考虑风力发电批特征的电力机组调度问题

Unit Commitment Problem for Wind Turbines Power Generation with Batching Characteristics Consideration

电力机组组合问题是在给定的计划周期内确定火电、风电和蓄电池机组的开关机状态及发电量,以满足系统的负荷需求、旋转备用等约束要求.为了降低风电在电网中的供电不稳定性,引入蓄电池储能系统与风机进行协调调度.由于大数量风机的介入,明显增加了问题处理的难度和复杂性.本文从一个新的视角将相近物理位置的风机进行组批,基于批的视角对问题建立了批模型.为了提高批模型的性能,提出了批模型参数的变换方法.根据问题的NP-难特征和模型的复杂结构,开发了拉格朗日松弛(Lagrangian relaxation,LR)算法进行求解.为了加速算法的求解效率,提出了子问题近似求解的代理次梯度的拉格朗日松弛算法.实验结果表明,提出的批模型明显优于传统的单机模型.基于批模型开发的拉格朗日松弛算法与CPLEX优化软件相比,能够在较短的时间内获得高质量的解.

The unit commitment problem is to determine the start-up/shut-down schedule and economical dispatch schedule of thermal generators, wind turbines and batteries to meet system load demand, reserved constraints, minimum up/down time constraints and other constraints within a certain time horizon. In order to reduce the power supply instability when wind power generation is plugged in the grid, coordinated scheduling of battery energy storage system introduced into the gird and wind turbines is performed. As a large number of wind turbines are plugged in the grid, the difficulties and complexities of the problem are increased significantly. In this paper, from a new batching perspective, we group wind turbines based on their physics locations to formulate the problem. In order to improve the performance of the batch model, a transformation method of model parameters is proposed. For tackling the complicated batch model and its NP-hardness, we develop a Lagrangian relaxation （LR） algorithm. In order to accelerate the algorithm, a surrogate subgradient Lagrangian relaxation algorithm is derived, in which subproblems are solved approximately. The experimental results show that the proposed batch model is superior to the ordinal single-unit model. Compared with CPLEX 11.0, the Lagrangian relaxation algorithm based on the batch model can obtain high quality solutions in a relatively short computation time.