A stochastic mixed-integer conic programming model for distribution system expansion planning considering wind generation

Abstract

This paper presents a stochastic scenario-based approach to finding an efficient plan for the electrical power distribution systems. In this paper the stochasticity for the distribution system expansion planning (DSEP) problem refers to the loads and wind speed behavior. The proposed DSEP model consist the expansion and/or construction of new substations, installation of new primary feeders and/or reinforcement the existing, installation of wind-distributed generation based, reconfiguration of existing network, and the proposed DSEP is solved considering uncertainty in electric demand and distributed generation. In this regard, a two-stage stochastic programming model is used, wherein the first stage the investment decision is made and the second stage calculates the expected operating value which depends on the stochastic scenarios. The mathematical approach is based on a mixed integer conic programming (MICP) model. By using this MICP model and a commercial optimization solver, finding the optimal global solution is guaranteed. Moreover, in this paper by using the Tabu Search algorithm and take the advantages of a stochastic conic optimal power flow model, an efficient hybrid algorithm is developed. With the aim of comparing the performance of the optimization techniques based on solution of MICP model directly and using a hybrid proposed methodology, they are tested in a 24-node distribution system and the results are compared in detail.