摘要
为了准确预测采煤工作面的瓦斯浓度,提出免疫遗传算法优化的动态模糊神经网络瓦斯浓度动态预测方法。用无线传感网络系统采集工作面瓦斯浓度数据作为样本,通过小波分析对样本数据进行降噪滤波预处理。采用IGA算法对DFNN网络参数进行优化,建立了瓦斯浓度的预测模型。通过MATLAB仿真研究表明,所建模型对采煤工作面的瓦斯浓度演变趋势预测合理,并且经过IGA算法优化DFNN网络比单纯的DFNN网络具有更快、更准确的预测功能,可以为防治煤矿瓦斯积聚提供更好的理论支持。
A method with Dynamic Fuzzy Neural Network Optimized by Immune Genetic Algorithm ( IGA-DFNN ) was proposed for predicting gas concentration in order to predict coal face gas concentration accurately. The simple of gas concentration is collected by wireless sensor networks,and is filtered and denoising through wavelet analysis before predicting. Then the prediction model sets up which use IGA algorithm to optimize the parameters of the DFNN network. The simulation of the MATLAB shows that the DFNN network optimized by IGA algorithm has faster and more accurate effect in predicting the gas concentration than simple DFNN network. It can provide better theoretical support for the prevention and control of gas accumulation in coal mine.
出处
《传感技术学报》
CAS
CSCD
北大核心
2014年第2期262-266,共5页
Chinese Journal of Sensors and Actuators
基金
国家自然科学基金项目(51274118
70971059)
辽宁省教育厅基金项目(L2012119)
辽宁省科技攻关项目(2011229011)