基于免疫Agent的液体火箭发动机故障诊断

为解决液体火箭发动机系统非线性导致的故障诊断困难问题,将免疫Agent(IA)技术应用于液体火箭发动机的故障诊断研究.结合可变半径人工识别球(ARB)方法,在学习人类举一反三的机制,采用最大识别概率原则的基础上,构建了免疫Agent实体,并进行了仿真实验.实验结果表明:该方法训练时间短、正确率高、有较强的抗噪能力和自学习能力,可以应用于液体火箭发动机的状态监测与故障诊断.

Fault diagnosis of liquid rocket engine based on immune Agent

In order to solve the fault diagnosis difficulties of the Liquid Rocket Engine (IRE) caused by non-linear character, the technique of Immune Agent (IA) was applied to fault diagnosis of IRE. After combining the alterable-radius Artificial Recognition Ball (ARB) and introducing a recognition principle called maximum recognized probability learned from the mechanism of drawing inferences about other cases from one instance, an IA entity was created and applied in the simulation experiment. The results show that this method is efficient with short training time and high rate of correct fault diagnosis. The capability of noise tolerance and learning ability are both powerful. It can work well in the state monitoring and fault diagnosis of the LRE.