首页|基于改进SVM的电力工程造价预测

基于改进SVM的电力工程造价预测

扫码查看
针对支持向量机求解速度较慢且用于预测电力工程造价的性能不理想等问题,提出了一种基于改进SVM的电力工程造价预测模型.该模型全面考虑了电力工程成本的组成要素并进行参数归一化处理,利用最小二乘估计改进SVM模型,同时采用遗传算法求解LSSVM的参数最优值,并通过优化后的GA-LSSVM模型实现对电力工程成本的预测.基于MATLAB仿真平台的仿真实验结果表明,模型预测的工程成本与实际值较为接近,归一化均方误差与平均绝对百分比误差分别为18.34 万元和3.58%,且预测时间仅为256 ms,证明了其整体性能优于其他对比模型.
Power engineering cost prediction based on improved SVM
Aiming at the problems of lower solving speed of support vector machine(SVM)and unsatisfactory performance in predicting power engineering cost,a power engineering cost prediction model based on improved SVM was proposed.This model comprehensively considered the constituent elements of power engineering costs and the normalization of parameters.The least squares estimation was utilized to improve the SVM model,and genetic algorithm(GA)was used to solve for the optimal parameter values of LSSVM.The prediction for power engineering cost was conducted by the optimized GA-LSSVM model.The simulation experiment results based on MATLAB simulation platform show that the predicted engineering cost value is relatively close to the actual value.The normalized mean square error and average absolute percentage error are 183400 yuan and 3.58%,respectively,and the prediction time is 256 ms.The overall performance is better than other comparison models.

power engineeringcost predictionsupport vector machineleast squares estimationgenetic algorithmGA-LSSVM modelnormalization processingerror analysis

刘云、李维嘉、赵子豪、董振亮、陈志宾

展开 >

武汉大学电气与自动化学院,湖北武汉 430072

河北省电力有限公司经济技术研究院,河北石家庄 050001

河北省教育考试院 信息处,河北 石家庄 050091

河北赛克普泰计算机咨询服务有限公司,河北 石家庄 050081

展开 >

电力工程 造价预测 支持向量机 最小二乘估计 遗传算法 GA-LSSVM模型 归一化处理 误差分析

河北省自然科学基金项目国网河北省电力有限公司科技项目

F20212100055204JY22000L

2024

沈阳工业大学学报
沈阳工业大学

沈阳工业大学学报

CSTPCD北大核心
影响因子:0.62
ISSN:1000-1646
年,卷(期):2024.46(4)
  • 12