鉴于预测边坡位移变化对边坡稳定性的重要意义,利用布谷鸟优化算法(COA)对最小二乘支持向量机(LSSVM)的核函数参数和惩罚因子进行寻优,从而建立了边坡位移时序预测的COA-LSSVM模型,并将该模型应用于锦屏一级水电站左岸高边坡变形预测中。与PSO-LSSVM模型的预测结果对比表明,COA-LSSVM模型具有更高的预测精度,预测结果更接近于实际的监测数据。 In view of the importance of prediction of slope stability on slope stability, the kernel function parameters and penalty factors of least square support vector machine (LSSVM) are optimized by cuonas optimization algorithm (COA), and the slope displacement The COA-LSSVM model of time series prediction is applied to the prediction of the high side slope deformation of the left bank of Jinping Hydropower Station. Compared with the prediction results of PSO-LSSVM model, the COA-LSSVM model has higher prediction accuracy and the prediction results are closer to the actual monitoring data.