基于改进猎人猎物算法的VMD-KELM短期负荷预测*

doi:10.11985/2023.04.025

摘要/Abstract

摘要：

为进一步提高负荷预测的准确性和可靠性，针对核极限学习机(Kernel extreme learning machine，KELM)参数选择影响预测能力的缺点以及负荷数据的波动性和非平稳性的特征，提出一种基于变分模态分解(Variational mode decomposition，VMD)与经种群混沌策略和莱维飞行策略改进的猎人猎物算法(Levy-hunter-prey optimizer，LHPO)优化KELM的预测模型。首先，使用灰色关联分析对原始数据的环境因素与负荷数据进行相关性分析；然后，使用VMD对负荷数据进行分解，分别将每个分解子序列输入经LHPO优化的KELM模型进行负荷预测；最终，将每个预测结果进行叠加。仿真试验验证该预测模型对短期负荷预测具有较高的适应性。

关键词: 短期负荷预测, 变分模态分解, 核极限学习机, 猎人猎物算法, 莱维飞行, 灰色关联模型

Abstract:

In order to further improve the accuracy and reliability of load forecasting, a forecasting model based on variational mode decomposition(VMD) and Levy-hunter-prey optimizer(LHPO) improved by population chaos strategy and Levy flight strategy is proposed to optimize kernel extreme learning machine(KELM), in view of the shortcomings of the KELM parameter selection that affects the forecasting ability and the characteristics of the volatility and non-stationary of load data. Firstly, the environmental factors and load data of the original data are analyzed by using the grey relational analysis. Then, VMD is used to decompose the load data, and each decomposition subsequence is input into the KELM model optimized by LHPO for short-term load forecasting. Finally, each prediction result is stacked. Simulation experiment results show that the prediction model has high adaptability to short-term load forecasting.

Key words: Short-term load forecasting, VMD, kernel extreme learning machine, hunter-prey optimizer, Levy flight, grey incidence model

中图分类号:

TM74

鲁英达, 张菁. 基于改进猎人猎物算法的VMD-KELM短期负荷预测^*[J]. 电气工程学报, 2023, 18(4): 228-238.

LU Yingda, ZHANG Jing. VMD-KELM Short-term Load Forecasting Based on Improved Hunter-prey Optimizer[J]. Journal of Electrical Engineering, 2023, 18(4): 228-238.

图/表 21

表1

基准测试函数"

函数表达式	取值范围	最优解
$f 1 = ∑ i = 1 n \| x i \| + ∏ i = 1 n x i$	[-10, 10]	0
$f 2 = ∑ i = 1 n [100 (x i + 1 − x i 2) 2 + (x i − 1) 2]$	[-30, 30]	0
$f 3 = 1 4 000 ∑ i = 1 n x i 2 − ∏ i = 1 n cos (x i i) + 1$	[-600, 600]	0

