论文部分内容阅读
利用改进的遗传算法对传统BP神经网络的隐含层节点个数和网络拓扑结构进行优化,提出利用遗传神经网络预测矿井工作环境中瓦斯体积分数的模型,并结合某矿瓦斯体积分数实例进行仿真试验研究,试验结果与实际比较吻合。
The improved genetic algorithm is used to optimize the number of hidden nodes and the network topology of the traditional BP neural network. The model of predicting the gas volume fraction in the working environment of the mine by using genetic neural network is proposed. The simulation is carried out based on the gas volume fraction of a certain mine Experimental study, the test results and the actual comparison.