基于分数阶神经网络的瓦斯涌出量预测

针对现有的煤与瓦斯涌出危险性区域预测模型存在收敛速度慢、极易陷入局部极值等问题,结合BP的局部搜索能力和分数阶算法快速的全局搜索能力,提出了一种基于分数阶神经网络的新预测模型,用于非线性瓦斯涌出量的动态预测。经训练和实验结果表明:该模型较其他模型具有更好的滤波效果、更强的抗干扰能力、更快的收敛速度、更高的收敛精度等特点,能够达到准确指导实践的要求。

Prediction of gas emission quantity based on fractional order neural network

A new predicting model based on fractional neural network combining global search ability of BP with fast local search ability of fractional order algorithm is presented to dynamically predict nonlinear gas emission quantity,aiming at present coal mine problems such as slow convergence speeding,easily trapped into local minima,etc.Both of training and the experimental results show that this model has better filtering effect,stronger anti-interference ability,faster convergence rate and higher prediction accuracy than the other models,which can meet requirements for exact guidance to practice.