基于准定向变异和变异回滚微分算法的风电场无功补偿

Quasi-Directional Variation and Rollback Mutation Based Differential Evolution Algorithm for Reactive Power Compensation of Wind Farms

电网无功补偿方案求解是高维非线性寻优问题,大规模风电场的接入增加了求解这一问题的难度。微分进化算法是求解此类问题的有效算法,但该算法对初始种群依赖性强且前期寻优逼近能力弱,因此提出准定向变异和变异回滚策略对其进行改进。以全网母线电压偏移和为目标函数,以基于准定向变异和变异回滚策略的改进微分进化算法为求解工具,建立求解大规模风电场无功补偿方案的计算模型。与普通微分进化算法相比,该算法在寻优初期具有寻优能力强且快速逼近最优解的优点,其确定的补偿方案可明显提高地区电网的电压水平。

Reactive power compensation of power grids is a kind of high-dimensional nonlinear optimization procedure that will become extremely complex with the integration of large-seale wind farms. Differential evolution algorithm is an effective method to solve the problem. To overcome its heavy dependence on the initial population and weak capability to approaeh to the optimization solution in the early process, a strategy based on quasi-directional variation and rollback mutation is presented to improve the performance of the algorithm. The sum of all bus voltage deviations is treated as the objective function and a computational model to design the compensation scheme is established. Compared with the ordinary differential evolution algorithm, the proposed algorithm is more efficient in searching for optimal solution with better convergence, such that the eompensatinn scheme can be obtained with significant improvements in the area voltage profiles.