论文部分内容阅读
为了提高气化配煤煤灰流动温度预测的精度和稳定性,提出将遗传算法(GA)与误差反向传播神经网络(BP)相结合的预测方法,采用GA优化BP神经网路的权值和阈值,再用BP算法训练网络,结合仿真实验分析比较了GA-BP网络算法与常规BP神经网络方法的精度和稳定性。结果表明:GA-BP网络改善了BP网络容易陷入局部极小值和收敛速度慢的缺点,经GA优化的BP神经网络预测方法的预测精度高于BP网络算法,将其应用于气化配煤灰熔点预测有效可行。
In order to improve the accuracy and stability of coal-ash flow temperature prediction for gasification coal blending, a prediction method combining genetic algorithm (GA) with error backpropagation neural network (BP) is proposed, and GA is used to optimize the weight of BP neural network And threshold, and then use BP algorithm to train the network. The accuracy and stability of the GA-BP network method and the conventional BP neural network method are analyzed and compared with simulation experiments. The results show that the GA-BP network has the shortcoming of BP neural network being easy to fall into local minima and slow convergence speed. The GA BP neural network forecasting method has higher prediction accuracy than BP network algorithm and is applied to gasification blending coal Ash melting point prediction is feasible.