Analysis Of Analog Circuits Network Features And Test Points Selection Based On The Bipartite Network Model

The problems of analog fault testing and diagnosis are hot issues in the research ofelectronic systems and the test points selection is an important direction of these studies.With the increasing complexity and intergration of circuit systems, the test pointsselection becomes more and more complicated. So far, there are a few research inanalyzing network properties and the test points selection on account of the complexnetwork theory. Therefore, in this paper, it focuses on the network characteristics andmeasuring point selection problem of analog circuit systems based on the theory ofcomplex networks with the bipartite network model. The main research results are asfollows:1The complex network features are studied for analog circuits based on thebipartite network model. Advantages for analog circuit systems are analyzed anddisplayed based on the bipartite network model in this paper. The complex networkparameters and characteristics are studied by two kinds of the bipartite network modelmethods (one is the projection method and anther is the direct analysis method) foranalog circuits. And the research results by these two methods are compared. It isshown that the results of the two methods can display the complex networkcharacteristics of analog circuits comprehensively, and these complex networkcharacteristics of circuits have close relationships with the fault propagation and the testpoints selection.2The networks of measuring points for the analog circuits are further studied bythe projection method. First of all, the importance of test points is researched from theperspective of both topology and data. On the one hand, the importance of test points isanalyzed by the way combining with the bipartite network model, judgment methods ofthe network node centricity and fuzzy decision methods based on the topological nodecentricity. On the other hand, to study the importance of measuring points, the data oftest points by simulation with the Pspice software are researched by using Grangercausality connection analysis (GCCA) based on the data node centricity. Second, thealgorithm with the node centricities for testing points selection is proposed based on the importance of test points. Finally, the vulnerability of testing points networks for analogcircuit systems are analyzed based on the analysis of the node centricity. It is verified byactual analog circuits that test points selection results by the proposed aigorithm are inaccord with the predecessors’ research results, which verify the correctness of thisalgorithm proposed.3The software system with bipartite network model anlysis and test pointsselection for analog circuit systems is designed and implemented. By using the softwareof Matlab, the complex network parameters and network chanracteristics of test pointsnetworks and components networks for analog circuit systems can be analyzed by thisanalysis system. And the importance for the test points and the test points selection alsocan be commplished with this system.