[1]王锋华,成敬周,文凡.快速双非凸回归算法及其电力数据预测应用[J].智能系统学报,2018,13(04):665-672.[doi:10.11992/tis.201708033]
 WANG Fenghua,CHENG Jingzhou,WEN Fan.Fast double nonconvex regression algorithm for forecast of electric power data[J].CAAI Transactions on Intelligent Systems,2018,13(04):665-672.[doi:10.11992/tis.201708033]
点击复制

快速双非凸回归算法及其电力数据预测应用(/HTML)
分享到:

《智能系统学报》[ISSN:1673-4785/CN:23-1538/TP]

卷:
第13卷
期数:
2018年04期
页码:
665-672
栏目:
出版日期:
2018-07-05

文章信息/Info

Title:
Fast double nonconvex regression algorithm for forecast of electric power data
作者:
王锋华1 成敬周1 文凡2
1. 国网浙江省电力公司, 浙江 杭州 310000;
2. 国网浙江省电力公司 经济技术研究院, 浙江 杭州 310000
Author(s):
WANG Fenghua1 CHENG Jingzhou1 WEN Fan2
1. State Grid Zhejiang Electric Power Company, Hangzhou 310000, China;
2. Economic Research Institute, State Grid Zhejiang Electric Power Company, Hangzhou 310000, China
关键词:
交替方向乘子法电力数据预测lp范数约束迭代阈值方法
Keywords:
alternating direction method of multiplier (ADMM)forecast of electric power datalp norm constraintiterative threshold method
分类号:
TP18;TM715
DOI:
10.11992/tis.201708033
摘要:
为适应产能输出、运营效益等电力数据预测应用,文中提出一种快速双非凸回归(double nonconvex regression,DNR)预测算法。首先,将经典稀疏编码分类技术解释为预测回归模型,并划分为训练阶段和测试阶段,使之适合标量预测应用;其次,针对经典Lasso模型存在的稀疏性不足以及噪声拟合单一问题,该算法通过lp范数约束逼近原始稀疏编码问题的误差重构项和系数正则项,具有更为灵活的模型形式和应用范围。最后,通过交替方向乘子框架实现了重构系数的优化升级策略。为确保ADMM优化子问题具有快速解,提出一种改进的迭代阈值规则用于更新非凸lp约束项,解决了原始算法陷入的局部最优问题。在电力企业实际运行产出和运营指标数据上的实验结果表明,DNR在预测效果和预测效率上均优于经典的支持向量机、BP神经网络以及非凸约束预测方法。
Abstract:
In this paper, we propose a new forecasting algorithm called double nonconvex regression (DNR) for the fast forecast of electricity power data such as the outputs of production ability and operational benefit. First, we reinterpret the typical sparse coding classification method as a regression model for forecasting, and further divide the model into training and testing phases to fit scalar-quantity forecasts. Next, we transform the constraints of representation residuals and coefficient regularization into a nonconvex lp norm for better approximation and broader application. Lastly, we adopt the alternating direction method of multipliers algorithm to optimize the formulated forecast problem. To achieve a fast update rule for lp norm constrained subproblems, we propose a new iterative threshold method that avoids the local minimum issue. Compared with typical methods such as the SVM, BP neural network, and nonconvex regularization methods, the proposed algorithm achieves surprisingly good experimental results for electricity power data.

参考文献/References:

