Contribution à l’optimisation de l’écoulement de puissance par les méthodes d’intelligence artificielle améliorées

Abstract

Optimal Reactive Power Dispatch (ORPD) is an important task for achieving more economical, secure and stable state of the electrical power system. It is expressed as a complex optimization problem that is stated to identify the optimal control variables of various regulating equipments for minimizing an objective function under constraints. Many meta-heuristic techniques have been proposed to overcome various complexities in solving ORPD problem, which are characterized by exploration and exploitation of the search mechanism. The balance between these two characteristics is a challenging problem to attain the best solution quality. The Artificial Bee Colony (ABC) algorithm as a reputed meta-heuristic has proved its goodness at exploration and weakness at exploitation where the enhancement of the basic ABC version becomes necessary. Salp Swarm Algorithm (SSA) is a II newly developed swarm-based meta-heuristic, which has the best local search capability by using the best global solution in each iteration to discover promising solutions. In this research, a novel hybrid approach based on ABC and SSA algorithms (ABC-SSA) is developed and applied for solving ORPD problem. The proposed approach tries to enhance the exploitation capability of the ABC algorithm using SSA. The efficiency of ABC-SSA is investigated using four standard test systems IEEE 30bus, IEEE 57bus, IEEE 118bus and large-scale IEEE 300bus and that by considering the famous objective functions in ORPD problem including total transmission active power losses (Ploss), Total Voltage Deviation (TVD) from the rated voltage magnitude and the Voltage Stability Index (VSI) of load buses. The reached simulation results have proved that the proposed ABC-SSA is more efficient overcoming ABC, SSA and other recently developed meta-heuristic optimization techniques in the literature.