Bridging Centrality:Identifying Bridging Nodes In
Scale-free Networks?
Woochang Hwang?Y oung-rae Cho?Aidong Zhang?Murali Ramanathan??
?Department of Computer Science and Engineering,State University of New York at Bu?alo,USA ??Department of Pharmaceutical Sciences,State University of New York at Bu?alo,USA
Email:{whwang2,ycho8,azhang}@cse.bu?https://www.doczj.com/doc/f4180460.html,,murali@acsu.bu?https://www.doczj.com/doc/f4180460.html,
ABSTRACT
Several centrality measures were introduced to identify es-sential components and compute components’importance in networks.Majority of these centrality measures are dom-inated by components’degree due to their nature of looking at networks’topology.We propose a novel essential com-ponent identi?cation model,bridging centrality,based on information?ow and topological locality in scale-free net-works.Bridging centrality provides an entirely new way of scrutinizing network structures and measuring compo-nents’importance.We apply bridging centrality on real world networks,including one simulated network,two bio-logical networks,two social networks,and one web network, and show that the nodes distinguished by bridging centrality are well located on the connecting positions between highly connected regions through analyzing the clustering coe?-cient and average path length of those networks.Bridging centrality can discriminate bridging nodes,the nodes with more information?owed through them and locations be-tween highly connected regions,while other centrality mea-sures can not.
Categories and Subject Descriptors
[Network Analysis]:Network metrics,Network compo-nent importance metrics,Essential component analysis General Terms
Degree,shortest path,betweenness,clustering coe?cient, average path length,singleton
Keywords
?This research is partly supported by NSF grants DBI-0234895,IIS-0308001and NIH grant1P20GM067650-01A1.All opinions,?ndings,conclusions and recommen-dations in this paper are those of the authors and do not necessarily re?ect the views of the funding agencies.
KDD’06August20-23,2006,Philadelphia,PA,USA Scale-free network,centrality,bridging node,bridging coef-?cient,bridging centrality,modularity,robusteness,paths protection
1.INTRODUCTION
Many real world systems,e.g.,internet,World Wide Web (WWW),social systems,biological systems,etc.,can be de-scribed as complex networks,which are structured as a set of nodes and a set of edges connecting the nodes.Scale-free network[4]is the most popular and emerging form of net-work in these real world network systems.Most of these real world networks have been proved to follow some topo-logical statistical features,i.e.,features of scale-free network, such as power law degree distribution,small world property, and high modularity[2,3,4,5].Power law degree distri-bution depicts the probability of?nding a highly connected node decreases exponentially with its own degree,which is the number of edges incident on the node.In other words, there are many low degree nodes,and only a small number of nodes have high degree.The second phenomenon,small world property,describes that the average distance between nodes in a network is relatively shorter than other network types, e.g.,random networks of the same https://www.doczj.com/doc/f4180460.html,ly, any node can be reached within small number of consec-utive edges from a node in a network.A module refers to a densely connected,functionally or physically,group of nodes in a network.For the last distinct and the most interesting property,these real world networks have high modularity which indicates that high clustering is one of dominating characteristics of these networks.
Over the past few years,empirical and theoretical studies of networks have been one of the most popular subjects of recent researches in many areas including technological,so-cial,and biological?https://www.doczj.com/doc/f4180460.html,work theories have been ap-plied with good success to these real world systems,and many centrality indices,measurements of the importance of the components in a network,have been introduced[6,9, 10,16,7,18].While these centrality indices have proved that they made outstanding achievements in the analysis and understanding of the roles of nodes in a network,ma-jority of these existing centrality indices focuses only on the extent how much nodes are well located on central positions or play central roles from the stand point of topology and information?ow.These existing centrality measures can not help being considerably dominated by nodes’degree due to their nature of computing components’importance.Even
though these approaches are very good at identifying cen-tral components,i.e.,central components from any central-ity viewpoint,of a network or of a module,they concentrate only on central components and overlook another essential topological aspect in networks.
In this research,we move the focus of the network analy-sis from the directions of identifying central nodes to an-other entirely new,fresh,and important direction.From our deeper observation of the high modularity property of scale-free networks,we claim that there should be“bridg-ing”nodes that are located between modules,and we found that there exist“bridging”nodes in real world scale-free net-works due to their high modularity phenomenon.So,we also claim that these bridging nodes,which bridge densely con-nected regions,should be attractive and important essential components in a network.We introduce a novel centrality metric,bridging centrality,that successfully identi?es the bridging nodes locating between densely connected regions, i.e.modules,using high modularity or high clustering prop-erty which is one of the most important property of scale-free networks.Experiments on several real world network sys-tems are performed to demonstrate the e?ectiveness of our metric.
Bridging centrality has many potential applications in sev-eral areas.First,it can be used to break up modules in a network for clustering purpose.Functional modules or physical modules in biological networks or sub community structures in social and technological networks can be de-tected using the bridging nodes chosen by bridging central-ity.Second,it also can be used to identify the most critical points interrupting the information?ow in a network for network protection and robustness improvement purposes for networks.Third,in biological applications,the bridging centrality can be used to locate the key proteins,which are the connecting nodes among functional modules.
2.METHOD
2.1Terminology and Representation
Real world systems can be represented using graph theoretic methods.The approach presented in this paper focuses on undirected graphs.An undirected graph G=(V,E)consists of a set V of nodes or vertices and a set E of edges,E?V×V. An edge e(i,j)connects two nodes i and j,e(i,j)∈E.
The neighbors N(i)of node i are de?ned to be a set of di-rectly connected nodes to node i.The degree d(i)of a node i is the number of the edges connected to node i.A path is de?ned as a sequence of nodes(n1,...,n k)such that from each of its nodes there is an edge to the successor node. The length of a path is the number of edges in its node se-quence.A shortest path between two nodes,i and j,is a minimal length path between them.The distance between two nodes,i and j,is the length of its shortest path.The clustering coe?cient C v for a node v is the proportion of links between the nodes within its neighbourhood divided by the number of links that could possibly exist between them, C v=2|{e(i,j)}|
d(v)(d(v)?1)
:i,j∈N(v),e(i,j)∈E[19].In other words, |{e(i,j)}|gives the number of triangles that go through node v,whereas d(v)(d(v)?1)/2is the total number of triangles that could pass through node v.Thus,clustering coe?cient of node v indicates how the neighbors of node v are well connected each other.The clustering coe?cient of a graph is the average of the clustering coe?cients of all nodes in the graph.The average path length of a graph is the average of the shortest paths between all pairs of nodes in the graph.
2.2Bridging Centrality
A bridging node is a node lying between modules,i.e.,a node connecting densely connected components in a graph. The bridging nodes in a graph are identi?ed on the basis of their high value of bridging centrality relative to other nodes on the same graph.The bridging centrality of a node is the product of the betweenness centrality C B[10]and the bridging coe?cient(BC),which measures the global and local features of a node,respectively.
Speci?cally,the bridging centrality C R(v)for node v of in-terest,is de?ned by:
C R(v)=BC(v)×C B(v)(1) The betweenness centrality is a measure of the global impor-tance of a node that assesses the proportion of the shortest paths between all node pairs that pass through the node of interest.The betweenness centrality,C B(v),for a node v of interest is de?ned by:
C B(v)=
X
s=v=t
s,v,t∈V
ρst(v)
ρst
(2)
In the above equation,ρst is the number of shortest paths from node s to t andρst(v)the number of shortest paths from s to t that pass through the node v.The higher C B(v), more number of shortest paths between all node pairs pass through the node v.So node v is more likely to be located on the shortest paths between all node pairs in the network, i.e.,more information travel through the node v.
The bridging coe?cient of a node determines the extent how well the node is located between high degree nodes.The bridging coe?cient of a node v is de?ned:
BC(v)=
d(v)?1
P
i∈N(v)
1
(3)
where d(v)is the degree of node v,and N(v)is the set of neighbors of node v.The bridging coe?cient assesses the lo-cal bridging characteristics in the neighborhood.The bridg-ing coe?cient understands a network as a simple electrical circuit.Intuitively,there should be more congestion on the smaller degree nodes if an unit electrical current arrives on a node since the smaller degree nodes have lesser number of outlets than the bigger degree nodes have.So,if we consider the reciprocal of the degree of a node as the“resistance”of the node,the bridging coe?cient can be viewed as the ratio of the resistance of a node to the sum of the resistance of the neighbors.Critical bridging nodes,typically representing rate limiting points in the network and because they con-nect its densely connected regions,have high“resistance.”Thus,higher C R(v)signi?es that more information?ows through node v,i.e.,higher betweenness centrality(C B(v)), and more resistance on node v,i.e.,higher bridging coe?-cient(BC(v)),by bridging densely connected regions.
