摘要
根据重庆市南桐矿务局砚石台煤矿的生产技术条件和开采地质条件,针对传统方法预测冲击地压存在的弊端,运用BP人工神经网络和遗传算法相结合的方法,通过改进激励函数来缩短训练时间,并利用给权值加入动量项和变速率学习方法,减少学习中的振荡,来对该矿冲击地压预测进行研究。工程实际应用表明,该方法能有效的确定网络结构和训练参数,并可以很好地应用在相关工程上。
Based on the research of mining circumstances and geological conditions of rockburst in Yanshitai coal mine, the influence factors of rockburst are provided. Because of the shortage of traditional methods to forecast rockburst, the method based on the back propagation neural network and genetic algorithm is applied to predict rockburst through reduction training time and instability in network training in Yanshitai coal mine. The results show that it is an effective method to predict rockburst and it can be applied to similar engineerings.
出处
《岩土力学》
EI
CAS
CSCD
北大核心
2003年第6期1016-1020,共5页
Rock and Soil Mechanics
基金
教育部高等学校优秀青年教师教学科研奖励计划资助项目。