Reactive power optimization is a key technology for maintaining voltage stability in power systems and reducing network losses. In operating scenarios with high penetration rates of new energy, traditional optimization algorithms such as Particle Swarm Optimization (PSO) and Genetic Algorithm (GA) are often prone to converge to local optimal solutions and suffer from insufficient coordination between devices. To address the aforementioned issues, this paper proposes an improved Adaptive Particle Swarm Optimization (APSO) algorithm. This algorithm constructs a reactive power optimization model containing multiple constraints, with minimizing system loss and voltage deviation as the dual objective functions. By dynamically adjusting the inertia weight and adopting an adaptive learning factor strategy, this method significantly improves the balance between global exploration and local optimization capabilities. Based on the IEEE 33 node system and actual distribution network cases involving photovoltaic and wind power, this paper compares the performance of the proposed algorithm with traditional methods. Experiments have shown that compared with PSO, GA, and differential evolution algorithm (DE), APSO reduces system network loss by 12.7% -18.3%, increases voltage qualification rate to 99.6%, and accelerates convergence speed by more than 30%. Even in high proportion scenarios where the penetration rate of new energy reaches 30%, this method can still reduce network loss by 28.7% and effectively smooth out voltage disturbances caused by fluctuations in new energy. This study provides an effective solution for reactive power coordination optimization in distribution networks under high proportion of new energy integration.
Tiantian Song (Thu,) studied this question.