Research On Fault Diagnosis Of Analog Circuit Modules Based On Cold Start SVDD Therory

With the continuous progress of science and technology,electronic technology has developed rapidly.Electronic systems play an increasingly important role in the fields of life,industry,medical treatment,etc.Therefore,the requirements for equipment reliability are gradually increasing,and there is a higher demand for the reliability detection technology of electronic systems.Many domestic and foreign scholars have carried out rich research on the topic of analog circuit fault diagnosis.The main idea is to use the sample of the system under test to train a diagnostic model,which can be used to detect and locate faults.However,in the actual application process,the number of negative samples in the tested electronic system is small and it is difficult to cover all the failure modes,resulting in a strong imbalance between positive and negative samples.This brings considerable difficulty for building accurate and effective diagnostic models.Therefore,in the case of circuit samples with strong imbalance,it is a difficult problem to detect and locate circuit faults in the field of analog circuit fault diagnosis.In order to solve the above problems,we deeply study the detection and location methods of analog circuit modular faults.This thesis proposes the fault diagnosis method of analog circuit modules based on Cold start Support Vector Data Description.First,divide the system under test into circuit modules with specific functions and structures according to the principle of module division of analog circuits.The feature selection method based on the Response Similarity Value is used to determine the type and parameters of the feature extraction method.Positive samples and a small number of negative samples are used to train the fault detection model that can identify the state of circuit module with Cold start SVDD.In order to realize the positioning of circuit fault modules,a method for location of analog circuit modular fault based on graph theory is proposed.The algorithm uses the fault detection results of the Cold start SVDD model to determine the location of the circuit modular fault to complete the fault diagnosis of the analog circuit system.In order to verify the effectiveness of this method,this article takes Sallen-Key circuit and Four-opamp circuit,dual bandpass filter circuit as examples to verify the effectiveness of the circuit module fault detection method.Finally,the method proposed in this paper is applied to the fault diagnosis of the Control System Simulation Platform and has achieved very good results.The results show that the method proposed in this paper has a good diagnostic effect on analog circuit modular faults.