论文部分内容阅读
为准确预测矿井煤与瓦斯突出的危险性,针对反向BP神经网络收敛差的缺点,分别采用基于MATLAB神经网络工具箱中的VLBP和LMBP算法的改进BP神经网络模型对煤与瓦斯突出的危险性进行了预测.根据煤与瓦斯突出的特点,选取开采深度、瓦斯压力、瓦斯放散初速度、煤的坚固性系数与地质破坏程度等五个关键影响因素作为煤与瓦斯突出的评判指标,建立了煤与瓦斯突出预测的神经网络模型.实际应用效果表明,采用基于MATLAB神经网络工具箱的BP网络模型,能克服一般BP网络收敛较慢的缺点,能加快收敛速度;运用LMBP算法比VLBP算法快,但需较大计算机内存;与常规预测方法相比较,该模型的预测准确性高,能有效地预测煤与瓦斯突出的危险性.
In order to accurately predict the danger of coal and gas outburst in coal mine, aiming at the shortcomings of reverse BP neural network with poor convergence, this paper respectively adopts the improved BP neural network model based on VLBP and LMBP algorithm in MATLAB neural network toolbox, According to the characteristics of coal and gas outburst, five key influencing factors, such as mining depth, gas pressure, initial gas velocity, coal solidity coefficient and degree of geological damage, are selected as the evaluation indexes of coal and gas outburst and established The neural network model of predicting coal and gas outburst was put forward.The practical application shows that using the BP network model based on the MATLAB neural network toolbox can overcome the disadvantage of the general BP network converging slowly and can accelerate the convergence speed.Using the LMBP algorithm than the VLBP algorithm Fast, but requires a large amount of computer memory. Compared with the conventional prediction method, the model has high prediction accuracy and can effectively predict the danger of coal and gas outburst.