神经树在煤与瓦斯突出预测中应用研究  

Study and Application on the Methods for Predicting Coal and Gas Outburst Based on Neural Tree

作　　者：沈学利[1] 张红岩[1] 黄秀[2] 

机构地区：[1]辽宁工程技术大学电子与信息工程学院,葫芦岛125105 [2]济南大学信息科学与工程学院,济南250022

出　　处：《世界科技研究与发展》2010年第3期303-306,共4页World Sci-Tech R&D

摘　　要：由于煤与瓦斯突出影响因素之间存在着复杂的非线性关系,为准确预测煤与瓦斯突出的危险性,本文提出了基于柔性神经树的煤与瓦斯突出预测模型,其中利用多表达式编程和粒子群优化算法分别优化了自身的结构及相关参数,使得神经树具有强大的预测和分类能力,与传统神经网络相比具有更加灵活的自动优化能力。通过采用实测数据对算法进行了验证,结果表明与常规预测方法相比较,该模型的预测准确性高,具有良好的适应性和有效性。Because the complicated non - linear relation between the coal and gas outburst and its affecting factors,for the purpose of predicting the danger of coal and gas outburst in mine coal layer correctly. In this paper a prediction method of coal and gas outburst based on neural tree model, and using multi expression programming and particle swarm optimization algorithms are optimized its own structure and relevant parameters, making a powerful prediction neural tree and classification capabilities, which is constructed by a structured by a structured tree and a set of computation symbols, and which flexible can automatically devise an artificial neural network. Then the algorithm is validated through experimental data. The results showed that the model is a very efficient prediction method for mine coal and gas outburst, compared with the popular predicting methods, and the results indicates the practicability and efficiency of the new prediction model.

关 键 词：神经树 多表达式编程 粒子群优化算法 煤与瓦斯突出 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]