Reactive Power Optimization in Power Systems Using Hybrid GA/SA/TS Algorithms

In this paper addressing reactive power optimization in power systems, we implemented a hybrid optimization methodology integrating three distinct algorithms: Genetic Algorithm (GA), Simulated Annealing (SA), and Tabu Search (TS) to improve system efficiency and performance. The implementation workflow involves the following stages:

First, the Genetic Algorithm (GA) is employed for initial system optimization, where population-based selection, crossover, and mutation operations explore the solution space to identify near-optimal solutions. In code implementation, chromosome encoding typically represents control variables like transformer tap positions and shunt compensator values. Following this, the Simulated Annealing (SA) algorithm refines GA results by applying a temperature-controlled stochastic acceptance criterion, preventing premature convergence to local optima through controlled probability-based uphill moves. Finally, Tabu Search (TS) enhances optimization further using memory structures (tabu lists) to avoid revisiting recent solutions while incorporating aspiration criteria to override tabu status when superior solutions emerge.

The synergistic combination of these three algorithms not only significantly improves optimization effectiveness and system performance but also effectively addresses various challenges in power systems, including voltage stability enhancement and loss minimization. Consequently, this methodology demonstrates broad application prospects and can make substantial contributions to power system optimization and development.