基于遗传算法的爆震传感器诊断标定优化

Optimization of diagnosis calibration of knock sensor based on genetic algorithm

为提升发动机爆震传感器性能诊断的可靠性,建立爆震传感器性能诊断结果判定的概率计算模型,以每100万次诊断中误诊断的次数为目标函数,采用遗传算法对标定参数进行优化。应用该方法对某发动机爆震传感器性能诊断标定进行优化,结果表明,遗传算法在标定中可以获得较好的效果,可为其他标定研究提供借鉴方法和参考。

In order to optimize the diagnosis reliability of engine knock sensor,a probabilistic calculation model of knock sensor performance diagnosis is established.The optimization based on the genetic algorithm used the number of misdiagnosis per million times of diagnosis as the objective function.This method applied to a engine knock sensor diagnostic calibration.The results show that this method performs quite well,which provides a valuable method for calibration development.