Interior point method based reactive power optimization of active distribution network

Abstract

High permeability distributed generation (DG) is accessed to active distribution network(ADN), and it has a great influence on the power flow, node voltage and network loss of distribution network. In distribution network, DG has a certain ability of reactive power regulation. Combined with the traditional voltage regulation method, the power flow of the active distribution network is optimized. According to the active power prediction of DG, the reactive power output of DG can be adjusted dynamically to maintain the local voltage stability and achieve the optimal allocation. Firstly, in this paper, the relationship between active power and reactive power of wind power, photovoltaic and other DG is studied. The objective function is constructed with total network loss and the penalty function of discrete variable. With the voltage and reactive power regulation capacity as the constraint, Distribution network reactive power optimization is established as a mixed integer optimization problem. The method of combining interior point method and penalty function was used to deal with the discrete variables. The continuous differentiable Gauss penalty function is introduced into the objective function, and it could change the mixed integer optimization problem into a continuous problem to solve. A calculation example of an ADN with photovoltaic and wind power shows that, reactive power optimization can reduce network loss, and effectively improve the performance of active distribution network.