Comparative Analysis of Particle Swarm Optimization and Salp Swarm Algorithm in improving the Voltage Profile of IEEE 9 Bus System

Abstract

Voltage stability is a critical aspect of modern power system operation, directly influencing system
reliability, operational efficiency, and protection against grid collapse. Voltage instability often
arises due to heavy reactive power demand, insufficient local reactive compensation, and weak
transmission infrastructure, especially under high loading or fault conditions. To address these
challenges, advanced optimization techniques are often employed for reactive power planning and
voltage profile enhancement. Hence, a comparative analysis of two nature-inspired metaheuristic
algorithms—Particle Swarm Optimization (PSO) and Salp Swarm Algorithm (SSA)—in mitigating
voltage instability of the IEEE 9-Bus test system is done. The aim is to compute the optimal
placement and sizing of reactive power compensation using shunt capacitors. The objective function
is formulated to minimize overall voltage deviation from the nominal value (1 p.u.), and ensure
computational efficiency of the algorithms. Both algorithms were evaluated on three main
performance metrics: voltage loss minimization, convergence characteristics, and recommended
capacitor sizes (MVAr ratings). PSO algorithm showed faster convergence towards near-optimal
solutions, making it suitable for time-constrained applications. On the other hand, SSA showed
superior performance in recommending more economical MVAr injection sizes. But SSA achieving
comparable voltage profile improvements, resulted with slightly higher iteration counts. Original
contribution of this work lies in the direct comparison of PSO and SSA for increasing voltage
stability under the same system conditions, offering insights into their relative strengths. The
significance of this study is its potential to guide power system engineers in selecting appropriate
optimization methods based on computational and economic trade-offs. Future research can extend
this approach to meshed and large-scale systems, integrating uncertainty in load demand and
renewable generation.