梯度采样下的市场用电短期能耗优化预测仿真

Optimal Prediction and Simulation of Short-term Energy Consumption of Market Electricity under Gradient Sampling

下载PDF

导出

摘要为了提高市场用电短期能耗预测正确率,降低二次规划与用电能耗,提出梯度采样下的市场用电短期能耗优化预测方法。根据最小二乘支持向量机回归算法,将湿度、气温、气压、节假日变量作为输入,构建市场用电短期能耗预测模型,依据梯度采样序列二次规划方法优化用电能耗预测模型参数,进行了梯度采样序列二次规划,逐步优化求解LS-SVMR模型目标函数,完成市场用电短期能耗优化预测。实验结果表明:湿度、温度、气压、节假日因素对用电能耗产生影响,采样数量越大,优化性能越好,且能耗预测误差小。实现市场短期用电能耗的预测,预测准确度高,预测能力突出。 In order to improve the accuracy of short-term energy consumption prediction in the market and reduce the secondary planning and energy consumption,an optimization prediction method of short-term energy consumption in the market under gradient sampling is proposed.According to the least squares support vector machine regression algorithm,humidity,air temperature,air pressure and holiday variables are used as inputs to build a short-term energy consumption prediction model for market electricity.Based on the gradient sampling sequence quadratic programming method,the parameters of the electricity consumption prediction model are optimized.The gradient sampling sequence quadratic programming is carried out,and the LS-SVMR model objective function is gradually optimized to complete the short-term energy consumption optimization prediction for market electricity.The experimental results show that humidity,temperature,air pressure and holidays have an impact on power consumption.The larger the number of samples is,the better the optimization performance is,and the prediction error of energy consumption is small.Realize the prediction of short-term power consumption in the market,with high prediction accuracy and outstanding prediction ability.

作者高迪梁东王骏杨峰王艺霏陆鑫 GAO Di;LIANG Dong;WANG Jun;YANG Feng;WANG Yi-fei;LU Xin(State Grid Jibei Electric Power Company Limitied,Beijing 100052,China;State Grid Jibei Information&Telecommunication Company,Beijing 100053,China;State Grid Info-Telecom Great Power Science and Technology Co.,Ltd.,Fuzhou 350003,China)

机构地区国网冀北电力有限公司国网冀北电力有限公司信息通信分公司国网信通亿力科技有限责任公司

出处《节能技术》 CAS 2023年第1期68-72,共5页 Energy Conservation Technology

关键词梯度采样用电短期能耗优化预测最小二乘支持向量机回归算法预测模型二次规划 gradient sampling short term power consumption optimize prediction least squares support vector machine regression algorithm prediction model quadratic programming

分类号 TP391 [自动化与计算机技术—计算机应用技术]