基于粒子群优化BP神经网络的速度观测器

Research on speed observer of BP neural network based on particle swarm optimization

利用BP神经网络可以逼近任意的非线性关系的特点,构建一个神经网络速度观测器。由于BP神经网络的初始权值和阈值的选择存在随机性,对网络的性能影响很大,但又无法准确获得,提出将PSO算法与BP算法结合的方法,通过粒子群算法优化BP神经网络,得到最佳初始权值和阈值,提高了速度观测器的精度。通过实验采集的数据来训练这个神经网络,实验结果证明了该方法的有效性。

BP neural network characteristic was used to approximate any nonlinear relationship and build a neural network speed observer. Due to the randomness of choice of the initial weights and thresholds, BP neural network performance was affected greatly. In order to improve the speed of identify effect, the combination of the PSO algorithm and BP algorithm was proposed to obtain the best initial weights and thresholds. The experimental data was collected to train the neural network. The experimental results demonstrate the effectiveness of the method.