Figure1:A small synthetic network example.Top six high bridging score nodes are colored.
Node Degree C B BC C R
E20.533330.857140.45713
B20.155550.857140.13333
D20.155550.857140.13333
F30.477770.222220.10617
A40.655550.100000.06555
J30.211110.166660.03518
Table1:Top six centrality values of Figure1,including Betweenness(C B),bridging coe?cient(BC),and bridging centrality(C R).
Figure1and Table1clearly illustrates the essence of bridg-ing centrality.Although node A has the highest degree and betweenness value,nodes E,B,and D have much higher bridging centrality values since node A is located on the center of a module not on a bridge which results in the lowest bridging coe?cient value.In other words,far more number of shortest paths goes through node A than other three nodes,but nodes E,B,and D position on bridges much better.So,nodes E,B,and D have higher bridg-ing centrality values since they are on the bridges between modules which leads much higher bridging coe?cient values than node A.Betweenness centrality decides only the extent how much important the node of interest is from information ?ow standpoint,and it does not consider the topological lo-cations of the node.On the other hand,nodes B and D have the same bridging coe?cient value with node E,but nodes B and D have much less betweenness centrality values since far more number of shortest paths passes through node E than through nodes B and D.Even though nodes E,B,and D are located on similar local topological positions,i.e.,similar lo-cal topological surroundings,node E is taking a much more important location than nodes B and D in the information ?ow viewpoint.Bridging coe?cient measures only the ex-tent how well the node is located between highly connected regions,and it does not deliberate the node’s importance from information?ow standpoint.Without a doubt,we can ?gure out that node E is taking a better bridging position than nodes B and D are in Figure1.Bridging nodes should be positioned between modules and also located on impor-tant positions in information?ow standpoint.So,bridging centrality combines these two measurements,betweenness centrality and bridging coe?cient,since none of these two indices can di?erentiate the bridging nodes alone,as we saw in the above.So bridging centrality combines global and lo-cal features,betweenness centrality and bridging coe?cient respectively,of the node not focusing only on one topologi-cal factor like other centrality indices do,and
discriminates Figure2:A synthetic network with36nodes and46edges. The nodes with the highest0-10th percentile of values for the bridging centrality are highlighted in black circles,the nodes with the10th-25th percentiles of bridging centrality are highlighted in gray circles.The letters are node labels. the bridging nodes which are located on the critical posi-tions for information?ow viewpoint and also are positioned on the bridges.
3.RESULTS
The focus of this research and performance analysis is mainly on the top25%high bridging centrality score components in all examples,since the signi?cance and the interest are rapidly reduced below top25percentile.Furthermore,bridg-ing centrality values and the range of the bridging nodes can be arbitrary according to the network topology dealt with.Empirical studies on several real world network sys-tems made us de?ne“bridging nodes”as the top25per-centile.
3.1Application on Simulated Data
To obtain a preliminary assessment of the underlying net-work characteristics identi?ed by the bridging centrality,we applied the metric to a synthetic network consisting of36 nodes and46edges shown in Figure2.The synthetic net-work investigated contains key elements such as hub nodes, peripheral nodes,cycles and bridging nodes that are com-monly found in biological networks.The overall degree dis-tribution followed a power law distribution but the overall size was kept small so that any patterns present could be easily detected by visual inspection.
In Figure2,we have highlighted the nodes in the highest 0-10th percentiles of bridging centrality values with black-?lled circles whereas nodes in the highest10th-25th per-centiles of bridging centrality values are shown in gray-?lled circles.Visual inspection of the synthetic network reveals
Figure3:The schematic of undirected graph network model for the p53protein with82nodes and106edges.The nodes with the highest0-10th percentile of values for the bridging centrality are highlighted in black circles;the nodes with the 10th-25th percentiles of bridging centrality are highlighted in gray circles.The labels are abbreviations of gene names. that the bridging centrality values of peripheral nodes(e.g., D,Y,Z,AA,AB,AI,AJ),hub nodes(e.g.,A,AF,X)and nodes in simple cycles(e.g.,A,B,C;M,N,O)do not occur in the highest percentiles of bridging centrality.The highest values of bridging centrality occur in the nodes that connect the modules and highly connected regions of the network.
3.2Application on the p53Network
Based on the encouraging performance of the bridging cen-trality metric on the synthetic network,we evaluated its performance on a simple undirected graph network model for the p53related regulatory network[13].The p53protein is a critical tumor suppressor molecule and because it is of-ten mutated in many human tumors,its interactions could potentially provide targets for anti-cancer drugs.
Figure3is the schematic of the undirected graph model for the p53regulatory network.Similar to Figure2,the nodes with the highest top10percentile bridging centrality values are shown with the black circles,whereas top25percentile with the grey circles.We can easily validate that the dis-tinguished bridging nodes(e.g.Mdm2,GRB2,SOS,CDK6, EIF4E,TGFB1,H-RAS,N-RAS)are clearly standing on the boundary between modules.
3.3Application on the Yeast Metabolic Net-
work
In the next step,we extended the promising results obtained with the simple p53regulatory network model to the undi-rected yeast metabolic network[17].The yeast
metabolic Figure4:The yeast metabolic network with359nodes and 435edges.The nodes with the highest0-10th percentile of values for the bridging centrality are highlighted in black circles;the nodes with the10th-25th percentiles of bridging centrality are highlighted in gray circles.
network is relatively well modularized and clustered accord-ing to their cellular functions.Figure4shows that bridging centrality successfully identi?es the bridging nodes and the nodes lying on the borders of modules.Importantly,the majority of its key bridging nodes can be readily identi?ed by visual inspection.
3.4Application on Social Networks Encouraged by the outstanding performance of bridging cen-trality metric on the biological networks,we shift gear to so-cial networks,an academic collaboration network[15]and a character relationship network in a novel,Les miserable[15]. Figures5and6exhibits the bridging centrality results on the academic collaboration network in physics research and the character network of Les Miserable,respectively.As can be seen in Figures5and6,the bridging nodes are well positioned on the road between modules,even though the networks are much more complex and highly cross connected in the core area than the previous examples.Furthermore, it is obviously shown that network information should pass through the highlighted bridging nodes if it tries to move from a module to a module.
3.5Application on a Web Network
One of the most emerging real world networks is a web net-work,i.e.,networks that connect web pages in World Wide Web.Figure7visualizes a small section web network[1][15]. This network is simple but highly modular with many pe-
Figure5:The academic collaboration network in physics research with142nodes and340edges.The nodes with the highest0-10th percentile of values for the bridging centrality are highlighted in black circles;the nodes with the10th-25th percentiles of bridging centrality are highlighted in gray
circles.
Figure6:The character network of Les Miserable with77 nodes and254edges.The nodes with the highest0-10th percentile of values for the bridging centrality are highlighted in black circles;the nodes with the10th-25th percentiles of bridging centrality are highlighted in gray
circles.Figure7:The web network with180nodes and228edges. The nodes with the highest0-10th percentile of values for the bridging centrality are highlighted in black circles;the nodes with the10th-25th percentiles of bridging centrality are highlighted in gray circles.
ripheral nodes,and importantly,the majority of its key bridging nodes can be readily identi?ed by visual inspec-tion.Most of the top10percentile scored nodes resides on the bridging positions and other relatively low scored nodes, which are between10%and25%and colored in grey,also well positioned on the boundary of modules.We can clearly di?erentiate modules by inspecting the bridging nodes as frontier of modules.
