风储一体化电站容量双层优化规划研究

Research on Bi-level Capacity Programming Optimization for the Integration of Wind Farm Energy Storage Power Station

在风储一体化电站系统规划问题中,为了达到系统总容量最大且全局经济性最优的目的。提出一种双层规划模型方法:上层规划模型以风储一体化电站总容量最大为目标,考虑当地风能资源与电网资产构成的风电消纳能力等约束条件;下层规划模型以风储一体化系统一次性投资费用最小为目标,优化风储配比,主要考虑风储出力目标和电力瞬时平衡等约束。并设计嵌套的遗传算法对该模型进行求解。其次,通过某风电场实例验证,得出该模型规划的风储一体化系统在无风或系统出力与调度计划出力不一致的情况时能以2.8 MW的风电场期望功率持续出力4.9 h的结果,表明该模型规划的系统在一定时间内具备响应电力系统出力需求的能力。最后,进一步分析了发电机组类型和风资源Weibull参数对风储容量规划结果的影响,结果表明当形状参数位于1-3和尺度参数位于4-10的区域时其规划结果具有很强的适应性,同时也指出规划模型没有能够全面反映储能作用的局限性。

In the system programming of wind power and energy storage integrated power station, a kind of bi-level programming model, which not only ensured maximized capacity of the wind power and energy storage integrated power station, but also ensured the optimal integral economy, was put forward. The target of upper programming model is to en- sure the maximized capacity of the wind power and energy storage integrated power station, considering the constraint conditions such as consumption capability of wind power which is constituted by local wind energy resources and grid assets; the target of lower programming model is to minimize the one-time investment of wind power and energy storage integrated system, optimizing the wind power and storage proportion, considering the constraints such as the transient equilibrium of electricity and the power output target of wind power and energy storage integrated power system. And nested GA （Genetic Algorithm） is designed to solve the given model, through example confirmation, we can conclude that the wind power and energy storage integrated power system can continuously work about 4.9 h with 2.8 MW pow- er（expected wind farm power） in the windless condition or when the system output and scheduling output plan are not consistent. The results show that the system programmed by above model has the ability to response to the output re- quirements of power system in a certain period of time. The effects of wind turbine type and Weibull parameters on the capacity planning results are further analyzed, and the result shows that the programming result of the model in the area between the shape parameter of 1-3 and scale parameter of 4-10 has a strong adaptability, revealing that the given pro- gramming model is not able to fully reflect the energy storage function.