表1

表2

图1

图2

图3

图4

表3

表4

图5

表5

图6

表6

表7

图7

图8

图9

图10

表8

图11

图12

图13

参考文献 21

[1]	孙华东, 徐式蕴, 许涛, 等. 电力系统安全稳定性的定义与分类探析[J]. 中国电机工程学报, 2022, 42(21):13.
	SUN Huadong, XU Shiyun, XU Tao, et al. Analysison the definition and classification of power system security and stability[J]. Proceedings of the CSEE, 2022, 42(21):13.
[2]	李泽文, 胡让, 刘湘, 等. 基于PCA-DBILSTM的多因素短期负荷预测模型[J]. 电力系统及其自动化学报, 2020, 32(12):32-39.
	LI Zewen, HU Rang, LIU Xiang, et al. Multi-factorshort-term load forecasting model based on PCA-DBILSTM[J]. Proceedings of the CSU-EPSA, 2020, 32(12):32-39.
[3]	杨胡萍, 余阳, 汪超, 等. 基于VMD-CNN-BIGRU的电力系统短期负荷预测[J]. 中国电力, 2022, 55(10):71-76.
	YANG Huping, YU Yang, WANG Chao, et al. Short-term load forecasting of power system based on VMD-CNN-BIGRU[J]. Electric Power, 2022, 55(10):71-76.
[4]	黄炜, 陈田. 基于SAE与CEEMDAN-BiLSTM组合模型的短期电力负荷预测[J]. 计算机应用与软件, 2022, 39(7):52-58.
	HUANG Wei, CHEN Tian. Short term power load forecasting based on the combined model of SAE andCEEMDAN-BILSTM[J]. Computer Applications and Software, 2022, 39(7):52-58.
[5]	龚飘怡, 罗云峰, 方哲梅. 基于Attention-BiLSTM- LSTM神经网络的短期电力负荷预测方法[J]. 计算机应用, 2021, 41(S1):81-86.
	GONG Piaoyi, LUO Yunfeng, FANG Zhemei. Short term power load forecasting method based on Attention BiLSTM-LSTM neural network[J]. Journal of Computer Applications, 2021, 41(S1):81-86.
[6]	汤庆峰, 刘念, 张建华, 等. 基于EMD-KELM-EKF与参数优选的用户侧微电网短期负荷预测方法[J]. 电网技术, 2014, 38(10):2691-2699.
	TANG Qingfeng, LIU Nian, ZHANG Jianhua, et al. Short term load forecasting method of user side micro grid based on EMD-KELM-EKF and parameter optimization[J]. Power System Technology, 2014, 38(10):2691-2699.
[7]	孙铭阳, 谢子殿, 韩龙, 等. 自适应阈值函数小波算法的电机振动信号去噪[J]. 电子科技, 2020, 33(1):63-67.
	SUN Mingyang, XIE Zidian, HAN Long, et al. Denoising of motor vibration signal using adaptive threshold function wavelet algorithm[J]. Electronic Science and Technology, 2020, 33(1):63-67.
[8]	张淑清, 李君, 姜安琦, 等. 基于FPA-VMD和BiLSTM神经网络的新型两阶段短期电力负荷预测[J]. 电网技术, 2022, 46(8):3269-3279.
	ZHANG Shuqing, LI Jun, JIANG Anqi, et al. New two-stage short-term power load forecasting based on FPA-VMD and BiLSTM neural networks[J]. Power System Technology, 2022, 46(8):3269-3279.
[9]	杨锡运, 邢国通, 马雪, 等. 一种核极限学习机分位数回归模型及风电功率区间预测[J]. 太阳能学报, 2020, 41(11):300-306.
	YANG Xiyun, XING Guotong, MA Xue, et al. A quantile regression model of nuclear limit learning machine and wind power interval prediction[J]. Acta Energiae Solaris Sinica, 2020, 41(11):300-306.
[10]	吴松梅, 蒋建东, 燕跃豪, 等. 基于VMD-PSO-多核极限学习机的短期负荷预测[J]. 电力系统及其自动化学报, 2022, 34(5):18-25.
	WU Songmei, JIANG Jiandong, YAN Yuehao, et al. Short term load forecasting based on VMD-PSO multi-core limit learning machine[J]. Proceedings of the CSU-EPSA, 2022, 34(5):18-25.
[11]	宋玉生, 刘光宇, 朱凌, 等. 改进的灰狼优化算法在SVM参数优化中的应用[J]. 传感器与微系统, 2022, 41(9):151-155.
	SONG Yusheng, LIU Guangyu, ZHU Ling, et al. Application of improved grey wolf optimization algorithm in SVM parameter optimization[J]. Transducer and Microsystem Technologies, 2022, 41(9):151-155.
[12]	闫少强, 刘卫东, 杨萍, 等. 基于K-means聚类的多种群麻雀搜索算法[J/OL]. 北京航空航天大学学报:1-13[2022-10-18]. http://doi.org/10.13700/j.bh.1001-5965.2022.0328.
	YAN Shaoqiang, LIU Weidong, YANG Ping, et al. A multi population sparrow search algorithm based on K-means clustering[J/OL]. Journal of Beijing University of Aeronautics and Astronautics:1-13[2022-10-18]. http://doi.org/10.13700/j.bh.1001-5965.2022.0328.
[13]	NARUEI I, KEYNIA F, SABBAGH MOLAHOSSEINI A., Hunter-prey optimization:Algorithm and applications[J]. Soft Computing, 2022, 26(3):1279-1314. doi: 10.1007/s00500-021-06401-0
[14]	王福忠, 王帅峰, 张丽. 基于VMD-LSTM与误差补偿的光伏发电超短期功率预测[J]. 太阳能学报, 2022, 43(8):96-103. doi: 10.19912/j.0254-0096.tynxb.2021-0043
	WANG Fuzhong, WANG Shuaifeng, ZHANG Li. Ultra short term power prediction of photovoltaic power generation based on VMD-LSTM and error compensation[J]. Acta Energiae Solaris Sinica, 2022, 43(8):96-103. doi: 10.19912/j.0254-0096.tynxb.2021-0043
[15]	丁瑞成, 周玉成. 引入莱维飞行与动态权重的改进灰狼算法[J]. 计算机工程与应用, 2022, 58(23):74-82. doi: 10.3778/j.issn.1002-8331.2205-0577
	DING Ruicheng, ZHOU Yucheng. Improved grey wolf algorithm with Levy flight and dynamic weights[J]. Computer Engineering and Applications, 2022, 58(23):74-82. doi: 10.3778/j.issn.1002-8331.2205-0577
[16]	王浩, 王艳, 纪志成. 基于SAIGM-KELM的短期风电功率预测[J]. 电力系统保护与控制, 2020, 48(18):78-87.
	WANG Hao, WANG Yan, JI Zhicheng. Short term wind power prediction based on SAIGM-KELM[J]. Power System Protection and Control, 2020, 48(18):78-87.
[17]	LIU Liping, WEI Ying, SONG Xiuyun, et al. Fault diagnosis of wind turbine bearings based on CEEMDAN- GWO-KELM[J]. Energies, 2023, 16(1):48. doi: 10.3390/en16010048
[18]	吴波, 崔耀中, 韦汉, 等. 基于赋权灰色关联分析的聚能药包炮孔参数优化[J]. 工程爆破, 2022, 28(5):1-8.
	WU Bo, CUI Yaozhong, WEI Han, et al. Optimization of shaped charge blasthole parameters based on weighted gray correlation analysis[J]. Engineering Blasting, 2022, 28(5):1-8.
[19]	李洪, 章立新, 任燕, 等. 基于灰色关联分析的BP神经网络对混流闭式冷却塔出水温度的预测[J]. 计算物理, 2022, 39(1):53-59. doi: 10.19596/j.cnki.1001-246x.8376
	LI Hong, ZHANG Lixin, REN Yan, et al. Prediction of outlet water temperature of mixed-flow closed cooling tower using BP neural network based on grey correlation analysis[J]. Chinese Journal of Computational Physics, 2022, 39(1):53-59. doi: 10.19596/j.cnki.1001-246x.8376
[20]	王修君, 莫磊, 郑啸, 等. 面向数据流滑动窗口的自适应直方图发布算法[J]. 计算机科学, 2022, 49(10):344-352. doi: 10.11896/jsjkx.210700242
	WANG Xiujun, MO Lei, ZHENG Xiao, et al. Adaptive histogram publishing algorithm for data flow sliding window[J]. Computer Science, 2022, 49(10):344-352. doi: 10.11896/jsjkx.210700242
[21]	张英杰, 李佳林, 孙庆帅, 等. 基于半定长滑动窗口数据的供水管网漏损检测[J]. 湖南大学学报, 2022, 49(10):43-50.
	ZHANG Yingjie, LI Jialin, SUN Qingshuai, et al. Leakage detection of water supply network based on semi-fixed-length sliding window data[J]. Journal of Hunan University, 2022, 49(10):43-50.