3.6Theoretical Analysis on Yeast Metabolic
Network
The main objective of this study is to analyze the potential of bridging centrality score to select the nodes that position on true bridging locations.We use the yeast metabolic net-work for further analyses since it has better scale-free net-work properties,e.g.,power law distribution,high modular-ity,than other examples in the above and also has moderate size that enables us to observe the performances precisely. In order to investigate the topological locality of the bridg-ing nodes picked up by bridging centrality in networks,we analyzed and compared the behaviors of the clustering co-e?cient,the average path length,and the number of single-tons occurrence with other two famous centrality measures. Figure8(a)compares the behaviors of the clustering coef-?cient of the network in the consequence of consecutive re-movals of top10percentile high centrality score nodes for three centrality measures,degree centrality(or node degree), betweenness centrality(BW),and bridging centrality(BR). The clustering coe?cient behaviors for these three centrali-
C l u s t e r i n g C o e f f i c i e n t
Node Cut
(a)
A v e r a g e P a t h L e n g t h
Node Cut
(b)
N u m b e r o f S i n g l e t o n s
Node Cut
(c)
Figure 8:Analysis on the Yeast Metabolic Network.(a)Average Path Length Changes,(b)Clustering Coe?cient changes,(c)Singleton Changes.Changes of the clustering coe?cient,average path length,number of singletons followed by the consecutive top 10percentile high score node removals for three centrality measures(degree,betweenness(BW),bridging(BR)).ties explain some interesting and important features of the nodes picked by these three di?erent centrality measures.For more clear understanding of the clustering coe?cient behaviors,one needs to observe the behaviors together with the changes of number of singletons simultaneously.Figure 8(c)shows the changes of the number of singletons produced by the same node removals.The removals by the other two centrality indices,degree centrality and betweenness cen-trality,did not show monotonic behaviors of the cluster-ing coe?cients,and they rather considerably decreased the clustering coe?cient about 20%.Furthermore,they pro-duced many more singletons than bridging centrality did in the same intervals.Needless to say,the nodes caught by the other two centrality indices are located on the center of modules and the removals of those nodes damaged the mod-ularity of the network and mass-produced singletons.How-ever,as we removed the highest bridging centrality score nodes one by one,the clustering coe?cient of the network was increased about 10%constantly for almost all intervals while only one singleton was produced in the same interval.In other words,cutting the high bridging centrality nodes enhanced the modularity of the network without producing many singletons,i.e.,the nodes picked up by bridging cen-trality are located between modules neither on the center of modules nor on the peripheral of the network.
As the second evidence of the bridging centrality’s superior-ity on targeting the bridging nodes,we observed the topolog-ical properties of the bridging nodes discriminated by bridg-ing centrality from the alternative paths availability and av-erage path length point of view.Figure 8(b)describes the changes of the average path length followed by the removals of top 10percentile high centrality score nodes.Increment of the average path length by a node removal means that there are some node isolations from the other part of the network or there are some alternative paths but longer than the re-moved path.The changes of the average path lengths should be also scrutinized along with the changes of the number of singletons,Figure 8(c),to comprehend more precisely.The changes of the average path length for the other two cen-trality indices,degree centrality and betweenness centrality,were increased more than the case of bridging centrality.But it is clear that their increment behaviors are caused by the mass-production of singletons in the same interval as can be seen in Figure 8(c)since the nodes distinguished by these two centrality indices mostly are located on the center of modules that have many peripheral nodes with one de-gree.Therefore,interrupting the nodes caught by these two centrality measures caused many single node isolations and turned out to be the larger increment of the average path length.On the other hand,the average path length of the interruptions on the bridging nodes discriminated by bridg-ing centrality were also increased signi?cantly with generat-ing only one singleton in the same intervals.This behavior insists that interruptions on the bridging nodes resulted in much longer alternative paths or isolations of larger modules not singletons.
4.DISCUSSION AND CONCLUSION
Jeong’s group has espoused the degree of a node as a key basis for essential components identi?cation[14].These high degree nodes are called hubs,and hubs have been found to be important determinants of survival in network perturbation.Power-law networks are very robust to random attacks but very vulnerable to targeted attack in this model[2].Hahn’s group looked for di?erences in degree,betweenness,and closeness centrality between essential and nonessential genes in three eukaryotic protein interaction networks:yeast,worm,and ?y[12].These three interaction networks are found to have remarkable similar structure and the proteins that have a more central position in networks,regardless of the number of direct interactors,evolve more slowly and are more likely to be essential for survival.Estrada’s group introduces a new centrality measure,which is called subgraph centrality that characterizes the participation of each node in all subgraphs in a network[8][9].The subgraph centrality is better able to discriminate the lethal nodes of a network than any other measures in protein interaction networks.Palumbo’s group tried to ?nd lethal nodes by arc deletion,which could lead to sub components isolation.They showed that lethality cor-
responds to the lack of alternative paths in the perturbed network linking the nodes a?ected by the enzyme deletion on yeast metabolic network which is a directed network[17]. Existing approaches are focusing only on?nding central and lethal nodes,and it has been proven that these existing ap-proaches can discriminate lethal nodes very well.We argue that identifying network’s essential components with these existing methods is likely to prove suboptimal because of their limited view of looking at the problem.Guimera’s group devised a clustering method to identify functional modules in metabolic pathways and categorized the role of each component in the pathway according to their topologi-cal location relative to detected functional modules[11].An-notating locality of components in network’s topology based on a certain clustering method is totally biased by the used clustering method.So identifying components’topological location,e.g.,hubs,peripheral nodes,or bridging nodes,in-dependent from any other methods is more preferable. While other existing approaches are focusing on targeting high degree,high central,and high lethal components in network topology,our bridging centrality discriminates the bridging nodes with more information?owed through them, i.e.,more central from the information?ow aspect,and also positioned between highly connected regions.We have shown that bridging centrality successfully distinguishes the bridging nodes in several real world scale-free networks in-cluding social,biological,and technical networks.Theoret-ical analysis of the yeast metabolic network,observing the clustering coe?cient changes and the average path length behaviors,were performed and showed that the nodes picked up by bridging centrality are well positioned on the connect-ing spots between modules.
Throughout the experiments we performed in this paper, bridging centrality did a great job on identifying the bridg-ing nodes in real world networks.Bridging centrality have many possible applications on many research areas.The recognition of the bridging nodes and information about the bridging nodes should be very valuable knowledge for further fruitful achievements in biological researches and in other?elds too.For example,identifying functional or physical modules or identifying the key components in bi-ological networks using the bridging centrality will provide a very e?ective and totally new way of looking biological network structures.This promising outcome should also be applicable to social networks for detecting sub community structures or discovering the key elements in them.As we observed in the previous section,while the perturbations on hubs or the nodes selected by other centrality indices caused a few local singleton isolations and might have many alter-native paths due to their high clustering property,which is one of the main properties of the scale-frre networks,among neighbors inside the module,the failures on the bridging nodes,unsurprisingly,caused whole module isolations from the rest of the network and might have longer alternative paths or no alternative path at all.So the interruptions on the bridging nodes could be much more lethal,and the cost of network failure by interrupting the bridging nodes would be much higher than the failure on the other nodes. Therefore,we claim that the bridging nodes picked up by bridging centrality also reside on the critical positions and also are worth getting attentions for the network robustness improvement and paths protection standpoint.
5.FUTURE WORK
It was clearly shown that the bridging nodes discriminated by bridging centrality are well positioned between highly connected modules in scale-free https://www.doczj.com/doc/f4180460.html,ing this bridg-ing centrality superiority,clustering analysis on scale-free networks can be accomplished through di?erentiating mod-ules by considering the bridging nodes as the boundary of clusters.
6.REFERENCES
[1]The graph is one of the test graphs from the graph
drawing competition held in conjunction with the
Symposium on Graph Drawing,Berkeley,California,
September18-20,1996.
[2]R.Albert,H.Joeng,and A.L.Barabasi.Error and
attack tolerance of complex networks.Nature,
406:378–381,Jul2000.
[3]E.Alm and A.P.Arkin.Biological networks.Current
opinion in Struc.Bio.,13:193–202,2003.
[4]A.L.Barabasi and E.Bonabeau.Scale-free networks.
Scienti?c American,pages60–69,May2003.
[5]A.L.Barabasi and https://www.doczj.com/doc/f4180460.html,work biology:
understanding the cell’s functional organization.
Nature Reviews:Genetics,5:101–113,Feb2004.
[6]P.Bonacich.Power and centrality:a family of
measures.Am.J.Sociol.,92:1170–1182,1987.
[7]S.brin and L.Page.The anatomy of a large-scale
hypertextual web search https://www.doczj.com/doc/f4180460.html,puter Networks
and ISDN Systems,30:107–117,1998.
[8]E.Estrada.Virtual identi?acation of essential proteins
within the protein interaction network of yeast.
Proteomics,6(1):35–40,Jan2006.
[9]E.Estrada and J.A.Rodiguez-Velazquez.Subgraph
centrality in complex networks.Physical Review,
71(5):056103,2005.
[10]L.C.Freeman.A set of measures of centrality based
on betweenness.Sociometry,40:35–41,1979.
[11]R.Guimera and L.A.N.Amaral.Functional
cartography of complex metabolic networks.Nature,
433:895–900,Feb2005.
