遗传BP网络在机载雷达故障诊断中的应用研究

摘要

This paper has a research on the application of GA- BP neural networks in fault diagnosis of a type of airborne radar. BP neural networks is likely to fall into the local minimum points which can be overcome by the global optimizing character of genetic algorithm.Take full use of the global optimization of genertic algorithm can improve the learning speed and precision of BPNN. The simulation with MATLAB further validates that GA-BPNN have the faster learning speed, higher prediction precision and better generalization performance than single BPNN, and GA-BPNN can be well applied to the fault diagnosis of electronic equipment such as radar.%对遗传算法优化的BP神经网络在某型机载雷达发射机中故障诊断应用进行了研究,目的是应用遗传算法的全局最优性解决BP神经网络容易陷入局部极小的问题,从而提高BP神经网络的学习速度和精度.通过MATLAB仿真,GA-BP神经网络在发射机故障诊断中网络训练收敛速度和误差精度都明显优于BP网络,进一步验证了遗传BP神经网络学习速度快、预测精度高、泛化效果好,很适合应用于雷达等电子设备的故障诊断.