基于正则化贪心森林的多维频率指标智能化预测方法

doi:10.12204/j.issn.1000-7229.2020.09.014

电力建设 ›› 2020, Vol. 41 ›› Issue (9): 124-131.doi: 10.12204/j.issn.1000-7229.2020.09.014

基于正则化贪心森林的多维频率指标智能化预测方法

黄明增¹, 文云峰¹, 汪荣华², 胥威汀², 李婷², 苟竞², 赵荣臻³

1.湖南大学电气与信息工程学院, 长沙市 410082
2.国网四川省电力公司经济技术研究院,成都市 610041
3.南京南瑞继保电气有限公司,南京市 211102

收稿日期:2020-03-12 出版日期:2020-09-01 发布日期:2020-09-03
作者简介:黄明增(1996) ,男,硕士研究生,主要研究方向为电力系统稳定性智能化评估与控制方法;|文云峰(1986),男,博士(后),副教授,博士生导师,主要研究方向为低惯性电力系统规划、运行与控制;|汪荣华(1964),男,高级工程师,主要从事电力系统规划与设计的研究工作;|胥威汀(1985),男,博士,高级工程师,主要从事电力系统规划与设计的研究工作;|李婷(1979),女,硕士,高级工程师,主要从事电力系统规划与设计的研究工作;|苟竞(1988),男,博士,高级工程师,主要从事电力系统规划与设计的研究工作;|赵荣臻(1995),男,硕士,工程师,主要从事低惯性电力系统频率稳定智能化评估的研究工作。
基金资助:
国家自然科学基金资助项目(51707017);湖南省自然科学优秀青年基金项目(2020JJ3011);国家电网公司科技项目(SGSCJY00GHJS1900010)

Intelligent Prediction for Multi-dimensional Frequency Indicators Based on Regularized Greedy Forests

HUANG Mingzeng¹, WEN Yunfeng¹, WANG Ronghua², XU Weiting², LI Ting², GOU Jing², ZHAO Rongzhen³

1. College of Electrical and Information Engineering, Hunan University, Changsha 410082, China
2. State Grid Sichuan Economic Research Institute, Chengdu 610041, China
3. Nanjing NR Electric Co., Ltd., Nanjing 211102, China

Received:2020-03-12 Online:2020-09-01 Published:2020-09-03
Supported by:
National Natural Science Foundation of China(51707017);Hunan Provincial Natural Science Foundation for Excellent Young Scholars(2020JJ3011);State Grid Corporation of China Research Program(SGSCJY00GHJS1900010)

摘要/Abstract

摘要：

为实现海量预想故障下电力系统频率响应性能的快速和精确感知,文章提出了一种基于正则化贪心森林(regularized greedy forests, RGF)的多维频率指标智能化预测方法。该方法采用RGF建立特征输入与多维频率指标之间的非线性映射关系,通过对全局参数进行优化,并引入3种正则化机制,使所构建的机器学习模型能够有效表征复杂函数,防止过拟合。为保证预测算法的性能,通过网格搜索遍历参数组合,以确定所构建RGF模型的最佳参数。在改进的IEEE RTS-79系统上开展了算例测试,与时域仿真、随机森林和梯度提升方法所得结果进行比较,验证了所提方法的准确性、快速性以及良好的泛化能力。

关键词: 频率, 惯性, 智能化预测, 正则化贪心森林, 网格搜索法

Abstract:

In order to realize fast and precise perception of the frequency response performance of the power system under the background of massive anticipated faults, this paper proposes a method for predicting the multi-dimensional frequency indicators based on regularized greedy forests (RGF). This method establishes the non-linear mapping relationship between input features and multi-dimensional indicators through RGF. By optimizing global parameters and introducing three regularization mechanisms to the decision-making forest, the RGF can effectively represent complex functions and prevent overfitting. To ensure the performance of the model, the combinations of parameters are traversed by the grid search to find the best parameter of the constructed RGF model. Case studies on the modified IEEE RTS-79 system demonstrate the high precision, rapidity, and well generalization ability of the proposed method.

