采用启发式策略的动态无功优化混合算法研究

Study on dynamic reactive power optimization hybrid algorithm using heuristic strategy

严格考虑控制设备的动作次数限制,建立一个完整的非线性混合整数动态无功优化模型。结合混合算法,把离散变量和连续变量分开优化,分别使用改进后的遗传算法和内点法求解,以改善算法的收敛性和计算速度。将动态约束分为设备一天内调节次数限制和变压器相邻时段调节次数限制,提出全局调整策略和局部调整策略来分别处理这两类动态约束,使离散变量严格满足动作次数限制。启发式策略的应用使算法既能严格保证离散变量的动态约束,又能充分保留其动作自由,得到可行的最优解。Ward-hale6和IEEE30节点系统的仿真结果证明提出算法的有效性。

Considering the action number constraints of action times of control devices,an integrated mixed-integer/nonlinear dynamic reactive-power optimization（DRPO）model is proposed.Combining with the hybrid algorithm,continuous variables and discrete variables are optimized seperately,which are solved by ameliorative interior point method and genetic algorithm respectively.These ameliorative optimization algorithms improve the algorithm in convergence performance and calculation rate.Dynamic constraints include the constraints of switching operating times between two adjacent time intervals for transformers with on-load tap changers and the constraints of daily operating times for reactive power control devices,which are disposed by part adjust strategy and whole adjust strategy respectively,thus the discrete variables satisify the action time constraints strictly.The application of heuristic strategies can not only satisfy action number constraints of discrete variables absolutely,but also approach the feasible best result by keeping action freedom of devices adequately.Test results on Ward-hale 6-bus and IEEE 30-bus systems demonstrate the efficiency of the proposed algorithm.