RBF神经网络在煤的热解特性研究中的应用

Application of RBF Neural Network to Pyrolytic Characteristics of Coal

在MATLAB5 .3环境下 ,利用神经网络工具箱中的RBF神经网络算法 ,通过煤的元素分析和工业分析对煤的热解特性进行了研究 ,对热解初析温度Ti、最大反应速度对应的温度T1max、热解前期热解量G1以及热解完成时间t进行了预报 .预测结果表明用RBF神经网络来研究煤的热解特性 ,可以达到较高精度 。

Under the condition of MATLAB5.3, a RBF neural network in the neural Networks Toolbox and essential analytic data: C ad , H ad , O ad , N ad , A ad , M ad are used to predict coal′s pyrolytic characteristics such as T i, T 1max , G 1 and t . The high precision of the prediction results shows that the RBF neural network is a useful method for the study of coal′s pyrolytic characteristics.