基于分布式模型预测控制的风电场有功控制系统的研究

With large-scale integration of wind power, active power control of wind farms is very important to improve the stability and reliability of power systems. The optimal wind farm control considering the mechanical fatigue loads can significantly reduce the mechanical fatigue loads experienced by wind turbines, while tracking the power reference specified by the system operator. Therefore, the service lifetime of wind turbines can be prolonged and the maintenance cost of wind farms can be reduced. This project introduces Distributed Model Predictive Control (DMPC) into the active power control of wind farms and aims to improve it in both theoretical and technical aspects. The contents include the following three parts. Firstly, the impact of time delay on the DMPC of wind farms will be analyzed and quantified, and a control law considering asymptotic stability and time delay will be designed. Secondly, the online estimation of wind turbine model parameters and state variables will be studied. Accordingly, the online joint estimator and feedback compensator will be designed. Thirdly, an adaptive law of weights of multi-objective optimization will be developed. With the proposed control structure, the present wind farms can be easily updated. Besides, the distributed control algorithm can significantly improve the online computation efficiency. Therefore, it is very promising for the online optimal control of various scales, especially large-scale of wind farms. Based on the studies above, a prototype of the D-MPC based controller will be designed and developed. The prototype will be verified by software simulation as well as Hardware In the Loop (HIL) test for the future industrial application.

大规模风电并网的条件下, 风电场有功功率控制对提高电网的稳定性和可靠性都具有重要意义。计及机械疲劳的风电场优化控制，可以在满足电网调度目标的条件下，有效减低引起机械疲劳加剧的变量变化率，减轻机械疲劳，延长机组寿命，降低风电场维护成本。本项目将分布式模型预测控制应用于风电场有功优化控制，并在理论和技术方面加以完善。内容主要包括: 针对时滞现象对风电场分布式控制的影响进行量化分析，设计考虑时滞和渐进稳定的控制率；开展对风机参数和状态变量的在线估计、实时校正的研究，设计在线联合估计和反馈校正算法；制定多目标权重自适应优化分配方案，设计权重自适应算法。所提出的控制框架易于现有风电场的升级，分布控制算法能明显提高在线计算效率，适用于各种规模、尤其是大规模风电场的实时在线优化控制，具有广阔的应用前景。在此基础上，开发分布式控制器样机，并进行软件和半实物仿真验证，为产业化应用打下基础。

大规模风电并网的条件下, 大规模风电场功率协调控制，尤其是有功功率的控制，对提高电网的稳定性和可靠性都具有重要意义。计及机械疲劳的风电场优化控制，可以在满足电网调度目标的条件下，有效减低引起机械疲劳加剧的变量变化率，减轻机械疲劳，延长机组寿命，降低风电场维护成本。本项目将模型预测控制应用于风电场有功优化控制，将风电机组的疲老载荷引入优化控制目标中，在风电场分布式控制架构及算法、模型预测控制的多目标权重计算、风电场精细化仿真平台搭建、硬件控制器的开发、以及多风电场经柔直并网稳定性分析等方面开展深入研究，通过实验平台和硬件控制器的仿真验证，控制框架易于现有风电场的升级，控制算法能明显提高在线计算效率，适用于各种规模、尤其是大规模风电场的实时在线优化控制，能够大幅降低风电场整体的载荷，具有广阔的应用前景。