基于主因子分析的改进BP神经网络瓦斯涌出量预测

Prediction of Gas Emission based on Principal Factor Analysis and Improved BP Neural Network

为提高回采工作面瓦斯涌出量预测效率和准确率,基于反向BP神经网络,采用主因子分析法对变量进行降维处理;结合遗传算法(GA)和附加动量法,采用遗传算法优化BP神经网络初始权值和阈值,建立基于主因子分析的GA—BP神经网络预测模型,并在权值反向更新过程中引入动量项。选取开滦矿业集团钱家营矿井瓦斯涌出量监测数据作为标签数据与输入数据,对不同网络模型进行了仿真与分析,结果表明:改进的GA—BP神经网络模型在603个时间步长里达到收敛,平均相对误差约为0.58%,预测精度和效率均优于其他神经网络模型,能更有效地实现瓦斯涌出量的准确预测。

In order to improve the prediction efficiency and accuracy of the gas emission in stope working face,based on reverse BP neural network,principal factor analysis was used to reduce the dimension of variable;combined genetic algorithm(GA)and additional momentum method,the initial weights and thresholds of BP neural network were optimized by genetic algorithm,the GA-BP neural network prediction model based on the principal factor analysis was established,and momentum term was introduced in the process of weight reverse update.The monitoring data of gas emission in Qianjiaying Coal Mine of Kailuan Mining Group was selected as tag data and input data,and the simulation and analysis were carried out for different network models.The results show that the improved GA-BP neural network model converged in603time steps,the average relative error is about0.58%,the prediction accuracy and efficiency are better than other neural network models,and the accurate prediction of gas emission can be realized more effectively.