Optimal reactive power dispatch using Smell Agent Optimization for transmission loss minimization considering multi-contingencies
Abstract
Optimal reactive power dispatch (ORPD) is a non-linear, non-continuous optimization problem that is central to secure and economic power system operation. This paper applies the Smell Agent Optimization (SAO) algorithm, a recently proposed nature-inspired metaheuristic based on the phenomenon of olfaction, to solve the ORPD problem while explicitly accounting for multi-contingency conditions. This appears to be one of the earliest applications of SAO to the ORPD problem under explicit multi-contingency conditions. The control variables optimized are generator terminal voltages, transformer tap ratios, and shunt capacitor sizes. The proposed approach is implemented and tested on the IEEE 30-bus reliability test system with 13 control variables using MATLAB R2017a and MATPOWER 5.1. Weak and secure buses are first identified through the Static Voltage Stability Index, and generator and line contingencies are ranked using the same index to select representative severe outages. A parametric study on the number of moles and iterations is conducted to establish a consistent SAO configuration (50 moles, 150 iterations), which is then used to minimize transmission loss under three scenarios: without outage, with a single line/generator outage (N-1), and with combined line-and-generator multi-contingency (N-2). At bus 26, SAO reduced transmission loss from 11.6957 MW to 6.9554 MW (40.5309%) at 25 MVAr loading without contingency, and from 13.9270 MW to 8.5120 MW under the combined multi-contingency case at the same loading, while the 13 control variables satisfied their equality and inequality constraints in every run; full post-optimization feasibility diagnostics are reported in Section 3.6. SAO reduces transmission loss consistently in every scenario tested, with and without contingencies, while every control variable stays within its operating limits.