基于人工免疫原理的系统异常检测与自适应容错控制

System Anomaly Detection and Adaptive Fault-tolerant Control Based on Artificial Immune Principle

介绍生物免疫系统的工作机理和人工免疫系统的发展与应用。基于免疫原理,定义控制系统的异常为抗原,与之对应的控制策略定义为抗体,将系统异常检测与诊断、控制策略匹配、自适应学习与记忆定义为免疫响应,这样,可将控制系统看成一个人工免疫系统。提出基于人工免疫原理的系统异常检测与自适应容错控制,并给出了控制系统总体结构。定义了表示系统异常特征的编码向量、特征空间;提出编码向量空间的划分方法,控制策略匹配算法与控制系统总体流程图。仿真实验和简单的工程应用验证了控制系统的可行性。

The operational principle of biological immune systems was described and the development and applications of artificial immune system were reviewed briefly. Based on immune principles, the control system anomalies were defined as antigens and the control policy for them were defined as antibodies. The anomaly detection and diagnosis, the control policy mating, adaptive learning and memory process were defined as immune response, so that control system would be regarded as a artificial immune system. System anomaly detection and adaptive fault-tolerant control based on artificial immune principles were proposed, and control system architecture was given. The coding vector of control system anomaly character and coding vector space were defined. A method divided the coding vector space, algorithm of control policies mating and a flowchart of the control system were given. The control system is verified practicability by simple engineering and simulations.