.sty。ctex.sty主要是提供整合的中文环境，可以配合大多数文档类使用。而ctexcap.sty则是在ctex.sty的基础上对L A T E X的三个标准文档类的格式进行修改以符合中文习惯，该宏包只能配合这三个标准文档类使用。ctexart.cls、ctexrep.cls、ctexbook.cls则是ctex.sty、ctexcap.sty分别和三个标准文档类结合产生的新文档类，除了包含ctex.sty、ctexcap.sty的所有功能，还加入了一些修改文档类缺省设置的内容（如使用五号字体为缺省字体）。 1https://www.doczj.com/doc/7614318057.html,/p/ctex-kit/
IP网络广播系统安装 调试部分 用户使用手册
目录 第一部分：软件安装 .......................................... 错误!未定义书签。 系统服务器软件............................................ 错误!未定义书签。 远程客户端软件安装........................................ 错误!未定义书签。第二部分：系统配置 .......................................... 错误!未定义书签。 预设置终端参数............................................ 错误!未定义书签。 配置系统服务器参数........................................ 错误!未定义书签。 配置远程客户端参数........................................ 错误!未定义书签。 在系统服务器中配置中继服务................................ 错误!未定义书签。 在系统服务器中添加终端.................................... 错误!未定义书签。 在系统服务器中添加用户.................................... 错误!未定义书签。 在系统服务器中播放节目.................................... 错误!未定义书签。第二部分：MP3节目制作工具................................... 错误!未定义书签。 一. 如何打开制作工具 ................................... 错误!未定义书签。 二. 如何录音 ........................................... 错误!未定义书签。 三. 如何CD抓轨 ........................................ 错误!未定义书签。 四. 如何转换音频文件格式 ............................... 错误!未定义书签。 五. 如何编辑音频文件 ................................... 错误!未定义书签。 1.分割音频文件 .................................... 错误!未定义书签。 2.合并音频文件 .................................... 错误!未定义书签。第三部分：硬件安装 .......................................... 错误!未定义书签。 壁挂式终端 ............................................... 错误!未定义书签。 机柜式终端 ............................................... 错误!未定义书签。第四部分：注意事项 .......................................... 错误!未定义书签。 网络规划 ................................................. 错误!未定义书签。 网络故障 ................................................. 错误!未定义书签。 关于跨VLAN或三层路由.................................. 错误!未定义书签。 关于组播协议 .......................................... 错误!未定义书签。 关于防火墙 ............................................ 错误!未定义书签。 音效异常 ................................................. 错误!未定义书签。 音频通路连接不良 ...................................... 错误!未定义书签。 声卡错误 .............................................. 错误!未定义书签。
