Multi-objective Expansion Planning of Electrical Distribution Networks Using Comprehensive Learning Particle Swarm Optimization

In this paper, a Pareto-based multi-objective optimization algorithm using Comprehensive Learning Particle Swarm Optimization (CLPSO) is proposed for expansion planning of electrical distribution networks. The two conflicting objectives are: installation and operational cost, and fault/failure cost. A novel cost-biased particle encoding/decoding scheme, along with heuristics-based conductor size selection, for CLPSO is proposed to obtain optimum network topology. Simultaneous optimization of network topology, reserve-branch installation and conductor sizes are the key features of the proposed algorithm. A set of non-dominated solutions, capable of providing the utility with enough design choices, can be obtained by this planning algorithm. Results on a practical power system are presented along with statistical hypothesis tests to validate the proposed algorithm.