论文部分内容阅读
为了对煤巷围岩稳定性进行科学、准确的预测,提出了一种将自适应差分进化算法(JADE)与极限学习(ELM)结合的煤巷围岩稳定性预测新方法。基于ELM训练速度快、泛化能力好和易获取全局最优解的优点,采用JADE优化ELM的输入权值矩阵和隐含层偏差,减少随机性造成的误差,建立JADE-ELM煤巷围岩稳定性预测模型。利用霍州煤矿区煤巷实测数据进行实例分析,并将预测结果与ELM、BP、RBF神经网络比较。结果显示:JADE-ELM模型预测平均精度为97.85%,比ELM、BP、RBF模型分别高出4.05%、17.85%、22.85%,JADE-ELM模型具有更高的预测精度,能够更准确的对煤巷围岩稳定性进行预测。
In order to predict the stability of surrounding rock in coal roadway scientifically and accurately, a new method to predict the surrounding rock stability of coal roadway based on adaptive differential evolution (JADE) and extreme learning (ELM) is proposed. Based on the advantages of ELM training speed, good generalization ability and easy access to the global optimal solution, the JADE-ELM input weight matrix and hidden layer deviation are optimized to reduce the error caused by randomness. Stability prediction model. The actual measurement data of coal roadway in Huozhou Coal Mining Area is used to carry out an example analysis, and the prediction results are compared with the ELM, BP and RBF neural networks. The results show that the average accuracy of JADE-ELM model is 97.85%, which is 4.05%, 17.85% and 22.85% higher than that of ELM, BP and RBF models, respectively. JADE-ELM model has higher prediction accuracy, Surrounding rock stability prediction.