Application of the arithmetic optimization algorithm to the phase-balancing problem in three-phase asymmetric distribution networks

Objective: This article presents the application of the arithmetic optimization algorithm (AOA) to the phase balancing problem in three-phase asymmetric distribution systems. The algorithm is implemented in MATLAB and its performance is compared with the losses from the results of the networks presented in specialized literature using other optimization algorithms. The objective is to determine the optimal load connections at each node to reduce the active power losses (the objective function).Methodology: Based on the specialized literature, the IEEE standard test networks with 25 and 37 nodes are selected. A master-slave strategy is proposed, where the iterative swept power flow is programmed in its three-phase version (slave stage). Power losses are characterized under a standard phase connection condition on the load side. The arithmetic optimization algorithm (AOA) (master stage) is then programmed and the most optimal connections to be implemented are defined. These are subsequently compared and analyzed with other methodologies and with the base case.Results: Using the selected IEEE test systems of 25 and 37 nodes from the specialized literature, the proposed master-slave methodology, incorporating AOA and the iterative sweep power flow in its three-phase version, effectively identifies an optimal solution. It determines the connections for the phases that balance the load in the nodes, reducing the losses of these systems by 4.155% and 19.249%, respectively.Conclusions: The study found that the AOA is effective in minimizing power losses in asymmetrical distribution systems, sligthly outperforming the CSA and CBGA optimization methods. The algorithm demonstrated robustness in solution accuracy and proved useful in validating results using standardized IEEE test systems. The findings highlight the importance of implementing efficient optimization techniques. Future work is suggested to explore the application of AOA to other electrical problems and to develop variants that enhance its parameters and reduce processing times.