GA-BP神经网络在钢铁加热炉建模中的应用

基于现场采集的大量的数据,采用BP神经网络建立现场加热炉炉温的非线性模型,并提出利用遗传算法优化BP神经网络的参数与阈值,有效避免了BP神经网络易陷入局部最小值、收敛速度慢的不足。仿真结果表明:在同样的数据集下,GA-BP神经网络的稳定性更好,预测精度更高。

Application of GA-BP Neural Network in Reheating Furnace Modeling

In this paper, basing on a large amount of data collected, the BP neural network was used to build a nonlinear model of the temperature of reheating furnace, including making use of genetic algorithm to opti-mize both parameters and thresholds of the BP neural network to effectively avoid the problem that BP neural network easily falls into local minimum or slow convergence speed.The simulation results show that, the GA-BP neural network is with more stability and higher prediction precision under the same data set.