基于PSO-ELM的尾矿坝变形预测

Deformation Prediction of Tailing Dam Based on PSO-ELM

鉴于尾矿坝的特殊建造方式、随机荷载及强非线性增大变形预测难度的问题,提出利用极限学习机(ELM)映射尾矿坝变形与影响因素之间的复杂非线性关系。针对ELM算法随机产生输入与隐含层之间的连接权值和阈值的缺点,利用粒子群算法的优化能力,确定最优的权值和阈值,结合影响尾矿坝变形的因素,建立极限学习机和粒子群算法(PSO)相结合的PSO-ELM尾矿坝变形预测模型,并以风水沟尾矿坝实测资料为例,通过与PSO-SVM位移预测模型作对比,验证了PSO-ELM模型预测精度。结果表明,与PSO-SVM模型的预测结果相比,PSO-ELM模型的预测值更接近实际观测值,预测精度更高。

Aiming at the problem that the special construction method of tailings dam,random load and strong nonlinearity make deformation prediction difficult,an extreme learning machine(ELM)is proposed to map the complex nonlinear relationship between tailings dam deformation and influencing factors.In view of the shortcomings of the ELM algorithm in randomly determining the connection weights and thresholds between the input and the hidden layer,the optimal weights and thresholds were determined by using the optimization ability of the particle swarm algorithm.Combined with the factors affecting the deformation of the tailings dam,the PSO-ELM model for tailings dam deformation prediction was established,which is a combination of ELM and PSO.Taking the actual measurement data of Fengshuigou tailings dam as an example,the prediction accuracy of the PSO-ELM model was verified by comparing with the PSO-SVM displacement prediction model.The results show that the predicted values by the PSO-ELM model are closer to the actual observations and the prediction accuracy is higher than that of the PSO-SVM model.