摘要
矿井主通风机风量、风速等参数与瓦斯浓度及其它工况密切相关,参数复杂,建立其非线性数学模型比较困难,传统的辨识方法无法精确描述模型特性,文章将神经网络和模糊系统应用于矿井主通风机的模型辨识。神经网络辨识采用了一种基于径向基(RBF)的神经网络、模糊辨识采用了一种基于三角形隶属函数的T—S模糊模型。仿真结果表明,这两种方法可以同时满足对辨识精度、收敛速度、稳定性和跟踪能力的要求。
The air flow, air velocity and other parameters of the mine main ventilator are closed related to the gas density and other performances. Due to the parameter complicated, to set up a nonlinear mathematic model would be quite difficult. The conventional identification method could not accurately describe the model features. The neural network and fuzzy system were applied to the model identification of the mine main ventilator. The neural network base on RBF was adopted to the neural network identification and a T - S fuzzy model base on the delta membership function was adopted to the fuzzy identification. The simulation results showed that both two methods in the paper could all meet the requirements of the identification accuracy, convergence rate, stability and tracing ability.
出处
《煤炭工程》
北大核心
2008年第12期80-83,共4页
Coal Engineering
