基于误差反向传播(BP)神经网络的函数逼近特点,提出了一种运用BP神经网络在毫米波被动探测中计算目标辐射温度的方法。仿真结果表明,该方法对毫米波被动式辐射计目标辐射温度的计算具有较高的精度,能够较好的拟出目标辐射温度因子的信息曲线及误差曲线,对目标中心的识别具有较好的效果。该方法对提高制导精度有一定的重要意义。

A method to calculate the radiometric temperature of the objective with a passive millimeter wave seeker based on the function approximation of the back propagation neural networks was proposed. The result of simulation shows that the method is reasonable and precise; it can get a good result for the target's radiometric temperatures and plot the temperature factor and error-curve, so it has the effective to recognize the center of targets. The method has the important significance in precision guide weapon.