摘要
针对飞行控制系统的传感器故障诊断问题,构造基于Mexicohat小波的小波神经网络,采用改进的自适应遗传算法优化所构造的小波网络的网络参数,并应用训练好的小波网络对飞行控制系统的传感器故障进行诊断。通过在Matlab/Simu-link中建立飞控系统传感器的数字仿真模型并进行计算机仿真,得出所构造的小波网络能很好地诊断出飞控系统传感器的三类故障。仿真结果表明,用遗传算法训练小波网络,收敛速度快,且不会陷入局部最优点,训练好的小波神经网络的收敛性、故障诊断能力及泛化性均强于传统的BP神经网络。
In order to realize the fault diagnosis for sensors of flight control system, a wavelet neural network （WNN） based on Mexico hat wavelet was built, the improved genetic algorithm was adopted to optimize the parameters of the WNN being built. The simulation model was built in Matlab/Simulink, and the results of applying the WNN to fault diagnosis of sensor were given. The results show that the diagnosis system built based on WNN has better convergence and robustness than BP network.
出处
《计算机仿真》
CSCD
2008年第3期83-85,89,共4页
Computer Simulation
关键词
遗传算法
小波网络
飞控系统
故障诊断
Genetic algorithm
Wavelet neural network
Flight control system
Fault diagnosis
