Research on Generator Tripping Control Strategy Based on Deep Reinforcement Learning

doi:10.16628/j.cnki.2095-8188.2023.03.002

Abstract

Abstract:

The power system will enter an emergency state after being greatly disturbed.The emergency control measures must be taken in time to restore the system to stable operation.Generator tripping control is the most effective and common control measure to maintain system stability.Aiming at the problem of fault mismatch in practical application of the traditional control method based on cure table,a decision method of power system transient stability generator tripping control based on deep reinforcement learning is proposed.Firstly,the deep deterministic policy gradient (DDPG) algorithm is introduced.Every element of the algorithm is redesigned in combination with the equal area criterion.Secondly,the decision model of generator tripping control based on DDPG algorithm is established.Finally,using PSA-BPA and Pycharm software,the generator tripping control simulation models of the single machine-infinite system and an IEEE39 node system are established.The effectiveness of the proposed method is verified by an example.

Key words: transient stability, generator tripping control, deep reinforcement learning, deep deterministic policy gradient

CLC Number:

TM712

LU Hengguang, LIN Bilin, WEN Buying. Research on Generator Tripping Control Strategy Based on Deep Reinforcement Learning[J]. LOW VOLTAGE APPARATUS, 2023, 0(3): 11-15.

Figures/Tables 7

References 17

[1]	杨海晶, 贾学翠, 高东学, 等. 分布式储能在电力系统的应用及现状分析[J]. 电器与能效管理技术, 2018(19):12-15.
[2]	郑广宇, 黄开艺, 艾芊. 电力市场改革下大用户购电电量交易模型[J]. 电器与能效管理技术, 2019(16):84-88.
[3]	翁智敏, 朱振山, 温步瀛, 等. 高比例新能源电力系统研究综述[J]. 电器与能效管理技术, 2021(11):1-7.
[4]	ZHANG C, LI Y, YU Z, et al. A weighted random forest approach to improve predictive performance for power system transient stability assessment[C]//2016 IEEE PES Asia-Pacific Power and Energy Engineering Conference (APPEEC), 2016:1259-1263.
[5]	WANG H, CHEN Q, ZHANG B. A transient stability assessment combined model framework based on cost-sensitive method[J]. IET Generation,Transmission and Distribution, 2020, 14(12):2256-2262. doi: 10.1049/gtd2.v14.12
[6]	刘强. 限定电力系统暂态过程最大摇摆角的控制策略整定方法[J]. 电器与能效管理技术, 2018(19):12-15.
[7]	顾卓远, 汤涌, 张健, 等. 基于相对动能的电力系统暂态稳定实时紧急控制方案[J]. 中国电机工程学报, 2014, 34(7):1095-1102.
[8]	王怀远, 张保会, 杨松浩, 等. 基于相平面特性的切机切负荷紧急控制方法[J]. 中国电机工程学报, 2016, 36(15):4144-4152.
[9]	杨少波, 刘道伟, 安军, 等. 基于长短期记忆网络的电网动态轨迹趋势预测方法[J]. 中国电机工程学报, 2020, 40(9):2854-2866.
[10]	王彤, 刘九良, 朱劭璇, 等. 基于随机森林的电力系统暂态稳定评估与紧急控制策略[J]. 电网技术, 2020, 44(12):4694-4701.
[11]	YU T, ZHEN W G. A reinforcement learning approach to power system stabilizer[C]//2009 IEEE Power & Energy Society General Meeting, 2009:1-5.
[12]	ZHANG D, HAN X, DENG C. Review on the research and practice of deep learning and reinforcement learning in smart grids[J]. 中国电机工程学会电力与能源系统学报(英文), 2018, 4(3):362-370.
[13]	李红, 朱立位, 郭朝辉. 基于深度强化学习的混合储能系统最优控制[J]. 电器与能效管理技术, 2021(11):42-49.
[14]	乔骥, 王新迎, 张擎, 等. 基于柔性行动器-评判器深度强化学习的电-气综合能源系统优化调度[J]. 中国电机工程学报, 2021, 41(3):819-833.
[15]	WEI F, WAN Z, HE H. Cyber-Attack Recovery Strategy for Smart Grid Based on Deep Reinforcement Learning[J]. IEEE Transactions on Smart Grid, 2020, 11(3):2476-2486. doi: 10.1109/TSG.5165411
[16]	HUANG Q, HUANG R, HAO W, et al. Adaptive power system emergency control using deep reinforcement learning[J]. IEEE Transactions on Smart Grid, 2020, 11(2):1171-1182. doi: 10.1109/TSG.5165411
[17]	刘威, 张东霞, 王新迎, 等. 基于深度强化学习的电网紧急控制策略研究[J]. 中国电机工程学报, 2018, 38(1):109-119.

项目	单机无穷大系统	IEEE39节点系统
网络层数/层	3	4
网络神经元个数/个	128,64	512,128,64
网络学习率/p.u.	0.000 1,0.001	0.000 5,0.001
最大训练次数/次	600	1 500

项目	单机无穷大系统	IEEE39节点系统
故障类型	三相短路	三相短路
故障线路	输电线路1	21-22
故障发生率/%	50	50
故障持续时间/周波	13、13.5、14、14.5	14.5、15、15.5、16、16.5

故障持续时间/ 周波	临界切机量/p.u.	训练400次		训练600次
故障持续时间/ 周波	临界切机量/p.u.	切机量/p.u.	相对误差/%	切机量/p.u.	相对误差/%
13	0.042	0.062 8	49.52	0.055 4	31.91
13.5	0.257	0.284 2	10.58	0.270 9	5.41
14	0.488	0.587 2	20.28	0.574 9	17.76
14.5	0.738	0.761 5	3.18	0.753 2	2.06

故障持续时间/ 周波	临界切机量/p.u.	训练1 000次		训练1 500次
故障持续时间/ 周波	临界切机量/p.u.	切机量/p.u.	相对误差/%	切机量/p.u.	相对误差/%
14.5	0.100	0.140 5	40.50	0.116 2	16.20
15	0.240	0.336 3	40.13	0.258 0	7.50
15.5	0.392	0.495 0	26.28	0.399 3	1.86
16	0.564	0.696 2	23.44	0.608 8	7.94
16.5	0.772	0.876 3	13.51	0.804 5	4.21