受生物免疫系统的启发,把免疫原理应用到故障诊断领域。分析了现有阴性选择算法在故障诊断应用中的不足之处,提出了改进后的阴性选择算法。介绍了田纳西-伊斯曼(TEP)过程仿真系统,用改进后的阴性选择算法对该系统的21种故障样本进行仿真试验,各故障的平均检出率达到85%以上,验证了此算法在工业控制系统故障诊断中的有效性。
Inspired by artificial immune system, the immune principle is applied to the fault diagnosis field. The improved negative selection algorithm was expatiated, after analyzing the shortcoming of the existent negative selection algorithm. The simulated system for Tennessee Eastman Process was introduced and the 21 fault samples of this system were diagnosed using the improved negative selection algorithm. The correct rate of the diagnosis is above 85%. It is proved that the improved algorithm is valid to diagnose faults of industry control system.