[12]M.W.Hahn and https://www.doczj.com/doc/f4180460.html,parative genomics
of centrality and essentiality in three eukaryotic
protein-interaction networks.Mol.Biology and
Evolution,22(4):803–806,2005.
[13]J.Hallinan.cluster analsys of the p53genetic
regulatory network:Topology and biology.2004IEEE Symposium on Computational Intelligence in
Bioinformatics and Computational Biology,Oct2004.
[14]H.Jeong,S.P.Manson,A.L.barabasi,and Z.N.
Oltvai.Lethality and centrality in protein networks.
Nature,411:41–42,May2001.
[15]M.E.J.Newman.Finding and evaluating community
structure in networks.arXiv:cond-mat,1:0308217,
Aug2003.
[16]M.E.J.Newman.A measure of betweenness
centrlality on random walks.arXiv:cond-mat,
1:0309045,Sep2003.
[17]M.C.Palumbo,A.Colosimo,A.Giuliani,and
L.Farina.Functional essentiality from topology
features in metabolic networks:A case study in yeast.
FEBS Letters,579:4642–4646,2005.
[18]G.Sabidussi.The centrality index of a graph.
Pyschometrika,31:581–603,1966.
[19]D.J.Watts and S.H.Strogatz.collective dynamics of
small world networks.Nature,393:440–442,June1998.
SUNDING自行车码表按键设置说明 (1)如何开机：按右键或按左键。 在开机状态下，码表总会显示瞬时速度。 如果4分30秒内没有运动信号输入，屏幕自动进入只显示时间的关闭状态，(2)如何转换显示模式：正常状态下，按右键依次变换显示模式， 码表左侧没有字母指示的是：时间模式 码表左侧显示字母ODO的是：总里程模式 (最多可显示99999公里) 码表左侧显示字母DST的是：单次里程模式 (最多可显示9999公里) 码表左侧显示字母MXS的是：单次最大速度模式 码表左侧显示字母AVS的是：单次平均速度模式 (当瞬时速度为0时，不计入平均速度) 码表左侧显示字母TM的是：单次行车时间模式(最多可显示100小时。当瞬时速度为0时，不计入行车时间) 码表左侧显示字母SCAN的是：自动循环显示模式 (每4秒变换一次↘DST →MXS→AVS→TM↖) (3)如何调整时间：在显示时间的模式下，长按左键3秒，进入时间调整状态。 第一步、24小时/12小时模式转换：按左键可进行24小时/12小时模式转换，按右键则跳过第一步，进入第二步。 第二步：调整小时：按左键可进行小时调整，按右键则跳过第二步，进入第三步。 第三步：调整分钟：按左键可进行分钟调整，按右键则跳过第三步，时间调整结束。 (4)如何调整总里程ODO的初始值：在显示总里程ODO的模式下，长按左键3秒，进入总里程ODO初始值的调整状态。 第一步、小数点后一位数闪动，按右键可进行小数点后一位数的初始值设定，按左键确认设定，并跳过第一步，进入第二步。 第二步：个位数闪动，按右键可进行个位数的初始值设定，按左键确认设定，并跳过第二步，进入第三步。 第三步：十位数闪动，按右键可进行十位数的初始值设定，按左键确认设定，
电商平台资料对接接口（草稿） *备注： 0.默认通信协议为https或者 http 1.如无特殊说明，传入参数与返回结果数据皆为标准Json格式 2.接口调用权限验证规则与方式由你方制定，为可选项 3.接口返回结果中的状态码（例如：200）与描述信息（例如：成功），应简洁明确，方便调试 一、基础商品资料上传接口 调用情景： ?-线下总部通过后台系统筛选出能够在线上电商平台销售的商品资料，通过后台系统交互界面批量同步到电商平台，等待后续电商平台上架操作 传入参数： ?-权限验证信息 ?-预设处理信息（1.单次传输商品数，用于校验；2.数据处理方式：测试、正常） ?-商品资料信息（单个商品信息包含：名称，价格，库存，条码，结算价格，零售价格，成本价格，品牌，商品编码，分类（大、中、小），规格，单位，包装数，物流包装，产地，等级） 返回结果： ?-状态码 ?-描述信息
二、商品下架接口 调用情景： ?-门店在线下后台系统进行商品下架后，后台系统自动告知电商平台商品同步下架 传入参数： ?-权限验证信息 ?-预设处理信息（1.门店信息；2.数据处理方式：测试、正常）?-下架商品信息列表 返回结果： ?-状态码 ?-描述信息 三、商品资料变更接口 调用情景： ?-门店在线下后台系统对商品信息进行修改（商品名称、结算价格、零售价等），后台系统定时传送商品变更日志到电商平台 传入参数： ?-权限验证信息 ?-预设处理信息（1.门店信息；2.数据处理方式：测试、正常）
?-变更商品信息列表 返回结果： ?-状态码 ?-描述信息 四、门店基础信息上传接口 调用情景： ?-门店使用线下后台系统交互界面将门店基础信息导入线上电商平台 传入参数： ?-权限验证信息 ?-预设处理信息（1.门店信息；2.数据处理方式：测试、正常）?-门店基础信息列表 返回结果： ?-状态码 ?-描述信息 五、商品价格、库存同步接口 调用情景： ?-门店后台系统定时同步各门店商品价格、库存信息到电商平
【品名】SIGMA/西格玛BC 506码表功能介绍： 1、可显示速度、骑行距离（最大99，999KM）、骑行时间（最大999：59小时）、累计骑行里程（最大99，999KM），时钟功能。 2、手动和自动显示功能。自动显示（AUTO）,自动显示功能开启后，能间隔1秒，顺序显示骑行距离、时间、总里程和时钟。 一、功能显示。 1 、自动显示。按动功能按钮（码表下方大按钮），只至出现AUTO字样，此时码表自动显示骑行距离、时间、总里程和时钟，间隔时间为1秒。 2、手动显示。按动功能按钮，出现KM字样，表示此时显示的是旅行距离；按动功能按钮，出现闹钟图案，表示此时显示的是旅行时间，精确显示到秒；按动功能按钮，出现Σ字样，表示此时显示的是从装上码表开始的累计骑行距离。 3、码表清零。按住功能按钮不小于3秒，此时码表上的数字闪烁，继续按住不动直到清零，清零不影响累计骑行距离。 二、设定。 此功能决定着码表的正确使用，非常关键。主要功能是输入车轮周长，调校时钟、改变速度单位、开启自动显示功能等。进入设置界面。按住码表背面的设置按钮3秒，出现SET字样，即进入设置界面，此时下面显示的四位数字为车轮周长，此时按动功能按钮，将进入下一个功能的设置界面。下面以设置车轮周长为例，介绍此功能的使用。在设置界面下，按动设置按钮，码表上的数字将开始闪动，表示此时可以输入数字，按动功能按钮，闪动的数字会改变，我的车胎是26*2.1，应在第一个数字位输入2，等待第二个数字闪动，按动功能按钮，第二位输入1，依次在第三位和第四位输入3，即周长为2133。按照以上的方法，可以调校时钟等，因方法一样不再重复。退出设置界面。按动设置键3秒，即可恢复到使用状态。 德国SIGMA BC506型自行车码表使用说明(5项功能) 功能介绍： 1、可显示速度、骑行距离（最大99，999KM）、骑行时间（最大999：59小时）、累计骑行里程（最大99，999KM），时钟功能,手动和自动显示功能。自动显示（AUTO）,自动显示功能开启后，能间隔1秒，顺序显示骑行距离、时间、总里程和时钟。 2、拥有7种语言显示，英里、公里转换，更大的显示数字，低电压显示功能，同样出色的防水设计，传输线90厘米 ★SIGMA BC506 中文使用说明 MODE功能: *AUT更改自动设定 *KM/M: 单一旅程距离 *RIDETIME : 骑乘时间 *TOTAL TRIP : 总哩程数 *CLOCK : 时间 RESET(重新设定)—需超过3秒: * KM/M: 单一旅程距离 *RIDETM : 骑乘时间
U S B接口电路电路 Company Document number：WUUT-WUUY-WBBGB-BWYTT-1982GT
