Multi-objective planning of electrical distribution systems using dynamic programming

Abstract

This paper presents a novel dynamic programming approach for multi-objective planning of electrical distribution systems. In this planning, the optimal feeder routes and branch conductor sizes of a distribution system are determined by simultaneous optimization of cost and reliability. The multiple planning objectives are minimization of: (i) installation and operational cost, and (ii) interruption cost. The first objective function consists of the installation cost of new feeder branches and substations, maintenance cost of the existing and new feeder branches, and the cost of energy losses. The second objective function measures the reliability of the distribution network in terms of the associated interruption costs for all the branches, which includes the cost of non-delivered energy, cost of repair, and the customer damage cost due to interruptions. A dynamic programming based planning algorithm for optimization of the feeder routes and branch conductor sizes is proposed. A set of Pareto solutions is obtained using a weighted aggregation of the two objectives with different weight settings. The proposed approach is evaluated on 21-, 54-, and 100-node distribution systems. The simulation test results are analyzed with various case studies and are compared with those of two existing planning approaches based on multi-objective evolutionary algorithm. � 2012 Elsevier Ltd. All rights reserved.