word文档打开是乱码有很多原因： 1.比如版本不同，往往高版本能打开低版本的，但低版本打开高版本就会出错。 2.源文件是否被破坏。 3.字体等不相同，出现不匹配，也会出现乱码。 1.可能是低版本打开高版本。或者是格式不相同。 2.文档损坏了。 解决办法： 方法一：利用word2003的“打开并修复”功能来修复文挡。 (1)启动word2003，单击“文件-----打开”，在“打开”对话框中选重要修复的word文挡。 (2)单击“打开”按钮右边的下三角按钮，在弹出的下拉菜单中选中“打开并修复”选项，即可对损坏的文挡进行修复，修复完成后，显示文挡内容。 提示：“打开并修复”是只有word2003以上的版本才具有的功能。 方法二：转换文挡格式来修复文挡。 如果使用的是word2003以下的版本(如word2000)，可以用此方法来修复文挡。 (1)启动word2000后单击“工具------选项-------常规”。在该选项卡中选中“打开时确认转换”复选框，并单击“确定”按钮。 (2)单击“文件---打开”，在弹出的“打开”对话框中选中要恢复的文件，并在“文件类型”框中选中“从任意文件中恢复文本”。 (3)单击“打开“按钮自动对孙换文挡进行转换修复。 (4)如果显示的文挡内容混乱，单击“文件---另存为”，将文挡保存为“RTF格式”或其他word 所识别的格式。 (5)保存后关闭文挡，再次打开以“RTF格式”保存的文挡即可看到完整的文挡。 (6)再将文挡保存为“DOC格式”即可，这样就完成了对孙换文挡的修复操作。 提示：此方法也同样只可以在word2002以上的版本中使用。 方法三：重设格式法 Word用文档中的最后一个段落标记关联各种格式设置信息，特别是节与样式的格式设置。这
tabularx宏包中改变弹性列的宽度\hsize 分类：latex 2012-03-07 21:54 12人阅读评论(0) 收藏编辑删除 \documentclass{article} \usepackage{amsmath} \usepackage{amssymb} \usepackage{latexsym} \usepackage{CJK} \usepackage{tabularx} \usepackage{array} \newcommand{\PreserveBackslash}[1]{\let \temp =\\#1 \let \\ = \temp} \newcolumntype{C}[1]{>{\PreserveBackslash\centering}p{#1}} \newcolumntype{R}[1]{>{\PreserveBackslash\raggedleft}p{#1}} \newcolumntype{L}[1]{>{\PreserveBackslash\raggedright}p{#1}} \begin{document} \begin{CJK*}{GBK}{song} \CJKtilde \begin{tabularx}{10.5cm}{|p{3cm} |>{\setlength{\hsize}{.5\hsize}\centering}X |>{\setlength{\hsize}{1.5\hsize}}X|} %\hsize是自动计算的列宽度，上面{.5\hsize}与{1.5\hsize}中的\hsize前的数字加起来必须等于表格的弹性列数量。对于本例，弹性列有2列，所以“.5+1.5=2”正确。 %共3列，总列宽为10.5cm。第1列列宽为3cm，第3列的列宽是第2列列宽的3倍，其宽度自动计算。第2列文字左右居中对齐。注意：\multicolum命令不能跨越X列。 \hline 聪明的鱼儿在咬钩前常常排祠再三& 这是因为它们要荆断食物是否安全&知果它们认为有危险\\ \hline 它们枕不会吃& 如果它们判定没有危险& 它们就食吞钩\\ \hline 一眼识破诱饵的危险，却又不由自主地去吞钩的& 那才正是人的心理而不是鱼的心理& 是人的愚合而不是鱼的恳奋\\
安装、调试及验收方案 1对系统调试和验收的响应 对系统调试的响应 调试工作是整个系统完成的最后技术阶段，也是技术性强、环节复杂、易出现各种问题的阶段。 我司缜密的制定调试计划，编写试运行及调试方案，填报详细日志，包括以下内容： 1. 对单项设备进行调试，确保单项产品质量过关，拟写测试报告； 2. 对分系统进行调试，确保各分系统安全可靠运行，拟写测试报告； 3. 整个系统联调，确保工程顺利完工，在测试中出现问题及时查找问题之所在，迅速及时地解决，拟写测试报告。 , 对系统试运行的响应 试验运行包括下面以下内容： 1. 系统试运行时间为3个月。 2. 在试运行开始日期之前，我司向采购人提供能证明系统联调成功、可正常运行的所有测量数据和资料。 3. 所有试运转期间设备的修改和软件变化都应在试运转结束后写入操作和维修手册中。 4. 我司给出任何缺陷或故障部件修复的全部细节。
