GA-BP网络与D-S证据相结合的多传感器信息融合

Multi-sensor information fusion based on integration of GA-BP network and D-S evidence theory

针对反向传播(back propagation,BP)网络与D-S(dempster-shafer)证据理论各自在处理不确定性信息方面的不足,提出了一种遗传算法(genetic algorithms,GA)优化的BP网络与D-S证据相结合的多传感器信息融合方法。一方面利用GA-BP网络获取D-S证据理论所需的基本概率赋值,另一方面通过D-S证据理论对GA-BP网络的输出进行融合。将此方法应用于高压电器设备故障诊断,仿真结果表明,该方法能克服传统BP网络易陷入局部最优问题,同时具有更好的识别结果。

As back propagation network and dempster-shafer evidence theory have individual defects in processing uncertain information,a method of multi-sensor information fusion is proposed based on the combination of genetic algorithms optimal BP neural network and D-S evidence theory.The basic probability assignment of D-S evidence theory can be obtained using GA-BP network.moreover,the results of GA-BP network output can be fused by D-S evidence theory.The method is applied to fault diagnosis of high-voltage electric equipment.Simulation shows that the new method can solve the locally optimal problem in single BP neural network training,and get a better recognition result.