针对传统智能诊断技术受限于先验知识、模拟电路故障多样性等不足，对基于否定选择算法的模拟电路故障诊断进行研究。分析人工免疫系统中的否定选择算法原理及应用，介绍实值否定选择算法的产生机制，提出与自体耐受和Monte Carlo相结合的优化算法，通过Fish’s Iris数据仿真显示，并将优化算法运用于电阻电路的8种软故障诊断。结果表明：优化算法的总体检测率达90%，能降低成熟检测器冗余，节省计算空间。%Research the fault diagnosis of simulation circuit based on negative selection algorithm to deal with traditional intelligent diagnosis technology shortages such as prior knowledge limit and simulation circuit fault variety and so on. Analyze principle and application of negative selection algorithm in artificial immune system. Introduce the mechanism of the real-valued negative selection algorithm, proposes optimized algorithm which combines Monte Carlo with self tolerance. By means of Fish’s Iris simulation result, use optimized algorithm in the redundancy of detections and saves space of computer. Use optimized algorithm in the fault diagnosis of resistance circuit which has eight soft fault kinds. The results show that the total detection rate of optimized algorithm reaches 90%, it can reduce redundancy of mature detector, and save computation space.
展开▼
