风电场有功功率多目标分层递阶预测控制策略

Stratified Progressive Predictive Control Strategy for Multi-objective Dispatching Active Power in Wind Farm

为了协调有功功率调度控制精度、控制稳定性和风机机械损耗等多方面的问题,风电场有功控制策略需要具备多目标闭环协调优化的能力。该文首先根据提升风电场对系统有功调度指令的追踪精度、降低调度的波动次数和优化风机的发电状态等多个控制目标,将风电场内有功调度分为场站优化分配层、分群控制层和单机有功功率管理层3个层次,并在分群控制层中根据风机负荷状态和未来有功功率变化趋势进行风机分群。在场站优化分配层建立基于在线序贯极限学习机、最小二乘支持向量机的超短期有功功率组合预测模型,结合功率预测值和系统有功调度值对各机群有功负荷值进行优化分配,并下发至分群控制层。各机群根据场站层下发的调度指标和群内风机发电状态进行进一步滚动优化,从而提高控制的针对性,降低风机调度指令的波动范围和次数。在此基础上,单机有功功率管理层实现风机有功功率的实时优化控制和校正。此外,系统根据风电场实时有功功率数据反馈,对功率组合预测模型系统进行误差反馈校正,从而整体提高有功功率预测精度。经过与目前常用风电场场内有功功率分配算法对比,提出的有功控制策略可在提升有功控制精度的同时,降低风机调度次数,提高风电场有功调控的鲁棒性。

Wind farm active power control strategy needs multi-objective closed-loop coordination optimization ability for improving active power control accuracy, stability and reducing wind turbines mechanical loss. The control strategy proposed in this paper is divided into three layers, which are wind farm optimized distribution layer, flocked control layer and single turbine active power managing layer, according to the targets of improving the system active power dispatching values tracking accuracy, reducing wind turbine dispatching instructions fluctuations and wind turbine generation state optimizing. The ultra-short term wind power forecasting combined model is built in wind farm optimized distribution layer by using online sequential extreme learning machine and least squares support vector machine. The power forecasting values and system active power dispatching values are used for each wind turbine group active power load, which is delivered to flocked control layer. The active power generations of wind turbine groups are further rolling optimized based on dispatching values and wind turbine generation condition. The pertinency of power control will be improved and wind turbine dispatching changing range and frequency are reduced. Moreover, the wind power combined model forecasting errors are adjusted by real time active power values. Compared with the common used active power dispatching algorithms, this control strategy can improve the power control accuracy, reduce the wind turbine dispatching frequency and enhance active power robustness in wind farms.