摘要
区域电网的互联互通使电网结构日趋复杂,弱阻尼模式引发的低频振荡会影响系统安全稳定运行,对电力系统小干扰稳定带来了严峻挑战。针对常规小干扰稳定评估方法计算复杂度高、不能满足实时分析需求的问题,提出基于随机森林算法的小干扰稳定评估校正方法。首先,利用随机森林算法建立发电机有功、无功出力,负荷的有功、无功功率和支路功率等数据与最小阻尼比之间的映射模型,得到基于随机森林算法的小干扰稳定评估模型;然后,对发电机参数进行微小摄动,放入随机森林模型中算出最小阻尼比的变化,用数值分析法得到发电机阻尼比灵敏度;接着,建立系统弱阻尼状态下发电机阻尼比的优化模型,优化调整发电机有功出力;最后,在IEEE-9节点系统和某区域615节点系统中进行仿真验证。仿真结果表明,所提模型能够准确估算出系统运行的最小阻尼比和近似阻尼比灵敏度,将系统的最小阻尼比校正到0.03以上,消除弱阻尼状态,使系统恢复稳定运行。
The interconnection of regional power grids makes the grid structure increasingly complex,and the low-frequency oscillation triggered by weak damping modes affects the safe and stable operation of the system,which causes serious challenges to the small signal stability of the power system.In order to overcome the problem that the traditional small signal stability assessment method has high computational complexity and can’t meet the requirements of real-time analysis,this paper proposes an assessment and correction method for small signal stability assessment based on random forest(RF).Firstly,it uses the RF algorithm to establish a mapping model between the data of active and reactive power output of the generator,active and reactive power and branch power of the load and the minimum damping ratio to get the small signal stability evaluation model based on RF algorithm.Then the generator parameters are slightly ingested and put into the RF model to calculate the change of the minimum damping ratio,and the generator damping ratio sensitivity is obtained by numerical analysis.It also establishes the optimization model of the generator damping ratio in the weakly damped state of the system to optimize active output adjustment of the generator.Finally,the paper carries out the simulation in the IEEE-9 bus system and a regional 615-bus system for verification.The simulation results show that the sensitivity of approximate damping ratio of the system can be assessed accurately,the minimum damping ratio can be improved to above 0.03,the weak damping state can be eliminated,and the system can be restored to stable operation based on the proposed model.
作者
姚海成
李豹
黄兆棽
YAO Haicheng;LI Bao;HUANG Zhaoshen(CSG Power Dispatching and Control Center,Guangzhou,Guangdong 510663,China)
出处
《广东电力》
2023年第7期60-69,共10页
Guangdong Electric Power
基金
中国南方电网有限责任公司科技项目(0000002022030103XT00028)。
关键词
小干扰稳定
阻尼比
灵敏度
随机森林算法
评估校正
small signal stability
damping ratio
sensitivity
random forest
assessment correction