Application Research On Fuzzy Immune Algorithm To Fault Diagnosis Of Analog Circuit

Higher requirements have been put forward to circuit fault diagnosis for the extensive application of electronic devices. Because most of electronic devices failures are caused by analog circuits, analog circuit fault diagnosis has become the key of circuit theory research. However, the research and application of analog circuit fault diagnosis are extraordinarily difficult as its complexity.Biology immune system has powerful information processing capabilities and a lot of unique mechanism. Based on this, artificial immune theory is developed rapidly, and provides an effective way for fault diagnosis, and therefore, gets prevalent concern and attention. Simultaneously, artificial immune system provides a new choice for analog circuit fault diagnosis research, too.Biology immune principle is more in-depth studied in this paper. The developments of existing artificial immune technology and application are also summarized detailedly. And a new breakthrough is attempted to be made based on the results of existing artificial immune researches.The core content researched in this thesis is the study of an immune algorithm based on Fuzzy Mathematics. And particle filter algorithm is used to optimize and improve the algorithm. First, some basic knowledge is introduced, including interrelated biology immune mechanism, basic artificial immune theory, particle filter algorithm and analog circuit fault diagnosis technology. The research works combining artificial immune theory and particle swarm are also summarized. Then, an algorithm based on the humoral immune is put forward. The generating, training, learning and diagnosing processes of the algorithm are given exhaustively. And, particle filter algorithm is used to guide the mutating process of antibodies. The theoretical advantages of the algorithm are analyzed. At last, the feature data of an analog circuit working on different states are collected through the simulation experiments of the analog circuit. And the process of using the algorithm to diagnose analog circuit faults is expounded. At the same times, the feasibility and validity of the algorithm are validated.