A Hybrid JAYA and Teaching-Learning Based Optimization Algorithm for Single- and Multi-Objective Optimal Power Flow

Abstract

This study presents a novel multi-objective optimization framework for the Optimal Power Flow (OPF) problem in an IEEE 30-bus system, aiming to simultaneously minimize generation costs, active power losses, and voltage deviations. Given the highly nonlinear and heavily constrained nature of OPF, conventional optimization techniques often fall short. To overcome these computational challenges, we propose a parameter-less hybrid metaheuristic approach that integrates Teaching-Learning-Based Optimization (TLBO) with the JAYA algorithm. This hybridization strategically synergizes TLBO’s global exploration capabilities with JAYA’s rapid, convergence-driven exploitation. By employing Pareto optimization and rigorously evaluating the outcomes using quantitative performance indicators (Spacing, Maximum Spread, and Hypervolume), a well-balanced and statistically robust set of optimal solutions is derived. Furthermore, the practical robustness of the framework is explicitly validated under a severe smart-grid stress scenario featuring high Electric Vehicle (EV) charging penetration. The comparative results reveal that the proposed hybrid TLBO-JAYA algorithm significantly enhances solution quality, convergence reliability, and overall system robustness compared to Particle Swarm Optimization (PSO) and standalone baselines. While the hybrid approach incurs a computational-time penalty due to its sequential phases, this research ultimately establishes the method as a highly effective tool for complex, multi-objective power system operational problems.