左边这张图，过了保险丝以后，接了一个470uF 的电容C16，右边这张图，经过开关后，接了一个100uF 的电容C19，并且并联了一个的电容C10。其中C16 和C19 起到的作用是一样的，C10 的作用和他们两个不一样，我们先来介绍这2 个大一点的电容。容值比较大的电容，理论上可以理解成水缸或者水池子，同时，大家可以直接把电流理解成水流，其实大自然万物的原理都是类似的。作用一，缓冲作用。当上电的瞬间，电流从电源处流下来的时候，不稳定，容易冲击电子器件，加个电容可以起到缓冲作用。就如同我们直接用水龙头的水浇地，容易冲坏花花草草的。我们只需要在水龙头处加个水池，让水经过水池后再缓慢流进草地，就不会冲坏花草，起到有效的保护作用。 作用二，稳定作用。我们一整套电路，后级的电子器件功率大小、电流大小也不一样，器件工作的时候，电流大小不是一直持续不变的。比如后级有个器件还没有工作的时候，电流消耗是100mA，突然它参与工作了，电流猛的增大到150mA 了，这个时候如果没有一个水缸的话，电路中的电压(水位)就会直接突然下降，比如我们的5V 电压突然降低到3V 了。而我们系统中有些电子元器
件，必须高于一定的电压才能正常工作，电压太低就直接不工作了，这个时候 水缸就必不可少了。电容会在这个时候把存储在里边的电流释放一下，稳定电 压，当然，随后前级的电流会及时把水缸充满的。有了这个电容，可以说我们 的电压和电流就会很稳定了，不会产生大的波动。这种电容 常用的有以下三种： 图3-这三种电容是我们常用的三种电容，其中第一种个头大，占空间大，单位容量价格最便宜，第二种和第三种个头小，占空间小，性能一般也略好于第一种，但是价格也贵不少。当然，除了价格，还有一些特殊参数，在通信要求高的场合也要考虑很多，这里暂且不说。我们板子上现在用的是第一种，只要在符合条件的情况下，第一种470uF 的电容不到一毛钱，同样的耐压和容值，第二种和第三种可能得1 块钱左右。 电容的选取，第一个参数是耐压值的考虑。我们用的是5V 系统，电容的耐压值要高于5V，一般推荐倍到2 倍即可，有些场合稍微高于也可以。我们板子上用的是10V 耐压的。第二个参数是电容容值，这个就需要根据经验来选取了，选取的时候，要看这个电容起作用的这块系统的功率消耗情况，如果系统耗电较大，波动可能比较大，那么容值就要选大一些，反之，可以小一些。 刚开始同学们设计电路也模仿别人，别人用多大自己也用多大，慢慢积累。比如咱上边讲电容作用二的时候，电流从100mA 突然增大到150mA 的时候，其实即使加上这个电容，电压也会轻微波动，比如从5V 波动到，但是只要我们板子上的器件在电压以上也可以正常工作的话，这点波动是没有问题的，但是如果不加或者加的很小，电压波动比较大，有些器件就会工作不正常了。但是如果加的太大，占空间并且价格也高，所以这个地方电容的选取多参考经验。
前言 助友U3ERP生产增强型，在MRP准确后，即可对其自动生成的生产计划进行排产，缺省只排已经齐套的生产任务,可选择排产全部生产任务。 U3ERP中的排产，根据需求日期，参考已经安排的情况，排出工艺路线的每个工序如何进行加工，给出设备，加工起止时间，加工数量等，这些数据是MES的数据来源。 助友MES使用方法： 0.因MES是助友U3ERP生产增强型的组成部分，因此是通过使用U3ERP实现的。 1.根据U3ERP排产结果，下发一定期限内的加工任务，并根据U3ERP操作提示对加工任务进行审批及确认 2.根据操作提示，在工作票派工模块，依据排产内容，按设备分别生成工单，并指定给某个人负责 3.打印制造件的流程单（“车间管理->车间任务令处理->在制品流程单打印”），找到要生产的WIP,打印其工艺流程单 4.加工人员，或车间物料员凭工艺流程单到仓库领料， 5.根据操作提示，在“车间任务令出库”模块，仓库人员根据加工件的物料编码找到对应的WIP，点“生成出库单”按钮，生成出库内容，库管员根据出库内容及提示的位置取出物料，交给领料人员，并在流程单中标识“已领料”，若出库数量与要求数量不一致，可多次出库，具体标识方法，企业可自行规定 6.根据工艺流程单上的工艺路线，按工艺指定的设备进行对应工序的加工 7.工序加工结果录入到U3ERP中（“车间管理->生成完工统计->工作票完工录入”），可作为员工计件的依据 8.若工序需要进行完工检验，则根据操作面板的提示进行检验 9.制造件的全部工序都加工完成，则该制造件可以进行完工入库 10.当前工序返工处理，可以在工序完工检验时，设置返工数量，系统自动增加当前工序的可完工数量，直接在工序完工录入模块录入返工的完成情况 11.加工件的不良返工 a.在“车间管理->任务令处理->任务令维护”模块，手工“新建维修任务令”，选择要维修的加工件编码，输入要维修的数量，选择具体的工艺，并指定需要的加工时间，提交 b.维修任务令批准后，应先在工单派工模块指定每道工序使用的设备及起止时间，然后再派工 c.维修任务令的工序完工后，也需要录入完工数据，但工序的加工单价也需要手工输入，因维修任务令的工序没办法事先确定，系统不能控制各个工序的加工顺序，应以实际加工为准
西格玛无线自行车码表BC1200 BC 1200中文說明書 MODE 功能： * DIST/DAY : 單一旅程距離 *RIDERTIME : 騎乘時間 *A VGSPEED : 平均速顯示 *MAXSPEED : 最高速率顯示 *STPWATCH : 自動計時功能 *TRIP UP/ DOWN : 可設定的固定旅程距離(可向前計程或以倒數的方式測量) *TOTALODO : 總哩程數 *CLOCK : 時間 RESET(重新設定): *DIST/DAY : 單一旅程距離 *RIDERTIME : 騎乘時間 總功能： DIST/DAY : 單一旅程距離 RIDERTIME : 騎乘時間 A VGSPEED : 平均速顯示 MAXSPEED : 最高速率顯示 STPWA TCH : 自動計時功能 TRIP UP/ DOWN : 可設定的固定旅程距離(可向前計程或以倒數的方式測量) TOTALODO : 總哩程數 CLOCK : 時間 Language : 語言設定 km/h , mph : 速度 語言設定： 語言設定及輪子尺寸設定原廠設定為德文，若要執行以下動作，請先更改為英文) *按MODE直到DIST/DAY出現於屏幕上。 *按背面的」WS1/2」直到「WS1」出現。 *用有尖頭的工具按住背面的」S 「3秒。 *出現」SET LANGUAGE」，按RESET 輸入你所想要的語言。 *按MODE確認語言，按RESET進入公里(KMH)/英哩(MPH)設定，按MODE確認。 設定輪圈周长尺寸： *顯示標準輸入的輪圈周长(SET WS) *用直尺量出輪圈直徑大小(请连轮胎厚度一起量度)，乘以3.14,得出车轮的周长,並輸入此號碼。单位:MM（毫米）*按MODEL進行下一步。 *按」S」完成設定。 *用尖頭工具按背面的(WS1/2)，屏幕上將會出現WS 2。 *相同方式選擇WS1。 設定時鐘： *按MODE 直到顯示CLOCK。 *用尖頭工具按住背面的「S」3秒。 *按RESET輸入小時。
常见串口接口电路设计集锦 六种常用串口接口电路1、并口接口(分立元件) 适用于Windows 95/98/Me 操作系统。这个电路与FMS 随软件提供的电路比多了一个200K 的电阻，这个主要是为了与JR 的摇控器连接，因为JR 的摇控器教练口好象是集电极开路设计的，需要加一只上拉电阻才能正常工作。 不过电路还是满简单的，用的元件也很少，很适合无线电水平不太高的朋友们 制作，只是不能用于Win2000/XP 上有点让人遗憾。 2、串口接口(分立元件)字串5 适用于Windows 95/98/Me 操作系统，电路也不是很复杂，当然元件比并口电路多了一些，而且串口的外壳比并口小很多，如何把这些元件都放到小 小的外壳里免不了要大家好好考虑一下了。当做体积小也是它的最大的优点， 而且不用占用电脑并口，因为现在还有一些打印机还是要用并口的。缺点同样 是不支持Win2000/XP。 3、串行PIC 接口(使用PIC12C508 单片机)字串9 适用于Windows 95/98/Me/2000/XP 操作系统。电路简单，只是用到MicroChip 公司的PIC12C508 型单片机，免不了要用到编程器向芯片里写程序了，这个东西一般朋友可能没有，不过大多卖单片机的地方都有编程器，你只 要拿张软盘把需要用的HEX 文件拷去让老板帮你写就可以了。这个接口最大 的优点就是支Win2000/XP 操作系统，还可以用PPJOY 这个软件来用摇控器虚拟游戏控制器玩电脑游戏。 4、25 针串行PIC 接口(使用PIC12C508 单片机) 适用于Windows 95/98/Me/2000/XP 操作系统。电路同9 针的接口基本一样，只不过是接25 针串口的，现在用的不是很多了。
