摘要
为了在提升飞轮-火电联合二次调频性能的同时控制调频成本、减少飞轮损耗,面向飞轮-火电联合系统,提出一种考虑自动发电控制(AGC)指令及电网频率偏差的基于多目标遗传算法的负荷分配策略。该策略以电网调频偏差、飞轮-火电出力能量和、飞轮损耗为三层最小化目标,考虑火电机组爬坡率及其负荷备用、飞轮荷电状态及其功率等约束条件,基于改进的非支配排序遗传算法(NSGA-Ⅱ),通过AGC功率指令变化速度及频率偏差动态调整飞轮与火电出力权重系数,以频率偏差及其变化方向调整飞轮虚拟下垂、惯性系数,求解飞轮-火电出力比例、飞轮下垂系数和飞轮惯性系数。结合中国宁夏某电厂飞轮储能系统辅助火电机组调频现场AGC指令等历史数据,在双区域两机系统模型上进行仿真实验,结果表明该负荷分配策略有效地减小了火电二次调频偏差和调节时间,提升了调频性能,同时也有效地控制了系统出力能量及飞轮损耗。
To enhance the performance of the flywheel-thermal power coordinated secondary frequency regulation while controlling the frequency regulation cost and reducing the flywheel system loss,this paper proposes a load allocation strategy based on a multiobjective genetic algorithm considering automatic generation control(AGC)commands and grid frequency deviations for flywheelthermal power coordinated system.The proposed strategy aims to minimize three layers of objectives:grid frequency deviation,the sum of flywheel-thermal power output energy,and flywheel losses,which takes into account constraints such as the ramp rate and load reserve of thermal power generating units,state of charge(SOC)and power of flywheels.The weighting coefficients of flywheel and thermal power output are dynamically adjusted by rate of change of automatic generation control(AGC)power commands and frequency deviations based on the improved non-dominated sorting genetic algorithm-Ⅱ(NSGA-Ⅱ).The virtual droop and inertia coefficients of the flywheel are also ajusted based on the frequency deviation and its changing direction.The proposed strategy solves the flywheel-thermal power output ratio,flywheel droop coefficient,and flywheel inertia coefficient.Combining historical data,including AGC commands at a certain power plant with flywheel energy storage-assisted thermal power frequency regulation,simulation experiments are conducted on a two-area two-machine system model.The results indicate that the load allocation strategy effectively reduces the secondary frequency regulation deviation and adjustment time of thermal power,improves the frequency regulation performance,and controls the system output energy and flywheel losses at the same time.
作者
房方
刘渝斌
王冰玉
魏乐
李承霖
FANG Fang;LIU Yubin;WANG Bingyu;WEI Le;LI Chenglin(School of Control and Computer Engineering,North China Electric Power University,Beijing 102206,China)
出处
《电力系统自动化》
EI
CSCD
北大核心
2024年第12期79-88,共10页
Automation of Electric Power Systems
基金
国家自然科学基金资助项目(52176005)
关键词
火电
飞轮储能
调频
负荷分配策略
多目标遗传算法
虚拟下垂
虚拟惯性
thermal power
flywheel energy storage
frequency regulation
load allocation strategy
multi-objective genetic algorithm
virtual droop
virtual inertia