论文部分内容阅读
文章针对太阳电池阵热真空试验非线性的特点,引入一种基于神经网络的系统辨识方法。该方法采用BP网络作为模型辨识器,而辨识器又采用L-M(Levenberg-Marquart)算法进行训练。仿真结果表明,该方法具有较高的训练速度与精度,可以对太阳电池阵热真空试验测点温度响应做出较为精确的预测。
In this paper, aiming at the non-linearity of thermal vacuum test of solar array, a system identification method based on neural network is introduced. This method uses BP network as model identifier, and identifier uses L-M (Levenberg-Marquart) algorithm to train. The simulation results show that this method has higher training speed and accuracy, and can make a more accurate prediction of the temperature response of the test points of the solar array thermal vacuum test.