1.社会服务系统对接接口设计 系统能提供兼容不同技术架构的数据接口，保证系统与省级各联合审批职能部门及其他电子政务系统进行数据交换。 1.1. 数据交换接口 数据交换平台基于Java技术和标准数据库接口（JDBC、ODBC等），为不同的数据库系统、应用系统、专用中间件系统提供接入组件，通过对接口协议需求进行抽象，使用TongIntegrator框架，就可以和特定系统的交互。另外提供组件定制接口，可以方便、快速地添加具有新的功能的组件。数据交换平台提供了大量的扩展接口，方便用户进行功能扩展。 1.1.1. 提供企业级需求的标准接口 数据压缩，减少带宽瓶颈；数据加密，提高系统安全性；异常处理，创建和维持了一个“消息异常处理器”的接口，它可以保存因为某种原因不能处理的消息，这些“异常”消息还可以被送回重新加以处理。 1.1. 2. 提供可扩展的告警方式接口 平台默认实现了邮件告警方式，只需要配置相应的邮件信息，当有警告产生时，会自动发送告警邮件给邮件接收者。同时平台还提供了可扩展的告警方式接口，可根据项目需要扩展不同的告警方式，如短信告警等。 1.1.3. 提供第三方的压缩和加密算法接口 提供数据压缩和加密功能，产品本身带有一套数据压缩、加密算法，同时也为第三方的压缩和加密算法提供了接口，用户可以方便的将自己指定的压缩和加密算法嵌入到系统中。 1.1.4. 系统特点 易于维护 通过使应用松耦合或分离，使系统环境中的接口更容易维护。同时通过数据交换平台对外提供统一接口，屏蔽了单个系统内部的改变，可以很容易替换过时的应用。 可扩展 数据交换平台提供了大量的扩展接口，方便用户进行功能扩展。
数据对接接口说明 1.试剂管理平台接口概述 试剂管理平台（以下简称“平台”）集试剂采购、审批、库房管理、废弃物处置、结算、资料查询、安全教育宣传于一体的、量身定制的信息化管理平台。“平台”以“方便师生，寓管理于服务，以服务促管理”作为指导思想，通过简化、优化采购、审批等各环节流程，透明、规范采购，实现试剂全程可追溯、全过程闭环管理。 为保证“平台”供货商产品数据更新的及时性，现将其中部分功能数据对接接口的方式向供货商提供，具体接口如下表所示： ，并获取一个秘钥（userKey）。接口成功部署后，可通过访问 http://ip:port/services/frontWebService?wsdl获取接口的详细描述。 2.数据对接方法 2.1.String sayHi(String name) 这是一个测试方法，返回"hello, " + name的字符串,测试地址为： http://ip:port/services/frontWebService/sayHello?name=J 2.2.String submit(String xmlData, String sign) 主要的业务处理方法，后面所说的xml报文，即该方法的xmlData参数，sign 为xmlData+userKey的md5密文。返回值为xml格式的字符串。 3.XML报文定义规则 3.1.请求报文
[考试]cateye猫眼码表设置说明书cateye猫眼码表设置说明书 CATEYE MITY3 和 ENDURO2码表的使用说明 1.MITY3和ENDURO2的初始设置 (a):第一次设置码表的微处理器，首先同时按住3个键清除所有设置 提示:把码表倒置，按住ST。/STOP 和MODE键，然后用铅笔或其他尖锐物戳SET键 (b):液晶屏上的显示会闪一下然后消失，只剩下一个闪动的K。K表示公里/小时,M表示英里/小时，用MODE键选择您所需的速度制式。注意您所选择的制式的指令时钟功能为12小时或24小时制，K为24小时制，M为12小时制。 按ST./STOP键确定您选择的制式。 (c):\"210\"将会闪动，这是为了调整700*23C轮胎设置的。请在您的手册或我们的网站中查询轮径参数。请注意这个数字事实上是您骑车时的车胎旋转滚动一周的距离(厘米)。如果您想输入最精确的参数请您动手测量。 (d):按MODE键增加参数，按ST./STOP键减少参数。按码表背后的SET键确定参数。 现在您的码表已设置完备。 2.按键使用说明 码表右边的按键是MODE键，左边的按键是ST./STOP键。您可以用MODE键通过模式完成控制。 当屏幕显示耗用时间功能(TM)时，按MODE键转换到平均速度功能(AV)，再次按MODE键转换到里程1功能，这是标准骑行里程功能。此时按MODE键回到时间功能。 耗用时间功能的附属功能是时钟功能。在耗用时间功能下按住MODE键2秒，即转为时钟功能，再次按MODE键则回到耗用时间功能。
平均速度功能的附属功能是最大时速功能(MX)，使用方法同上。 里程1有2个附属功能。按住MODE键2秒即转为里程2。 里程2是一个单独的计程功能，它能在别的参数保留的情况下单独清空。只要同时按住ST./STOP和MODE键一秒即可。 在里程2功能下，按住MODE键2秒转为总里程功能(ODO)。 在里程2或总里程功能下按MODE键即可回到里程1功能。 3 (1) 设置功能和里程功能 同时按住ST。/STOP和MODE键一秒即可清零。 在标准模式下，您必须按ST./STOP(S)键从而开始记录耗用时间和里程 在自动模式(AT)下，当微处理器接收到前叉上传感器的信号则开始或结束记录按码表背后的SET键即可开启或关闭耗用时间或里程功能的自动功能。一个小的AT符号指示该功能。 (2) 设置时钟功能 当使用该功能时码表必须停止工作，速度符号(K或M)必须停止闪动。在耗用时间功能下按住MODE键2秒，即转为时钟功能。 您将看到一个小的时钟符号。小时开始闪动，按MODE键增加数字。按 ST./STOP键将从小时该到分钟。最后按SET键确认。 时钟是12小时或24小时制取决于您所选择的制式的指令时钟功能的制式。K 为24小时制，M为12小时制。(参见 b 注释) (3)设置耗用时间，最大时速和平均速度功能为较高的显示 当电脑处于自动功能时，按ST。/STOP键则耗用时间，最大时速和平均速度功能将转换为较高的显示，再次按ST./STOP键你 能更改速度为较高的显示。 在标准模式下(自动功能关闭)按住ST./STOP键2秒亦可。
典型接口电路EMC设计 一、以太网接口EMI设计 100M网口设计时必须设计Bob smith 电路：可以产生10dB的共模EMI衰减，为了更好的抑制共模信号通过线缆对外的辐射应注意下面几点： 1 、不用的RJ45管脚4 、5、7、8按下图的方法处理。 2 、物理芯片侧的变压器中心抽头需通过0.01uF－0.1uF的电容接地。 3 、物理芯片侧的差模电阻（收端）应等分为二（100分为两个49.9），中心点通过1000pF 电容接地。 以太网口Bob smith电路原理图 以82559为例说明网口设计PCB注意点，布局如下： 以太网口布局示意图
A、B要求尽量短，A不得超过1英寸，B可以根据实际情况放宽。接口变压器PCB设计如下： 以太网口变压器布局示意图 布局要求： PCB布局示意图 布线要求： 1、变压器下面全部掏空处理，其余隔离带的宽度大于100mil； 2、连接器与隔离变压器之间距离小于1000mil； 3、晶振距离接口变压器和板边大于1000mil； 4、灯线不要走到变压器下面，并且尽量不要与差分信号线同层走线，如果同层走线，需要与差分信号线相距30mil以上； 5、差分信号线与变压器输出侧的过孔距离大于40mil。
二、以太网口的防护设计 加防护电路的设计： 增加防护器件电路原理图 以上器件选型要求： 1、变压器要选用隔离耐压3000Vac要求的。 2、气体放电管尽量选用3端气体放电管，启动电压为90V的； 3、TVS管选用SLV2.8-4； 三、485接口电路设计 对于出户外的485端口,进行如下设计，采取气体放电管加TVS管加限流电阻组合方式。选用90V陶瓷管（3R090）可承受10/700us，8KV雷击测试；64V固体管（P0640）只能承受10/700us，3KV雷击测试 。TVS的选择为P6KE6.8CA ，去耦电阻选择为10Ω/1W 。
BC1009 STS 码表使用说明书 说明书中所涉及英文 一、初始设定 1.BC1009码表装入电池后起始ENGLISH 画面（图1） 图1 图2 2、按“MODE 1”键（图2）选择ENGLISH （语言设定）、KMH （速度单位）、WHEEL SIZE （自行车轮圈尺寸）、CLOCK （时钟设定）、TOTAL ODO （自行车总骑行里程）、TOTAL TIME （自行车总骑行时间）、CONTRAST （显示对比度）其中一项进入设定。 1、ENGLISH （语言设定）： 按一下“SET ”键(图3)进入语言设定，画面能改动的地方开始闪动，按“RESET ”键或“MODE 2”键（图4）在ENGLISH （英语）、FRANCAIS （法语）、ITALIANO （意大利语）、ESPANOL （西班牙语）、SVENSK （瑞典语）、HOLLANDS
