论文部分内容阅读
针对煤与瓦斯突出灾害中瓦斯涌出量的辨识预测问题,结合采煤工作面瓦斯涌出量系统的现场实际特点,提出了混沌免疫遗传优化算法(CIGOA)与Elman神经网络相结合的耦合算法(CIGOA-ENN)。利用GIGOA的全局寻优能力替代梯度下降法,以克服Elman神经网络固有的缺陷。并根据输入的数据,构造基于CIGOA和ENN耦合算法的瓦斯涌出量系统辨识预测模型。利用矿区采集的现场监测数据进行仿真预测,实验表明该预测模型与BPNN,GA-ENN等神经网络预测模型相比,其收敛速度更快、收敛精度更高、鲁棒性更强,为解决煤矿瓦斯涌出量的预测问题提供了一个行之有效的方法。
Aimed at the problem of gas emission in coal and gas outburst disasters, combined with the actual characteristics of gas emission system in coalface, the coupling algorithm of chaos immune genetic algorithm (CIGOA) and Elman neural network (CIGOA-ENN). Global optimization ability of GIGOA is used instead of gradient descent method to overcome the inherent defects of Elman neural network. Based on the input data, a system identification prediction model of gas emission based on CIGOA and ENN coupling algorithm is constructed. Experiments show that the prediction model is faster than the neural network prediction models such as BPNN and GA-ENN, the convergence speed is faster, the convergence accuracy is higher and the robustness is better. In order to solve the problem of coal mine Prediction of gas emission provides a proven method.