基准函数	HPO		LHPO
基准函数	均值	标准差	均值	标准差
f₁	2.497×10^-56	3.862×10^-56	0	0
f₂	3.065×10^-5	3.746×10^-5	3.339×10^-7	7.284×10^-7
f₃	0	0	0	0

特征参数	参数类型	详细描述
经济因素	电价	使用1度电的价格
环境因素	湿度露点温度干球温度湿球温度	表示空气中大气干燥程度水蒸气与水达到平衡状态的温度被测量空气的真实热力学温度热力学绝热饱和温度
电力负荷	瓦	发电机组单位时间发电量

模型	RMSE/kW	MAPE(%)	MAE/kW
KELM	393.52	3.75	326.94
GRA-KELM	312.71	3.17	251.65

模态K	K=1	K=2	K=3	K=4	K=5
模态1	0.020	0.020	0.021	0.022	0.021
模态2	—	0.125	0.125	0.128	0.119
模态3	—	—	10.067	11.348	10.660
模态4	—	—	—	17.813	16.680
模态5	—	—	—	—	19.880

模型	RMSE/kW	MAPE(%)	MAE/kW
LSTM	324.54	2.915	252.64
SVR	383.38	3.364	306.05
VMD-SSA-KELM	196.87	2.123	180.60
VMD-WOA-KELM	152.87	1.553	133.65
LHPO-KELM	232.90	2.610	220.25
VMD-HPO-KELM	135.07	1.320	114.29
VMD-LHPO-KELM	77.61	0.665	59.06