Research On Intelligent Method Of Analog Circuit Fault Diagnosis Based On Cloud Model Optimization

The continuous development of integrated circuit technology has made the circuits in the system continue to develop on a large scale.While increasing the complexity of the circuit system,it has also brought a series of problems in circuit testing.According to research,the test cost of circuit systems now far exceeds the development cost of circuit systems.Although the proportion of analog circuits in circuit systems is only 20% of the entire circuit,the failure rate accounts for 80%.Moreover,due to the tolerance and nonlinearity of the analog circuit components,the fault test technology of the analog circuit develops slowly compared with the digital circuit.With the improvement of circuit integration,the traditional analog circuit fault diagnosis technology is more and more unable to meet the needs of modern circuit testing,and it is urgent to research new analog circuit testing technology.In recent years,due to the rise of artificial intelligence algorithms,most scholars have introduced artificial intelligence algorithms into fault diagnosis of analog circuits,thus bringing new ideas to the development of analog circuit fault diagnosis technology.In this paper,the artificial intelligence algorithm is used as the test method,and two international benchmark circuits,Sallen?key and CTSV,are used as verification objects.Based on the cloud model,the test methods related to analog circuits are studied from the aspects of circuit feature extraction and circuit fault diagnosis.The work and main contents involved in this article are as follows:(1)Research on feature extraction of analog circuits.Feature extraction is a key technology for fault diagnosis of circuits.The quality of feature extraction directly affects the fault diagnosis results of analog circuits.Aiming at the nonlinearity and non-stationarity of analog circuits,and the ambiguity and randomness of circuit fault diagnosis,and through the study of local mean decomposition algorithm(LMD)and cloud model,a method of feature extraction of analog circuit combining Local Mean Decomposition algorithm and cloud model is proposed.This method firstly decomposes the original signal of the analog circuit by the LMD algorithm,and then uses the inverse cloud generator to extract the three digital eigenvalues of the selected decomposed signal.When using this method to extract cloud digital feature values,not only the feature dimensions of the analog circuit are greatly compressed,but also some of the detailed information of the circuit is lost.Moreover,the traditional inverse cloud generator has large errors and weak stability when solving cloud digital features.Therefore,in order to make the cloud feature data better reflect the essential information of the analog circuit,as well as increase the stability of the reverse cloud generator and reduce the estimation error of solving the cloud feature data,another feature extraction method,phase space reconstruction(PSR)optimization multi-step reverse cloud algorithm,is proposed.During feature extraction,this method improves the solving process of three cloud digital feature values and optimizes them with PSR technology.Compared with the feature extraction method combined with LMD cloud model,the fault diagnosis rate of analog circuit is effectively improved.(2)Research on fault classification of analog circuits.In this paper,through the in-depth study of the Least Squares Support Vector Machine(LSSVM),it is found that the kernel function parameters and penalty factors of LSSVM are important factors that affect the classification effect.Because the Artificial Bee Colony algorithm(ABC)has a good global optimization capability,the ABC algorithm is used to optimize the parameters of LSSVM during fault diagnosis,and the ABC-LSSVM fault diagnosis model is successfully constructed.However,the standard ABC algorithm uses the method of random assignment to optimize the parameters of LSSVM,which will have a certain impact on the stability of abc-lssvm network.Therefore,this paper uses the reverse learning mechanism to improve the initialization of ABC algorithm.In addition,in order to increase the diversity of the ABC algorithm solution and enhance its exploration ability,the cloud model is used to improve the probability selection mechanism of the ABC algorithm.At the same time,in order to improve the optimization speed of the ABC algorithm,the optimal individual is used to guide the search method of the standard ABC algorithm.Therefore,the Optimal Guidance Reverse Cloud Model(GRC)is proposed to optimize the ABC algorithm,and the GRCABC-LSSVM analog circuit fault diagnosis model is constructed.Compared with ABC-LSSVM,GRCABC-LSSVM has a faster convergence speed,and it also has a better analog circuit fault diagnosis effect.