Research On The Typical Electroniccircuit Fault Diagnosis Technology Based On The Multidimensional Characteristics

With the rapid development of system integration, the functions and performances of electronic system are further improved in the key areas of national security and military forces. However, such achievements are accompanied by several unfavorable problems, such as the lack of accessible detection points, the difficulties in gaining circuit information and exercising test excitation and the low fault detection rate, etc. Therefore, further researches into the diagnosis technology of typical electronic circuits are necessary and meaningful. By the help of as few test points as possible, we hereby tried to extract more characteristic fault information, improved the reliability of the electronic system and managed to detect the system faults timely and accurately.In this paper, based on the analysis of structural characteristics of circuit, research was carried out into the fault diagnosis technology by extracting and integrating the multidimensional fault characteristics of the time domain and frequency domain of the circuits. And typical circuit fault diagnosis system is designed to validate the effectiveness of the fault diagnosis algorithm based on multidimensional characteristics. The main contents are as follows:Firstly, research into the fault diagnosis algorithm based on multidimensional characteristics was carried out. The multidimensional information extracted from a single point of the feature circuit helps to resolve the problems of shortage of accessible test points and insufficient information of the integrated circuit extracted from single voltage signal; the correlation of multidimensional characteristics was analyzed, and then the principal components of characteristic signal features were extracted to simplify the algorithm complexity; tolerance analysis and margin setting of the circuit were carried out and therefore the influences on the circuit characteristics, from both the component tolerance and the external environment conditions, can be identified and excluded.Secondly, the principles and fault features of typical electronic circuits were analyzed. According to the characteristics of circuit output response, the paper extracted and optimized the fault features by applying the fault diagnosis algorithm based on multidimensional characteristics. The Fau1 lt Mode-Characteristics Reliance Matrix was established and then the actual fault diagnosis time was also shortened. Actual applications of the algorithm mentioned above in calculating the diagnosis indexes proved that they are practical and effective.Thirdly, the design of typical electronic circuit fault diagnosis system was implemented and completed. The upper computer based on WPF and the lower computer based on MCU and FPGA place the system completely under control and lead to a visualized and automatized fault diagnosis process. The functions of subsidiary systems and modules are properly allocated and tested, thus guarantee the completeness and reliability of the system.Finally, a typical application system of electronic circuit was successfully established. By the help of hierarchical and measurable modeling methods, we evaluated the design for testability and completed the physical verification of the system’s fault diagnosis functions on typical electronic circuits. The results show that the fault diagnosis algorithms based on multidimensional characteristics brought about in this paper can achieve fast and effective circuit fault diagnosis functions.