协同进化PSO算法在瓦斯含量预测中的应用研究

Study on Application of Co-operative Particle Swarm Optimization Algorithm to Forecasting In-situ Gas Content

针对影响瓦斯含量的各种因素之间的复杂非线性关系,提出了利用协同进化粒子群优化(HCPSO)算法优化带开关权值的神经网络,来实现煤层瓦斯含量预测。通过使用二进制值0和1来表示神经网络的节点间有无连接,并用二进制编码来调节神经网络的结构;同时使用协同进化粒子群优化(PSO)算法优化神经网络权值,来获得精度较高、结构精简的神经网络模型。实验结果表明,该方法有效提高了瓦斯含量预测的精度,为煤矿瓦斯预测提供了一种新方法。

Because there is a complex nonlinear relationship among the various factors of influencing the gas content, the hybrid-coding co-operative particle swarm optimization algorithm （HCPSO） to optimize the neural network with switch is proposed to forecast the seam gas content. By using the binary values 0 and 1 to indicate whether or not the neural network between nodes connects, and using the binary code to adjust the neural network structure, at the same time, using the co-evolution particle swarm optimization algorithm to optimize the weight of the neural network, thus the neural network model with high accuracy and simple structure has been obtained. The experimental results show that the proposed method improves the prediction accuracy of gas content, and provide a new method for the coal mine gas forecasting.