瓦斯突出模型预测控制的应用研究

Application Research of Model Predictive Control of Gas Outburst

煤体瓦斯涌出量的动态变化是一个复杂的非线性系统,传统的瓦斯监测方法准确率较低。针对该问题,文章提出了一种基于BP人工神经网络模型的瓦斯突出危险性预测控制方法。该方法运用BP人工神经网络预测模型对输入的多组样本进行训练学习、建立预测准则,并以此辨识瓦斯突出危险性类型。仿真结果表明,该方法有效解决了传统的瓦斯突出预测模型在事故预测中误差大、稳定性差的缺陷,提高了预测精度。

Dynamic change of coal gas emission is a complex nonlinear system and the traditional gas monitoring method has low accuracy. To solve the problem, the paper proposed a predictive control method for gas outburst hazard based on BP artificial neural network model. The method uses predictive model of BP artificial neural network to train and study multi-group input samples and build predictive criterion by which types of gas outburst hazard can be identified. The simulation result showed that the method can effectively solve the defects of large error and bad stability of the traditional gas outburst predictive model in accident prediction and can improve prediction precision.