摘要
利用BP神经网络算法对电子对抗干扰效果进行评估,根据干扰效果评定的诸因素的隶属函数确定其隶属度,以此作为神经网络输入层数据。然后,根据选定的训练样本对所构造的三层BP神经网络进行训练,再运用几组典型干扰对抗效果实验数据对该BP神经网络性能进行检验。结果表明BP网络已经训练好,用此BP神经网络评估电子对抗干扰效果是合理的,该方法减少了评估中的人为因素,使得评估结果更为可信。
The BP neural network is used in the evaluation of jamming effectiveness for electronic warfare, and the degrees of membership for jamming effectiveness' factors are determined according to the factors' subjection functions, so the degrees can be regarded as input data for BP neural network. Then training samples are chosen to train the BP neural network. Some typically experimental data for jamming effectiveness arc used to test the performance of the BP neural network. The result shows this BP neural network has been trained well, and it can be utilized to evaluate the jamming effectiveness for electronic warfare. The method limits artificial factors during the evaluation, makes the solution more reasonable and creditable.
出处
《电光与控制》
北大核心
2007年第1期69-71,75,共4页
Electronics Optics & Control
关键词
电子对抗
干扰效果
BP神经网络
electronic warfare
jamming effectiveness
Back Propagation neural network