Artificial Electric Field Algorithm-Pattern Search for Many-Criteria Networks Reconfiguration Considering Power Quality and Energy Not Supplied

Considering different objectives and using powerful optimization methods in the distribution networks reconfiguration by accurately achieving the best network configuration can further improve network performance. In this paper, reconfiguration of radial distribution networks is performed to minimize the power loss, voltage sag, voltage unbalance, and energy not supplied (ENS) of customers using a new intelligent artificial electric field algorithm-pattern search (AEFAPS) method based on the many-criteria optimization approach. The voltage sag and voltage unbalance are defined as power quality indices and the ENS is the reliability index. In this study, the pattern search (PS) algorithm enhances the artificial electric field algorithm’s (AEFA) flexibility search both globally and locally. AEFAPS is applied to determine the decision variables as open switches of the networks considering the objective function and operational constraints. The proposed methodology based on AEFAPS is performed on an unbalanced 33-bus IEEE standard network and a real unbalanced 13-bus network. The reconfiguration problem is implemented in single-criterion and many-criteria optimization approaches to evaluate the proposed methodology’s effectiveness using different algorithms. The single-criterion results demonstrated that some power quality indices might be out of range, while all indices are within the permitted range in the many-criteria optimization approach, proving the effectiveness of the proposed many-criteria reconfiguration with logical compromise between different objectives. The results show that AEFAPS identified the network configuration optimally and different objectives are improved considerably compared to the base network. The results confirmed the superior capability of AEFAPS to obtain better objective values and lower values of losses, voltage sag, voltage unbalance, and ENS compared with conventional AEFA, particle swarm optimization (PSO), and grey wolf optimizer (GWO). Moreover, the better performance of AEFAPS is proved in solving the reconfiguration problem compared with previous studies