对验收的响应 1. 系统验收 ^ 系统自测完成后，我方首先拟出一个测试方案，具体到每一个测试步骤，与业主和监理讨论通过后，方可按计划进行测试。由投标人提供测试方法、测试工具、测试数据。系统每一项测试必须有详细的测试记录，须有业主、监理和投标人三方代表签字确认，并附有详细的分析报告。 2. 项目竣工验收 （1）系统开通后须正常试运行3个月。 （2）项目的验收必须经过工程验收（由我司组织）、用户验收（由用户单位组织）、公安技防管理部门验收（由技防管理部门组织）、项目终验等阶段，所有验收费用由我司承担。 （3）我司在投标文件中，根据《广东省安全技术防范管理条例》、《广东省安全技术防范管理条例实施办法》、《广州市社会治安视频监控系统验收指南》的要求，提交详细描述验收的组织和实施办法，测试方案，试运行时间，用户验收条件等。 2 系统安装方案 软件系统安装的主要目标不仅是使所有软件能够在相应平台上正常运行，而且必须具有对软件系统运行的监控测试手段，以证明系统优化运行。投标人有责任且必须承诺使项目单位的系统达到以上目标。
信息系统安装调试手册 一、手册对象： 本手册提供给需要对信息系统的源代码进行编写和修改的人员，以及对源代码有兴趣的人员。 本手册假定使用人员具备一定的软件知识，了解JAVA语言，了解ORACLE数据库，了解面向对象编程的思想。 二、需求硬件： PC电脑一台，预装WINDOWS系列操作系统： 安装的软件： 1.1必须安装的软件： 1.1.1JAVA虚拟机：JDK 1.6 及以上 1.1.2JAVA的IDE集成环境：ECLIPSE，版本不限 1.1.3数据库客户端：ORACLE数据库客户端,版本不限；或者数据库服务器，版本为ORACLE 11G 1.1.4数据库连接工具: PL/SQL 9，版本不限 1.1.5 WEB服务器软件:TOMCAT 6.0 1.2可以安装的软件： 1.2.1 数据库设计查看软件：PowerDesigner 1.2.2JAVA的窗体设计软件：netbeans 1.2.3 源代码管理工具：SVN 四、需要的资源： 数据库服务器连接地址：见详细地址 信息系统WEB端源代码：见详细地址 信息系统子系统端源代码：见详细地址
五、环境安装步骤： 第一步：安装JDK 1.6，详细如下： JDK1.6的下载-安装安装JDK很简单，只需要按照安装向导一步一步进行即可，安装好后重点如何设置环境变量 安装Java JDK1.6最重要的也是最容易出错的是设置环境变量。一共需要设置3个环境变量：Path，Classpath和Java_Home（大小写无所谓）。Windows系统中设置环境变量如下图：右击“我的电脑”，选择“属性” 点击“高级”选项卡，选择“环境变量”。
软件开发的过程 摘要：软件开发过程即软件设计思路和方法的一般过程，包括设计软件的功能和实现的算法和方法、软件的总体结构设计和模块设计、编程和调试、程序联调和测试以及编写、提交程序。 生产一个最终能满足需求且达到工程目标的软件产品所需要的步骤。软件开发过程覆盖了需求、设计、实现、确认以及维护等活动。需求活动包括问题分析和需求分析。问题分析获取需求定义，又称软件需求规约。需求分析生成功能规约。设计活动一般包括概要设计和详细设计。概要设计建立整个软件系统结构，包括子系统、模块以及相关层次的说明、每一模块的接口定义。详细设计产生程序员可用的模块说明，包括每一模块中数据结构说明及加工描述。实现活动把设计结果转换为可执行的程序代码。确认活动贯穿于整个开发过程，实现完成后的确认，保证最终产品满足用户的要求。维护活动包括使用过程中的扩充、修改与完善。 1.需求分析 1.1 需求分析的特点和任务 需求分析是软件开发的第一步。获取需求的一个必不可少的结果是对项目中描述的客户需求的普遍理解。一旦理解了需求，分析者、开发者和客户就能探索出描述这些需求的多种解决方案。参与需求获取者只有在他们理解了问题之后才能开始设计系统，否则，对需求定义的任何改进，设计上都必须大量的返工。把需求获取集中在用户任务上—而不是集中在用户接口上—有助于防止开发组由于草率处理设计问题而造成的失误。有几种原因使需求分析变得困难：（1）客户说不清楚需求；（2）需求自身经常变动；（3）分析人员或客户理解有误。 需求获取、分析、编写需求规格说明和验证并不遵循线性的顺序，这些活动是相互隔开、增量和反复的。当你和客户合作时，你就将会问一些问题，并且取得他们所提供的信息（需求获取）。同时，你将处理这些信息以理解它们，并把它们分成不同的类别，还要把客户需求同可能的软件需求相联系（分析）。然后，你可以使客户信息结构化，并编写成文档和示意图（说明）。下一步，就可以让客户代表评审文档并纠正存在的错误（验证）。这四个过程贯穿着需求分析的整个阶段。需求获取可能是软件开发中最困难、最关键、最易出错及最需要交流的方面。需求获取只有通过有效的客户—开发者的合作才能成功。分析者必须建立一个对问题进行彻底探讨的环境，而这些问题与产品有关。为了方便清晰地进行交流，就要列出重要的小组，而不是假想所有的参与者都持有相同的看法。对需求问题的全面考察需要一种技术，利用这种技术不但考虑了问题的功能需求方面，还可讨论项目的非功能需求。确定用户已经理解：对于某些功能的讨论并不意味着即将在产品中实现它。对于想到的需求必须集中处理并设定优先级，以避免一个不能带来任何益处的无限大的项目。 1.2.需求分析的一般方法 跟班作业。通过亲身参加业务工作来了解业务活动的情况。这种方法可以比
