基于改进PSO-BP神经网络的回弹预测研究

Study on springback prediction based on improved PSO-BP neural network

在V形自由折弯中,准确地预测板料回弹,有利于实际生产中精确地控制回弹以提高生产效率。由于板料回弹的影响因素众多,呈现出复杂的非线性变化特征,采用传统的BP神经网络难以满足高精度的预测要求,因此为了进一步有效预测板料的回弹,提出基于改进粒子群算法优化的BP神经网络预测模型。对标准粒子群算法的缺陷进行改进,利用改进粒子群算法的全局搜索能力对BP神经网络的权值和阈值进行优化求解,提高了BP神经网络预测模型的收敛精度和泛化能力。将改进PSO-BP神经网络预测模型应用在板料回弹预测中,并与LM-BP神经网络预测模型进行对比仿真,结果表明改进PSO-BP神经网络预测模型具有更高的非线性拟合优度和预测精度。

The accurate prediction of sheet metal springback in V-shape air bending is conducive to the accurate springback control in actual production,and improve the production efficiency.The springback of sheet metal is influenced by multiple factors,and appears as the complex nonlinear change characteristic.The traditional BP neural network is difficult to meet the high.precision forecasting requirements.Therefore,the prediction model based on improved particle swarm optimization algorithm optimizing BP neural network is proposed to further predict the springback of sheet metal effectively.The defect of standard particle swarm optimization algorithm is improved,and the global search ability of the improved particle swarm optimization algorithm is used to optimize and solve the weights and thresholds of the BP neural network,which can improve the convergence accuracy and generalization ability of the BP neural network prediction model.The improved PSO-BP neural network prediction model is used in sheet metal springback prediction,and compared with LM-BP neural network prediction model for simulation.The simulation results show that the improved PSO-BP neural network prediction model has higher nonlinear fitting goodness and prediction precision.