The planning of distribution network containing distributed generators based on mixed integer linear programming

Currently, the methods for solving the planning of distribution network (DN) are meta-heuristic algorithms whose computational efficiency is low and the results are not accurate. Therefore, this paper uses the advanced mixed integer linear programming (MILP) technology to model the multi-conductor DN planning containing the allocation of distributed generator (DG). Firstly, replace power flow calculation with an equivalent cone programming, then remove the product terms of binary and continuous variables, and finally use polyhedral models which are linear to approximate the planning model. In addition, as the siting and sizing of DG are added to the network planning, the situation that part of the loads are supplied by the DG separately perhaps appear thus leading to the appearance of loops. So the spanning tree (ST) constraints are introduced to the model to ensure the radiation of the DN. Finally, 23-node distribution system is used to simulate the above-mentioned model, and the results show the effectiveness and efficiency of the proposed model.