（荷兰语）、DEUTSCH（德语）之间任一选择→按“SET”键（图5 ）确认，画面显示SET OK。 图3 图4 图5 2、KMH公里/MPH英里（速度单位） 按“MODE 1”键（图6）移到KMH选项,按一下“SET”键（图7）进入速度单位设定，画面能改动的地方开始闪动，按“RESET”键或“MODE 2”键（图8）在KMH公里/MPH英里之间选择→按“SET”键（图9）确认，画面显 示SET OK（设置成功） 注：从“KMH”切换至“MPH”时，距离格式会自动从“公里”切换至“英里”，时间格式也会从“24h”切换成“12h”。正常习惯使用“KMH” 图6 图7 图8 图9 3、WHEEL SIZE（自行车轮圈尺寸） 按“MODE 1”键（图10）移到WHEEL SIZE选项,按一下“SET”键（图11）进入自行车轮圈尺寸（26*1.95 205厘米）设定，画面能改动的地方开始闪动，按“RESET”键或“MODE 2”键（图12）选择您的轮圈尺寸，按“MODE 1”键（图13）在轮圈尺寸的4位数间切换→按“SET”键（图14）确认，画面显示SET OK（设置成功） 图图图图图
接口表、错误信息表 Table List: *不要使用xxx_temp接口表，如mtl_material_transactions_temp；用它们会绕过系统id验证，直接进入数据表 Table Relation: 仅启用批次 mtl_transactions_interface.transaction_interface_id = mtl_transaction_lots_interface.transaction_interface_id 仅启用序列 mtl_transactions_interface.transaction_interface_id = mtl_serial_numbers_interface.transaction_interface_id 同时启用批次、序列 mtl_transactions_interface.transaction_interface_id = mtl_transaction_lots_interface.transaction_interface_id mtl_transaction_lots_interface.serial_transaction_temp_id = mtl_serial_numbers_interface.transaction_interface_id mtl_transactions_interface select mti.error_code,mti.error_explanation from mtl_transactions_interface mti; 并发程序Process transaction interface N: INV/Setup/Transactions/Interface Managers/Process transaction interface/Tools/Launch Manager 这个是库存事务处理主程序，正式环境一般是Schedule运行的。
系统对接设计 1.1.1对接方式 系统与外部系统的对接方式以web service方式进行。 系统接口标准： 本系统采用SOA体系架构，通过服务总线技术实现数据交换以及实现各业务子系统间、外部业务系统之间的信息共享和集成，因此SOA体系标准就是我们采用的接口核心标准。主要包括： 服务目录标准：服务目录API接口格式参考国家以及关于服务目录的元数据指导规范，对于W3C UDDI v2 API结构规范，采取UDDI v2的API的模型，定义UDDI 的查询和发布服务接口，定制基于Java和SOAP的访问接口。除了基于SOAP1.2的Web Service接口方式，对于基于消息的接口采用JMS或者MQ的方式。 交换标准：基于服务的交换，采用HTTP/HTTPS作为传输协议，而其消息体存放基于SOAP1.2协议的SOAP消息格式。SOAP的消息体包括服务数据以及服务操作，服务数据和服务操作采用WSDL进行描述。 Web服务标准：用WSDL描述业务服务，将WSDL发布到UDDI用以设计/创建服务，SOAP/HTTP服务遵循WS-I Basic Profile 1.0，利用J2EE Session EJBs实现新的业务服务，根据需求提供SOAP/HTTP or JMS and RMI/IIOP接口。 业务流程标准：使用没有扩展的标准的BPEL4WS，对于业务流程以SOAP服务形式进行访问，业务流程之间的调用通过SOAP。 数据交换安全：与外部系统对接需考虑外部访问的安全性，通过IP白名单、SSL 认证等方式保证集成互访的合法性与安全性。
数据交换标准：制定适合双方系统统一的数据交换数据标准，支持对增量的数据自动进行数据同步，避免人工重复录入的工作。 1.1.2接口规范性设计 系统平台中的接口众多，依赖关系复杂，通过接口交换的数据与接口调用必须遵循统一的接口模型进行设计。接口模型除了遵循工程统一的数据标准和接口规范标准，实现接口规范定义的功能外，需要从数据管理、完整性管理、接口安全、接口的访问效率、性能以及可扩展性多个方面设计接口规格。 1.1. 2.1接口定义约定 客户端与系统平台以及系统平台间的接口消息协议采用基于HTTP协议的REST风格接口实现，协议栈如图4-2所示。 图表错误!文档中没有指定样式的文字。-接口消息协议栈示意图系统在http协议中传输的应用数据采用具有自解释、自包含特征的JSON数据格式，通过配置数据对象的序列化和反序列化的实现组件来实现通信数据包的编码和解码。 在接口协议中，包含接口的版本信息，通过协议版本约束服务功能规范，支持服务平台间接口协作的升级和扩展。一个服务提供者可通过版本区别同时支持多个版本的客户端，从而使得组件服务的提供者和使用者根据实际的需要，独立演进，降低系统升级的复杂度，保证系统具备灵活的扩展和持续演进的能力。
AP_INVOICES_INTERFACE AP_INVOICE_LINES_INTERFACE 涉及的请求： 应付款管理系统开放接口导入 涉及案例：运费导AP、费用导AP PO接口表： 申请： PO_REQUISITIONS_INTERFACE_ALL 涉及请求： 导入申请 采购： po_headers_interface po_lines_interface po_distributions_interface 涉及的请求： Import Standard Purchase Orders 接收： rcv_headers_interface rcv_transactions_interface mtl_transaction_lots_interface
接收事务处理处理器 涉及案例：运费导采购、MRP导申请、POP导申请 GL接口表： gl_interface 涉及案例：ADI导日记账、返利导日记账 FA接口表： fa_mass_additions FA API： 增加fa_addition_pub.do_addition 修改：fa_adjustment_pub.do_adjustment fa_asset_desc_pub.update_desc 涉及案例：电子资产清理 库存事务接口：mtl_transactions_interface 1）一般用来做各类杂收发、Cost Update，对于和业务有关的事务一般不建议使用，比如SO发货，如果自己发会导致Workflow没有往下走
2）成功导入之后运行Cost Manager生成会计分录 3）平均成本更新也通过此接口；如果该Item没有交易，则成本数据不会进入cst_item_costs 固定资产接口：fa_mass_additions 1）fa_mass_additions这个表有几个Trigger需要注意；会自动去插其他表，删除的时候也是 2）折旧方法接口表无法给，而是自动从Category继承下来，所以导入之后需要Update 表 3）不是通过AP引过来的FA，是没有Source Lines信息的；如果需要，可以通过插表来实现 4）数据进接口之后从Navigator:Mass Additions/Post Mass Additions提交请求集，这样会有个报表显示导入结果。不过我的测试程序可以直接提交请求集。 总账分录接口：gl_interface 1）最简单的接口，不说了 应付发票接口：ap_invoices_interface/ap_invoice_lines_interface 1）Profile AP: Use Invoice Batch Controls定义是否使用批控制 2）仅支持CREDIT/STANDARD；注意金额正负，小于零用CREDIT，大于零用STANDARD 3）PREPAYMENT/DEBIT/EXPENSE REPORT/MIXED不支持，先用CREDIT或者STANDARD，进去之后再Update正式表ap_invoices_all 4）导入之后可再调用请求Invoice Validation自动Validate发票
