电工技术学报2024，Vol.39Issue(13)：3975-3989,15.DOI:10.19595/j.cnki.1000-6753.tces.230788
基于非线性仿射的风电场电压实时计算和优化方法
Real-Time Voltage Calculation and Optimization Method for Wind Farms Based on Nonlinear Affine Transformation
摘要
Abstract
The uncertainty of wind speed leads to fluctuations in wind power,resulting in voltage fluctuations at the sending end and grid-connected point.Traditional methods for evaluating the impact of wind power fluctuations on the system state online either entail frequent power flow calculations to obtain real-time system voltages,which creates a substantial computational burden,or employ linear affine methods that yield significant errors when wind power fluctuations are large.To optimize the sending-end voltage and grid-connected point voltage considering the wind power fluctuations,centralized optimization methods require frequent optimal power flow calculations,leading to heavy computational burdens.On the other hand,distributed algorithms require advanced communication facilities,which many wind farms may struggle to meet.Therefore,this paper proposes a real-time voltage calculation and optimization method based on nonlinear affine transformation to rapidly calculate the voltage state and optimize the system voltage considering wind power fluctuations,with limited computation and communication resources. Firstly,the method establishes a multi-period dynamic reactive power optimization model to minimize the power loss and the voltage fluctuations of the grid-connected point.The cumulant method is used to analyze the sending-end voltage influenced by uncertain wind generation.Secondly,an improved interior point method,based on the analytic gradient matrix and Hessian matrix,is utilized to solve the established model,with the obtained results serving as a reference for real-time voltage optimization.Thirdly,at each moment the wind generator measures the local active power fluctuation and uses the designed event-triggered algorithm to determine whether real-time voltage optimization is necessary.Then,the nonlinear affine method is used to rapidly calculate the real-time system voltage state,and obtain the reactive power adjustment to prevent potential voltage violations at the sending end,while also reducing voltage fluctuations at the grid-connected point.Finally,the reactive power of each wind generator is regulated simultaneously by controlling the converter to achieve real-time voltage optimization of the whole wind farm.Overall,the proposed real-time voltage optimization method can improve the voltage quality of wind farms in real time. The following conclusions can be drawn from the simulation analysis:(1)The solution speed is greatly improved by adding a penalty term to the objective function of the multi-period dynamic reactive power optimization model.The improved interior point method based on analytic gradient and Hessian matrix can solve the model efficiently.(2)Based on the third-order nonlinear affine transformation of system voltage w.r.t.active/reactive power derived in this paper,the proposed method could rapidly and accurately obtain the system voltage state considering the wind power fluctuations.The computational time cost is about 0.58％of the power flow calculation,and the calculation results are more accurate compared with the linear affine method.(3)The proposed real-time voltage optimization method based on nonlinear affine transformation could effectively prevent sending-end voltage violations and further reduce the grid-connected point voltage fluctuations,thus improving the voltage quality of wind farms.Statistically,the sending-end voltage violations are eliminated throughout the simulation,and the voltage fluctuation of the grid-connected point is reduced to 35.12％of the conventional method.关键词
多时段动态无功优化/非线性仿射变换/实时计算和优化/事件触发Key words
Multi-period dynamic reactive power optimization/nonlinear affine transformation/real-time calculation and optimization/event-triggered分类
信息技术与安全科学引用本文复制引用
张兆毅,胡浩,王子江,商犇,樊友平,舒印彪..基于非线性仿射的风电场电压实时计算和优化方法[J].电工技术学报,2024,39(13):3975-3989,15.基金项目
国家重点研发计划资助项目(2022YFF0610600). （2022YFF0610600）
