MISOCP-BASED OPTIMAL DG LOCATION AND SIZING CONSIDERING ZIP LOAD MODEL IN POWER DISTRIBUTION SYSTEMS

Abstract

The increasing penetration of Distributed Generation (DG) is likely to result in many difficulties in planning and operating power distribution systems. One of these
challenges is to determine the DG’s optimal location and capacity. This paper presents a mathematical model of optimally allocating DG with the aim of minimizing
the total DG’s investment cost, DG’s operation cost and the power exchange cost with the upper-level transmission network. Constraints considered in this
optimization model are power flow equations, DG’s power output limits, voltage magnitude bounds, thermal limits of branches, the number of installed DGs, and the
number of buses for DG placement. In addition, this optimization formulation consists of modelling the voltage-dependent loads (ZIP load modelling). Since this
optimal problem is denoted as a Mixed-Integer Non-Linear Programming (MINLP) model, the solution to be found can not be globally optimal. Therefore, this paper
deploys methods to convert the MINLP model into a Mixed-Integer Second-Order Cone Programming model (MISOCP). The globally optimal solution of the MISOCPbased model can be obtained by commercial solvers such as CPLEX/GAMS. An IEEE 15-bus system is deployed to evaluate the proposed model. The calculation results
reveal that the optimal placement, sizing, generating output of DG and voltage profile in the distribution system are contingent on the DG cost function and the power
exchange price with the transmission grid.