基于IPSO优化RBF神经网络的带钢厚度控制预测新方法

A new method of the steel strip thickness control forecasting based on optimized RBF neural network using IPSO algorithm

提出了一种基于IPSO算法优化RBF神经网络的带钢厚度控制预测的新方法.先将PSO算法中的权重和学习因子进行优化,再将改进后的新粒子群算法应用于RBF神经网络的参数确定中,实现了RBF神经网络隐含层高斯函数的中心值和宽度向量以及隐含层与输出层之间权值的优化,改善了RBF神经网络的预测精度.仿真结果表明,将优化的RBF网络用于带钢厚度控制预测中,获得了可靠的精度和较好的收敛速度,说明该方法具有广阔的应用推广前景.

A new method of the thickness control forecasting of steel strip was put forward,which was based on the optimized RBF neural network using IPSO algorithm. First,the weight functions and the learning factors of PSO algorithm were optimized. Then,the improved IPSO algorithm was applied to optimize the positions of data centers,width vectors and weight functions of RBF neural network. The prediction precision of RBF neural network was improved. The simulation results show that the improved RBF neural networks applied in the thickness control forecasting of the steel strip are qualified with reliable accuracy and better convergence rate. This method possesses great practical value and future prospects.