系统名称 安装调试方案 一、系统基本情况 二、安装调试内容 三、安装调试成功/失败的标准 定义每个施工项目通过或失败的准则。 四、实施挂起及恢复 指明挂起实施的准则，即在什么情况下将挂起全部或部分实施项目，并指明恢复实施的必要条件。 五、组织形式 说明实施参与人员及各自的角色、工作权限范围。 六、实施任务安排 具体的工作对应到具体的人员、角色。 七、实施进度安排 实施阶段的总体时间：起始日期-终止日期，根据具体工作安排如下： 1．实施动员：日期。 2．硬件设备安装调试 1)服务器、工作站安装调试：起始日期-终止日期，完成系统后台硬件平台及软件平台的搭建。 2)网络布线及调试：起始日期-终止日期，完成IC卡管理系统专用网络的布线、施工、调试。 3)读写设备安装调试：起始日期-终止日期，在指定位置完成IC卡读写设备的安装、调试。 4)设备联网调试: 起始日期-终止日期，完成设备联网后的设备、网络测试。 3．软件系统安装调试 1)软件模块调试阶段：起始日期-终止日期，完成各软件模块的调试工作。 2)软件集成调试阶段：起始日期-终止日期，完成所有软件模块的联合调试工作。 3)基础数据准备阶段：起始日期-终止日期，完成所有基础数据的整理和录入工作。 4)基础数据校验阶段：起始日期-终止日期，对录入的基础数据做全面校对。 5)系统初始化阶段：起始日期-终止日期，完成系统试运行必须的初始化工作，如起初库存和 起初余额等设置。 4．系统集成调试阶段：起始日期-终止日期，完成系统软、硬件结合调试。 5．系统培训阶段：起始日期-终止日期，根据具体时间安排完成系统操作、管理、维护培训。 6．系统运行测试阶段：起始日期-终止日期，根据项目合同、客户需求说明书及相关文档，引导业务人员根据实际业务完成系统运行测试工作，并给出系统确认测试报告。 7．系统试运行阶段：起始日期-终止日期，引导业务人员根据实际业务完成系统试运行。 8．系统正式运行阶段：起始日期-终止日期，引导业务人员根据实际业务完成系统正式运行。 9．项目验收阶段：起始日期-终止日期。 八、支持条件 8.1 系统支持(软件、硬件、网络环境) 8.2 需由客户承担的工作 8.3 需由子合同承担的工作
%=== 配合前面的ntheorem宏包产生各种定理结构,重定义一些正文相关标题===% \theoremstyle{plain} \theoremheaderfont{\normalfont\rmfamily\CJKfamily{hei}} \theorembodyfont{\normalfont\rm\CJKfamily{song}} \theoremindent0em \theoremseparator{\hspace{1em}} \theoremnumbering{arabic} %\theoremsymbol{} %定理结束时自动添加的标志 \newtheorem{definition}{\hspace{2em}定义}[chapter] %\newtheorem{definition}{\hei 定义}[section] %!!!注意当section为中国数字时，[sction]不可用！ \newtheorem{proposition}{\hspace{2em}命题}[chapter] \newtheorem{property}{\hspace{2em}性质}[chapter] \newtheorem{lemma}{\hspace{2em}引理}[chapter] %\newtheorem{lemma}[definition]{引理} \newtheorem{theorem}{\hspace{2em}定理}[chapter] \newtheorem{axiom}{\hspace{2em}公理}[chapter] \newtheorem{corollary}{\hspace{2em}推论}[chapter] \newtheorem{exercise}{\hspace{2em}习题}[chapter] \theoremsymbol{$\blacksquare$} \newtheorem{example}{\hspace{2em}例}[chapter] \theoremstyle{nonumberplain} \theoremheaderfont{\CJKfamily{hei}\rmfamily} \theorembodyfont{\normalfont \rm \CJKfamily{song}} \theoremindent0em \theoremseparator{\hspace{1em}} \theoremsymbol{$\blacksquare$} \newtheorem{proof}{\hspace{2em}证明} \usepackage{amsmath}%数学 \usepackage[amsmath,thmmarks,hyperref]{ntheorem} \theoremstyle{break} \newtheorem{example}{Example}[section]