Selection of low-voltage side current-limiting reactor for interconnection transformers in hydropower station based on particle swarm optimization algorithm
Abstract：To reasonably select the reactance percentage for the low-voltage side current-limiting reactor of the interconnection transformer in hydropower station, a strategy based on the particle swarm optimization algorithm is proposed to effectively reduce short-circuit current levels while ensuring the sensitivity of transformer differential protection. A linearly decreasing inertia weight is used to dynamically balance the algorithm’s global exploration and local exploitation capabilities, preventing premature convergence. An optimization model that takes the reactance percentage as the decision variable and differential protection sensitivity as the constraint is constructed, and a fitness function that integrates the constraint conditions is designed, directly embedding the engineering constraints into the optimizing process of particle swarm. A function model is established in Matlab for simulation and compared with the traditional back-calculation method. Results show that this proposed method has global optimization capability, which can quickly obtain the optimal reactance percentage, and intuitively compare the effects of different reactance percentages on transformer differential protection sensitivity and short-circuit current magnitude. Further research verifies the dynamic and thermal stability of the selected current-limiting reactor, assesses its impact on transformer backup protection, low-voltage switchgear breaking capacity, and the accuracy class of current transformers, confirming the rationality and feasibility of the new parameters in engineering applications.
汪平, 杨令. 基于粒子群优化算法的水电站联络变压器低压侧限流电抗器选型分析[J]. 电气技术, 2026, 27(8): 44-52.
WANG Ping, YANG Ling. Selection of low-voltage side current-limiting reactor for interconnection transformers in hydropower station based on particle swarm optimization algorithm. Electrical Engineering, 2026, 27(8): 44-52.
2026, Vol. 27 
