Abstract
This paper aims to propose a new multiobjective algorithm for multiobjective distributed network reconfiguration (DNR) with the placements of distributed generation (DG) in radial distribution networks (RDNs). The new proposed algorithm, called the nondominated sorting stochastic fractal search (NSSFS), is a new multiobjective version of the original SFS algorithm. NSSFS incorporated fast nondominated sorting strategies, crowding distance computation, and selection mechanism into SFS to find and maintain the best nondominated solutions. The proposed NSSFS algorithm was tested with eight multiobjective benchmark test functions to validate its performance. The NSSFS was then implemented to define the optimal network configuration, positions, and sizes of DG units in the RDNs, where real power loss, voltage profile, and voltage stability index were optimized simultaneously. The implementation of multiobjective DNR-DG (MODNR-DG) significantly enhanced the performance of the system. Based on the comparison outcomes, the NSSFS algorithm obtained better solution quality than other multiobjective techniques, proving the effectiveness of NSSFS in dealing with the MODNR-DG problem.
Highlights
Academic Editor: Chien Ming Chen is paper aims to propose a new multiobjective algorithm for multiobjective distributed network reconfiguration (DNR) with the placements of distributed generation (DG) in radial distribution networks (RDNs). e new proposed algorithm, called the nondominated sorting stochastic fractal search (NSSFS), is a new multiobjective version of the original SFS algorithm
NSSFS incorporated fast nondominated sorting strategies, crowding distance computation, and selection mechanism into SFS to find and maintain the best nondominated solutions. e proposed NSSFS algorithm was tested with eight multiobjective benchmark test functions to validate its performance. e NSSFS was implemented to define the optimal network configuration, positions, and sizes of DG units in the RDNs, where real power loss, voltage profile, and voltage stability index were optimized simultaneously. e implementation of multiobjective DNR-DG (MODNR-DG) significantly enhanced the performance of the system
In Ref. [10], Fireworks Algorithm was applied for voltage profile improvement and power loss minimization. e normalized objectives were formulated as the primary objective, which could not acquire the best solution
Summary
Compute objective function values according to equation (26) for all particles of population. Perform the updating process as follows: All solutions are assigned the value of Pai according to equation (17). E generational distance (GD) metric measures the convergence of Pareto optimal front generated (PFg) by a multiobjective algorithm [42] as follows:. E spacing (SP) metric measures the distribution of solutions in PFg as follows [43]:. A lower value for the SP metric denotes better distribution of a multiobjective algorithm. A lower value for the Δ metric shows a better spread of a multiobjective algorithm. E hypervolume (HV) metric measures the volume covered by solutions in PFg in the objective space for a multiobjective problem with two objectives. A multiobjective algorithm with the highest HV metric obtains a better nondominated set
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