Key words: frequency, inertia, intelligent prediction, regularized greedy forests, grid search

中图分类号:

TM71

黄明增, 文云峰, 汪荣华, 胥威汀, 李婷, 苟竞, 赵荣臻. 基于正则化贪心森林的多维频率指标智能化预测方法[J]. 电力建设, 2020, 41(9): 124-131.

HUANG Mingzeng, WEN Yunfeng, WANG Ronghua, XU Weiting, LI Ting, GOU Jing, ZHAO Rongzhen. Intelligent Prediction for Multi-dimensional Frequency Indicators Based on Regularized Greedy Forests[J]. ELECTRIC POWER CONSTRUCTION, 2020, 41(9): 124-131.

图/表 13

图1 由3颗决策树组成的决策森林

Fig.1 A decision forest composed of three decision trees

表1 输入特征集

Table 1 Input feature set

图2 基于RGF的频率指标预测模型

Fig.2 Model of frequency matrix prediction based on RGF

图3 基于RGF的频率指标预测方法流程

Fig.3 Flowchart of frequency matrix prediction

图4 改进后的IEEE RTS-79系统

Fig.4 Modified IEEE RTS-79 system

图5 网格搜索法调参可视化分析

Fig.5 The visualization of grid search

表2 极值频率误差结果

Table 2 Errors of the fnadir Hz

图6 极值频率误差分布

Fig.6 Error distribution of the fnadir

图7 准稳态频率误差分布

Fig.7 Error distribution of the fss

图8 最大频率变化率误差分布

Fig.8 Error distribution of the RoCoFmax

表3 准稳态频率误差结果

Table 3 Errors of the fss Hz

表4 最大频率变化率误差结果

Table 4 Error results of the RoCoFmaxHz/s

表5 RGF与时域仿真的计算时间比较

Table 5 Comparison of RGF and time-domain simulation in times

参考文献 21

[1]	周孝信, 陈树勇, 鲁宗相 , 等. 能源转型中我国新一代电力系统的技术特征[J]. 中国电机工程学报, 2018,38(7):1893-1904, 2205.
	ZHOU Xiaoxin, CHEN Shuyong, LU Zongxiang , et al. Technology features of the new generation power system in China[J]. Proceedings of the CSEE, 2018,38(7):1893-1904, 2205.
[2]	林鸿基, 闫园, 文福拴 , 等. 高比例可再生能源电力系统中计及灵活调节产品的实时调度[J]. 电力建设, 2019,40(10):18-27.
	LIN Hongji, YAN Yuan, WEN Fushuan , et al. Real-time dispatch considering flexible ramping products in A power system with high proportion of renewable energy generation[J]. Electric Power Construction, 2019,40(10):18-27.
[3]	王琦, 郭钰锋, 万杰 , 等. 适用于高风电渗透率电力系统的火电机组一次调频策略[J]. 中国电机工程学报, 2018,38(4):974-984, 1274.
	WANG Qi, GUO Yufeng, WAN Jie , et al. Primary frequency regulation strategy of thermal units for a power system with high penetration wind power[J]. Proceedings of the CSEE, 2018,38(4):974-984, 1274.
[4]	ULBIG A, BORSCHE T S, ANDERSSON G . Impact of low rotational inertia on power system stability and operation[J]. IFAC Proceedings Volumes, 2014,47(3):7290-7297.
[5]	WEN Y F, LI W Y, HUANG G , et al. Frequency dynamics constrained unit commitment with battery energy storage[J]. IEEE Transactions on Power Systems, 2016,31(6):5115-5125.
[6]	方勇杰 . 英国“8·9”停电事故对频率稳定控制技术的启示[J]. 电力系统自动化, 2019,43(24):1-7.
	FANG Yongjie . Reflections on frequency stability control technology based on the blackout event of 9 August 2019 in UK[J]. Automation of Electric Power Systems, 2019,43(24):1-7.
[7]	张恒旭, 李常刚, 刘玉田 , 等. 电力系统动态频率分析与应用研究综述[J]. 电工技术学报, 2010,25(11):169-176.
	ZHANG Hengxu, LI Changgang, LIU Yutian , et al. Reviews on power system dynamic frequency analysis and its application[J]. Transactions of China Electrotechnical Society, 2010,25(11):169-176.