库存事务接口：mtl_transactions_interface 1）一般用来做各类杂收发、Cost Update，对于和业务有关的事务一般不建议使用，比如SO发货，如果自己发会导致Workflow没有往下走 2）成功导入之后运行Cost Manager生成会计分录 3）平均成本更新也通过此接口；如果该Item没有交易，则成本数据不会进入cst_item_costs 固定资产接口：fa_mass_additions 1）fa_mass_additions这个表有几个Trigger需要注意；会自动去插其他表，删除的时候也是 2）折旧方法接口表无法给，而是自动从Category继承下来，所以导入之后需要Update表3）不是通过AP引过来的FA，是没有Source Lines信息的；如果需要，可以通过插表来实现 4）数据进接口之后从Navigator:Mass Additions/Post Mass Additions提交请求集，这样会有个报表显示导入结果。不过我的测试程序可以直接提交请求集。 总账分录接口：gl_interface 1）最简单的接口，不说了 应付发票接口：ap_invoices_interface/ap_invoice_lines_interface 1）Profile AP: Use Invoice Batch Controls定义是否使用批控制 2）仅支持CREDIT/STANDARD；注意金额正负，小于零用CREDIT，大于零用STANDARD 3）PREPAYMENT/DEBIT/EXPENSE REPORT/MIXED不支持，先用CREDIT或者STANDARD，进去之后再Update正式表ap_invoices_all 4）导入之后可再调用请求Invoice Validation自动Validate发票 应收发票接口：ra_interface_lines_all/ra_interface_salescredits_all 1）几个接口表之间的关系和其他接口不同，不是通过ID关联，而是通过描述性弹性域Line Transaction Flexfield定义 2）接口给ID还是Name，要根据Transaction Source上的定义，不能随便给 3）ra_interface_distributions_all这个表可以不插，让系统自动生成分配行 4）如果没有销售员，可以提供一条ID=-3的No Sales Person记录给接口。这个要看AR System Option Miscellaneous上的设置是否要求销售员 应收收据接口：ar_interim_cash_receipts_all 1）AR Receipt标准接口是Lockbox，需要预先定义Format；Oracle根据Format定义把接口数据写到ar_interim_cash_receipts_all；我们可以跳过第一步，直接把数据插入到ar_interim_cash_receipts_all，不过Batch需要我们直接插表 2）在Batch界面查看：数据在ar_interim_cash_receipts_all的Batch Type为 Mannual
最近一段时间一直在做APP接口，总结一下APP接口开发过程中的注意事项： 1、效率：接口访问速度 APP有别于WEB服务，对服务器端要求是比较严格的，在移动端有限的带宽条件下，要求接 口响应速度要快，所有在开发过程中尽量选择效率高的框架，PHP建议使用YAF框架。 2、数据格式 最好使用JSON格式数据，因为JSON有较好的跨平台性。对于 3、数据量 按需分配，APP客户端需要什么数据就返回什么数据，过多的数据量影响处理速度，最重要的 是影响传输效率。 4、接口、参数命名准确 无论是接口还是参数，命名都应该有意义，让人一目了然。 5、一个页面尽可能就用一个接口 现在很多的APP页面都有广告、焦点图、文章列表等，对于这些不同格式的数据，不可能都分 配一个接口，这样加大了APP请求接口数，影响响应速度。建议服务器端尽可能处理好数据后 通过一个接口返回给APP客户端。 6、缓存 这点比较重要，不管是文件缓存还是memcache缓存。 7、接口要有可扩展性 8、接口安全 目前一般都是在APP客户端和服务器通过约定的算法，对传递的参数值进行验证匹配。但是如 果APP程序被反编译，这些约定的算法就会暴露，特别是在安卓APP中，有了算法，完全就 可以通过验证模拟接口请求。 9、接口版本控制 对于接口版本控制，自己目前也没有找到一个好的方法，怎么去应对不断的APP版本升级，新、旧接口的处理。 10、接口数据、状态 接口必须提供明确的数据状态信息，不管是成功的，还是失败的，都必须返回给APP客户端。 以上10点就是自己在这端时间做APP接口过程中注意的事项，写的有点乱，想到什么就写什么。
一款简单实用、稳定可靠、成本低廉的 主站M-BUS接口电路 随着智能表越来越多的使用，各种类型的抄表器（既M-BUA主站）需求也随之增加。M-BUS 接口电路作为抄表器的一个主要模块，决定了抄表器性能的好坏，也较为影响抄表器的成本高低。现今大多数抄表器都是延用TI推荐的M-BUS接口电路方案（或是做了一些小的修改），该方案电路复杂，成本也较高，并不太适合大众化抄表器的使用。 笔者根据M-BUS的工作原理，结合自身多年的电路开发经验，设计出一款简单实用、稳定可靠、成本低廉的主站M-BUS接口电路。这款接口电路经电路模拟仿真以及实际抄表测试，性能良好，工作可靠，完全可以替代TI的M-BUS接口电路方案。 电路原理 根据主站M-BUS的工作原理： 发送：传号电压：24V~36V（CJ-T188-2004：20.8V~42V） 空号电压：传号电压－12V（CJ-T188-2004：传号电压－10V）接收：传号电流：≤1.5mA 空号电流：11~20mA 1．发送电路 发送电路的设计主要需要考虑的问题有：发送传、空号电压的变化量要大于等于12V （10V）；电路的驱动能力，几十上百个智能表不能影响发送电压低于12V。 用一个直流稳压器应该可以满足这些要求。图1是发送电路框图。 V1输出正12V稳压 电压，V2输出负18V稳 压电压。当搬动开关K 时，总线上的电压即可 在30－18V间变换。由 于采用稳压电源，负载 变化只要不超过稳压电 源的驱动能力，总线上 K 图1 图2
的电压不会有大的变化。一个5～10W 的直流稳压电源足以驱动100只智能表。 图2是发送电原理图。在这个电路中Q2设定为具有简单稳压能力的射随器，发射极的电压是随基极电压而变化。传号时，串口TX 端为高电平，Q3不导通，R4、R5分压使Q2的基极电平在12.7V 左右，BUS+端的电平即可保持在12V 左右。由于BUS-端的电平为-18V，所以R1上获得电压为30V 左右。空号时，串口TX 端为低电平，Q3导通，Q2的基极电平为零，BUS+端的电平即可保持在－0.7V 左右。由于BUS-端的电平始终为-18V，R1上获得电压为17.3V 左右。空号比传号电压低了12.3V 左右。 当调整VSS 电平为－12V 时，R1上的传号电压为24V。若增加VSS 电平为－24V 时，R1上的传号电压可达36V。事实上无论VSS 电平是多少，只要不大于0，空、传号电压差值都是12V，不会受VSS 影响。 改变R4、R5分压值为10.7V，即可使空、传号电压差值为10V 左右。 稳压电源的过载能力和短路保护能力直接对BUS 总线施加影响，不必再另外设计保护电路。 2． 接收电路 接口电路接收部分的主要难点是不要误读。在一个稳压电路中要读取它的电流，就象是在测它的纹波一样，串一个取样电阻是需要的。稳压电路本身纹波很小，取样电阻大了影响稳压效果，小了有取不出可用值。更要命的是，负载大小引起的电流变化远比接收电流引起的电流变化来的大，直接影响直流电平偏移，造成接收波形失真。一个25欧姆的采样电阻，当接收电流变化20MA 时，会产生0.5V 的接收电压；当负载变化100MA 时，会产生2.5V 的接收电平偏移变化。 负载变化往往是缓慢的，智能表安装完毕通常不会再变，而接收信号在300～9600Hz 。采用电容耦合的方式，可以有效地避免负载变化对接收波形的影响。 图3是接收波形采集。 接收电流采样电阻R7被放置在Q2的集电极回路，不会对发送电压产生影响。同时R7一端是交流接地，也便于电容直接耦合。 最终的设计见图4。Q5、Q6完成接收波形的放大整型。C2和R14组成截止频率很低（1Hz 左右）的高通滤器，阻止负载变化对接收波形的影响，而让300～9600Hz 的接收信号顺利通过。 这一设计的另一特点是，提高稳压电源VCC 的电压和功率（有效电流）就可以很容易的增加M-BUS 接口的驱动能力，而电路的其它器件无需做任何改动。 其实，这个设计和原始的电话机电路非常接近，只是电话机收发的是模拟型号，而M-BUS 接口收发的是数字信号。 图4的电路可直接放入Multisim 仿真。注意收发不要同时进行，实际使用也是这样的。 图3