[1] TAYLOR J W. Multi-item sales forecasting with total and split exponential smoothing[J]. Journal of the operational research society, 2011, 62(3):555-563.
[2] 彭敏, 张泰玮, 黄佳佳, 等. 基于回归模型与谱聚类的微博突发话题检测方法[J]. 计算机工程, 2015, 41(12):176-181. PENG Min, ZHANG Taiwei, HUANG Jiajia, et al. Microblog sudden topic detection method based on regression models and spectral clustering[J]. Computer engineering, 2015, 41(12):176-181.
[3] ARYA F K, ZHANG Lan. Time series analysis of water quality parameters at Stillaguamish River using order series method[J]. Stochastic environmental research and risk assessment, 2015, 29(1):227-239.
[4] 丁宏飞, 李演洪, 刘博, 等. 基于BP神经网络与SVM的快速路行程时间组合预测研究[J]. 计算机应用研究, 2016, 33(10):2929-2932, 2936. DING Hongfei, LI Yanhong, LIU Bo, et al. Expressway’s travel time prediction based on combined BP neural network and support vector machine approach[J]. Application research of computers, 2016, 33(10):2929-2932, 2936.
[5] ZHENG Jianwei, YANG Ping, CHEN Shengyong, et al. Iterative re-constrained group sparse face recognition with adaptive weights learning[J]. IEEE transactions on image processing, 2017, 26(5):2408-2423.
[6] CHEN Liang, SUN Defeng, TOH K C. A note on the convergence of ADMM for linearly constrained convex optimization problems[J]. Computational optimization and applications, 2017, 66(2):327-343.
[7] CUI Zhuoxu, FAN Qibin. A nonconvex nonsmooth regularization method for compressed sensing and low rank matrix completion[J]. Digital signal processing, 2017, 62:101-111.
[8] YANG Meng, ZHANG Lei, YANG Jian, et al. Regularized robust coding for face recognition[J]. IEEE transactions on image processing, 2013, 22(5):1753-1766.
[9] WRIGHT J, YANG A Y, GANESH A, et al. Robust face recognition via sparse representation[J]. IEEE transactions on pattern analysis and machine intelligence, 2009, 31(2):210-227.
[10] LUO Lei, YANG Jian, QIAN Jianjun, et al. Robust image regression based on the extended matrix variate power exponential distribution of dependent noise[J]. IEEE transactions on neural networks and learning systems, 2017, 28(9):2168-2182.
[11] 郑建炜, 黄琼芳, 杨平, 等. 特征加权组稀疏判别投影分析算法[J]. 自动化学报, 2016, 42(5):746-759. ZHENG Jianwei, HUANG Qiongfang, YANG Ping, et al. Feature weighted group sparse discriminative projection algorithm[J]. Acta automatica sinica, 2016, 42(5):746-759.
[12] CANDÈS E J, WAKIN M B, BOYD S P. Enhancing sparsity by reweighted l1 minimization[J]. Journal of Fourier analysis and applications, 2008, 14(5):877-905.
[13] SHE Yiyuan. Thresholding-based iterative selection procedures for model selection and shrinkage[J]. Electronic journal of statistics, 2009, 3:384-415.
[14] CAI Jianfeng, CANDÈS E J, SHEN Zouwei. A singular value thresholding algorithm for matrix completion[J]. SIAM journal on optimization, 2010, 20(4):1956-1982.
[15] CHAÂBANE N. A novel auto-regressive fractionally integrated moving average-least-squares support vector machine model for electricity spot prices prediction[J]. Journal of applied statistics, 2014, 41(3):635-651.
[16] LIU Ke, GUO Wenyan, SHEN Xiaoliu, et al. Research on the forecast model of electricity power industry loan based on GA-BP neural network[J]. Energy procedia, 2012, 14:1918-1924.
[17] TÜFEKCI P. Prediction of full load electrical power output of a base load operated combined cycle power plant using machine learning methods[J]. International journal of electrical power & energy systems, 2014, 60:126-140.
[18] ZHANG Yong, YE Wanzhou, ZHANG Jianjun. Sparse signal recovery by accelerated lq (0<q<1) thresholding algorithm[J]. International journal of computer mathematics, 2017, 94(12):2481-2491.

备注/Memo

备注/Memo:
收稿日期:2017-08-31。
基金项目:国家电网浙江省电力公司科技项目(5211JY15001V);国家电网公司科技项目(5211011600RJ).
作者简介:王锋华,男,1977年生,硕士研究生,主要研究方向为电网数据融合和处理分析技术。牵头国家电网公司科技项目5项。出版专著2部;成敬周,男,1980年生,博士研究生,主要研究方向为电力系统交直流动态系统、电网数据挖掘与分析应用技术。参与国家电网公司科技项目4项,发表学术论文10余篇;文凡,男,1982年生,硕士研究生,主要研究方向为电力系统自动化、电网数据分析技术。参与国家电网公司科技项目5项,发表学术论文10余篇。
通讯作者:王峰华.E-mail:wangfenghua0627@126.com.
更新日期/Last Update: 2018-08-25