Taguchi-Based Parameter Tuning of PSO for Optimal Capacitor Placement in Unbalanced Distribution Systems

This paper presents a Taguchi-tuned Particle Swarm Optimization (PSO) approach for the optimal placement and sizing of shunt capacitor banks (CBs) in unbalanced distribution systems. The optimization aims to minimize the total operational cost by reducing power losses and improving voltage profile. A systematic parameter tuning was carried out using the Taguchi method based on an L25 orthogonal array, with five PSO parameters evaluated through Signal-to-Noise (SN) ratios and Analysis of Variance (ANOVA). The IEEE 13-bus test feeder was used as a benchmark. The results show that the installation of four optimally placed CBs reduces active power losses by 29.6% (from 133.16 kW to 93.71 kW), improves the minimum bus voltage from 0.942 p.u. to 1.014 p.u., and decreases operating costs by 6,144.55$ compared to the base case. Validation using DIgSILENT PowerFactory confirms the consistency of the proposed method. Moreover, the Taguchi-optimized PSO demonstrated superior performance over the classical PSO in terms of convergence speed, solution quality, and result consistency across multiple independent runs, confirming its effectiveness and robustness for practical distribution system optimization.