[8]	ANDERSON P M, MIRHEYDAR M . A low-order system frequency response model[J]. IEEE Transactions on Power Systems, 1990,5(3):720-729.
[9]	MITCHELL M A, LOPES J A P, FIDALGO J N , et al. Using a neural network to predict the dynamic frequency response of a power system to an under-frequency load shedding scenario [C]//2000 Power Engineering Society Summer Meeting. Seattle: IEEE, 2000: 346-351.
[10]	胡益, 王晓茹, 滕予非 , 等. 基于多层支持向量机的交直流电网频率稳定控制方法[J]. 中国电机工程学报, 2019,39(14):4104-4118.
	HU Yi, WANG Xiaoru, TENG Yufei , et al. Frequency stability control method of AC/DC power system based on multi-layer support vector machine[J]. Proceedings of the CSEE, 2019,39(14):4104-4118.
[11]	XU Y, DAI Y, DONG Z Y , et al. Extreme learning machine-based predictor for real-time frequency stability assessment of electric power systems[J]. Neural Computing and Applications, 2013,22(3/4):501-508.
[12]	仉怡超, 闻达, 王晓茹 , 等. 基于深度置信网络的电力系统扰动后频率曲线预测[J]. 中国电机工程学报, 2019,39(17):5095-5104, 5290.
	ZHANG Yichao, WEN Da, WANG Xiaoru , et al. A method of frequency curve prediction based on deep belief network of post-disturbance power system[J]. Proceedings of the CSEE, 2019,39(17):5095-5104, 5290.
[13]	赵荣臻, 文云峰, 叶希 , 等. 基于改进堆栈降噪自动编码器的预想事故频率指标评估方法研究[J]. 中国电机工程学报, 2019,39(14):4081-4093.
	ZHAO Rongzhen, WEN Yunfeng, YE Xi , et al. Research on frequency indicators evaluation of disturbance events based on improved stacked denoising autoencoders[J]. Proceedings of the CSEE, 2019,39(14):4081-4093.
[14]	纪昌明, 周婷, 向腾飞 , 等. 基于网格搜索和交叉验证的支持向量机在梯级水电系统隐随机调度中的应用[J]. 电力自动化设备, 2014,34(3):125-131.
	JI Changming, ZHOU Ting, XIANG Tengfei , et al. Application of support vector machine based on grid search and cross validation in implicit stochastic dispatch of cascaded hydropower stations[J]. Electric Power Automation Equipment, 2014,34(3):125-131.
[15]	FRIEDMAN J H . Greedy function approximation: a gradient boosting machine[J]. The Annals of Statistics, 2001,29(5):1189-1232.
[16]	JOHNSON R, ZHANG T . Learning nonlinear functions using regularized greedy forest[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2014,36(5):942-954. URL pmid: 26353228
[17]	NG A Y. Feature selection, L1 vs. L2 regularization, and rotational invariance[C]//Twenty-First International Conference on Machine Learning. New York: ACM Press, 2004: 78-86.
[18]	刘克天, 王晓茹, 薄其滨 . 基于广域量测的电力系统扰动后最低频率预测[J]. 中国电机工程学报, 2014,34(13):2188-2195.
	LIU Ketian, WANG Xiaoru, BO Qibin . Minimum frequency prediction of power system after disturbance based on the WAMS data[J]. Proceedings of the CSEE, 2014,34(13):2188-2195.
[19]	王钦, 蒋怀光, 文福拴 , 等. 智能电网中大数据的概念、技术与挑战[J]. 电力建设, 2016,37(12):1-10.
	WANG Qin, JIANG Huaiguang, WEN Fushuan , et al. Concept, technology and challengs of big data in smart grids[J]. Electric Power Construction, 2016,37(12):1-10.
[20]	候智圆 . 电力系统频率动态行为特征分析[D]. 济南: 山东大学, 2012.
	HOU Zhiyuan . Dynamic characteristic analysis of power system frequency[D]. Jinan: Shandong University, 2012.
[21]	全国电压电流等级和频率标准化技术委员会. 电能质量电力系统频率偏差: GB/T 15945—2008[S]. 北京: 中国电力出版社, 2008.

