基于Labview和Matlab的改进神经网络煤与瓦斯突出预测系统

Improved neural network for coal and gas outburst prediction system based on Labview and Matlab

为研究煤与瓦斯突出与其影响因素之间复杂的非线性关系并对其进行预测,基于共轭梯度法的改进神经网络,建立了煤与瓦斯突出预测模型。经过实际数据检验,此模型有很高的预测精度,避免了传统神经网络收敛速度慢的缺点。同时,为克服传统神经网络操作复杂和不便于现场使用的缺点,利用Labview图形化编程和数据采集能力以及Matlab强大的数值计算功能,通过ActiveX技术,实现了在Labview中调用神经网络工具箱,构建了一个煤与瓦斯突出预测系统,方便现场操作使用。最后指出使用Labview和Matlab混合开发可以极大缩短开发周期,为煤矿智能化系统的开发提供了一个新途径。

In order to study the complicated nonlinear relationship between coal and gas out-burst andits influence factors and to predict it,the coal and gas outburst prediction model was established by improved neural network based on conjugate gradient method. Through actual test data,this model has high prediction accuracy and avoids the shortcoming of slow convergence speed of traditional neural network. At the same time,in order to overcome complex operation and inconvenience for field use of traditional neural network,invoking the neural network toolbox in Labview was achieved via ActiveX technology by taking advantage of Labview graphical programming and data collection capacity as well as powerful numerical calculation function of Matlab. A coal and gas outburst prediction system was built,being convenient in operation on the scene. In summary,the combination of Labview and Matlab can shorten the development cycle, providing a new path for the development of intellectual system for coal mines.