PSO-BP网络算法在运动要素解算中的应用

为了改善被测目标运动要素计算精度,提出了采用 PSO-BP 神经网络算法作为运动要素解算的方程。 该算法将粒子群算法作为 BP 神经网络的学习算法,提高 BP 网络的全局收敛性和收敛速度,将观测到的运动目标参数作为 PSO-BP 神经网络的输入,并将运动目标的方位作为主要输出量,将运动目标的方位值与误差期望值进行比较并作为 PSO 的输入修改 BP 网络权值,进而得到高精度 BP 神经网络。 对该算法进行仿真计算,结果表明:基于该算法的运动目标运动要素解算,尤其是运动方位的解算器精度可以达到 0. 128°,提高了运动要素解算的精度和速度。

Application of PSO-BP Network Algorithm in Solving Motion Elements

In order to improve the calculation precision of the motion elements of the measured target, a PSO-BP neural net- work algorithm is proposed to solve motion elements. The algorithm uses the particle swarm optimization algorithm as the learning algorithm of the BP neural network to improve the global convergence and convergence speed of the BP network, And the observed moving target parameters are used as the input of PSO-BP neural network. The orientation of the moving target is taken as the main output, the azimuth value of the moving target is compared with the expected value of the error, and the weight of the BP network is modified as the input of the PSO, and then the high precision BP neural network is obtained. Through the simulation calculation of the algorithm, the results show that the precision and speed of the motion elements calculation can be improved, especially the moving azimuth can be improved by 0. 128°.