A comprehensive analysis for multi-objective distributed generations and capacitor banks placement in radial distribution networks using hybrid neural network algorithm

Abstract

This paper proposes a new methodology based on the combination of symbiosis organism search (SOS) and neural network algorithm (NNA), named SOS-NNA, for the optimal planning and operation of distributed generations (DGs) and capacitor banks (CBs) in the radial distribution networks (RDNs) considering single- and multi-objective optimization with various equality and inequality constraints. The multi-objective framework is a weighted combination of five component objectives including active power loss, voltage deviation, voltage stability, load balancing, and supply reliability. In addition, practical voltage-dependent non-linear load models are also examined. Two benchmark instances 33 and 69-bus networks have been utilized to evaluate the effectiveness and feasibility of the proposed SOS-NNA via various case studies. The obtained outcomes for different operating cases and test scenarios reveal that the proper combination of optimal power factor DGs (OPF-DGs) and CBs can boost the network performance indexes to an ever-highest degree for all test networks. A cost–benefit analysis is further implemented to evaluate the economic feasibility of the obtained multi-objective solutions. As a result, the proposed SOS-NNA shows a marked improvement regarding the solution quality compared to the recently well-established optimization algorithms as well as outweighs the original NNA in the performance indexes of the solution quality, convergence speed, and statistical results. In addition, the proposed SOS-NNA has been employed for allocating different DG types in RDNs with the consideration of actual 24-h load profiles and the obtained outcomes contribute to the further improvement of yearly energy loss mitigation as well as cost savings.