基于改进灰狼算法优化支持向量机的边坡位移预测  

Slope Displacement Prediction Based on Improved Grey Wolf Algorithm and Support Vector Machine

作　　者：刘晖 朱德康 许凌杰 闫少霞 LIU Hui;ZHU De-kang;XU Ling-jie;YAN Shao-xia(Fuchunjiang Hydropower Plant of State Grid Xinyuan Group Co.,Ltd.,Hangzhou 311000 China;Guangzhou South Satellite Navigation Instrument Co.,Ltd.,Guangzhou 510000 China)

机构地区：[1]国网新源集团有限公司富春江水力发电厂,浙江杭州311000 [2]广州南方卫星导航仪器有限公司,广东广州510000

出　　处：《自动化技术与应用》2024年第9期30-33,171,共5页Techniques of Automation and Applications

基　　金：广东省重点领域研发计划项目(232023021021900001);国网新源集团有限公司科技项目资助项目(SGXYKJ-2022-057)。

摘　　要：为了提高边坡位移预测精度,采用帐篷映射和位置收敛参数变化方式调整两种对改进了灰狼算法,得到改进灰狼算法。采用改进灰狼算法对支持向量机进行优化,建立基于改进灰狼算法优化支持向量机的边坡位移预测模型,采用实际水利工程边坡位移数据进行仿真分析,IGWO-SVM模型的平均相对百分误差和均方根误差分别为2.41%和0.21,预测效果优于PSO-BP模型,验证了该模型在边坡位移预测方面的实用性和优越性。In order to improve the prediction accuracy of slope displacement,the Grey Wolf Algorithm is improved by using tent mapping and position convergence parameter change adjustment.The Improved Grey Wolf Algorithm is used to optimize the support vector machine,and the slope displacement prediction model based on the Improved Grey Wolf Algorithm is established.The actual hydraulic engineering slope displacement data is used for simulation analysis.The average relative percentage error and root mean square error of the IGWO-SVM model are 2.41%and 0.21,respectively.The prediction effect is better than PSO-BP model,which verifies the practicability and superiority of this model in slope displacement prediction.

关 键 词：边坡位移 预测 改进灰狼算法 支持向量机 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]