基于正则化贪心森林的多维频率指标智能化预测方法

Intelligent Prediction for Multi-dimensional Frequency Indicators Based on Regularized Greedy Forests

RichHTML

PDF

可视化

摘要/Abstract

引用本文

使用本文

图/表 13

参考文献 21

相关文章 15

编辑推荐

Metrics

本文评价

[1]	崔红芬, 王德顺, 杨波, 吴福保, 朱天天, 赵晶晶. 支撑电网黑启动的风光储新能源电站协调控制策略[J]. 电力建设, 2020, 41(9): 50-57.
[2]	姚建华, 胡晟, 王冠, 沈云, 姜林林, 冯宇立, 龚成亚, 张照轩, 柳伟. 基于强化学习的孤岛微电网多源协调频率控制方法[J]. 电力建设, 2020, 41(9): 69-75.
[3]	赵波, 章德, 胡柳君, 马俊杰, 李勇, 王炜宇. 基于机电扰动传播特性的电网惯量分布辨识方法[J]. 电力建设, 2020, 41(8): 25-31.
[4]	肖友强, 林晓煌，文云峰. 直流和新能源高渗透型电网惯性水平多维度评估[J]. 电力建设, 2020, 41(5): 19-27.
[5]	袁丽丽，阳育德，覃智君，刘辉. 基于谐波阻抗特性的风电次同步振荡分析[J]. 电力建设, 2020, 41(4): 117-125.
[6]	陆秋瑜，杨银国，王中冠，谭嫣，夏天，伍双喜. 基于次梯度投影分布式控制法的虚拟电厂经济性一次调频[J]. 电力建设, 2020, 41(3): 79-85.
[7]	刘东明，杨杨，王军波，史海涛，马鸿君，张虹. 基于事件触发的互联电网负荷频率模型预测输出反馈控制[J]. 电力建设, 2020, 41(2): 108-117.
[8]	宋洁，侯继彪. 基于压缩空气储能系统的直流配电网调压控制方法[J]. 电力建设, 2019, 40(6): 49-56.
[9]	夏世威，陆涛，杜婷，李庚银，卜思奇. 考虑风电波动性的微网中电动汽车分布式协调运行方法[J]. 电力建设, 2019, 40(6): 86-94.
[10]	宋振浩，陆立民，胥峥，吕志鹏，吴鸣，赵婷. VSG虚拟惯量动态区间控制策略[J]. 电力建设, 2019, 40(2): 45-53.
[11]	钟永洁，翟苏巍，孙永辉. 孤岛模式下互联微电网的自适应同步频率控制[J]. 电力建设, 2019, 40(10): 75-83.
[12]	朱全胜,常烨骙，张振安，赵华，晋萃萃，吴玉琼，刘娆. 面向各类功率扰动的频率控制及评估问题探讨[J]. 电力建设, 2018, 39(9): 129-136.
[13]	熊雄1，孙庆凯1，井天军1，孙可2，王坤3. 基于多分段P/f特性曲线的独立微电网协调控制策略[J]. 电力建设, 2018, 39(4): 51-.
[14]	李盟盟1，王育飞1，薛花1，张宇2，方陈2. 不平衡电网下无锁相环逆变器模型预测控制[J]. 电力建设, 2018, 39(4): 100-.
[15]	徐政, 杨健, 盛能进, 陆韶琦. 半波长级与全波长级输电系统的过电压与同步稳定性分析[J]. 电力建设, 2018, 39(3): 23-.