【摘要】 将空空导弹攻击区的拟合表示成一非线性逼近问题 ,从而利用模糊逻辑系统的万能逼近性质 ,并结合误差反向传播学习算法 (BP算法 )实现了对导弹攻击区的自动学习。在此基础上 ,提出了利用扩展的 T- S推理进行攻击区分组逼近与综合的方法。仿真结果表明该方法具有精度高、参数少、通用性强等优点。更多还原

【Abstract】 Considering the universal approximation theorem of fuzzy logic system, the fitting of air to air missile launch envelopes(MLEs), which is a nonlinear approximation problem in nature, is studied by using the method of fuzzy logic system. To make the system learn from automatically standard MLEs, the famous error Back Propagation algorithm (BP algorithm) is introduced. And because classifying the whole MLE spacemen into teams makes the fitting more accurate, Takagi and Sugeno′s fuzzy inference method (T S inference) is extended and used to classify/combine the MLEs. Numerical result shows that the method has many advantages of more accurate, less fitting parameters and more common in use over general methods.更多还原