基于PSO-BP算法的神经网络模型预测策略研究

Study of Neural Network Model Prediction Strategies Based on PSO-BP Algorithm

在非线性模型预测中,往往难以获得精确的非线性数学模型,从而对预测精度造成一定的影响。该文将粒子群算法与BP算法相结合,提出了一种PSO-BP算法,改进了BP算法的不足,并将其应用于神经网络模型预测当中,提高了非线性模型预测的精度。

In nonlinear model in the prediction,often difficult to obtain precise nonlinear mathematical model,and the prediction accuracy cause certain effect.This paper will particle swarm algorithm and the BP algorithm is proposed,which combines a PSO-BP algorithm,improved BP algorithm is insufficient,and applied to the neural network model of forecasting,improve the prediction precision of the nonlinear model.