Research On Distribution Network Reconfiguration To Enhance Consumption Of Diverse Renewable Energy

With the depletion of traditional fossil fuels, renewable energy like wind and solar energy develops quickly, and worldwide installed capacity continues to grow. However, due to the limitation in transmission capacity and leveling capacity of the grid, only a limited portion of diverse renewable energy is consumed by the grid, leading to a phenomenon where renewable energy can generate but cannot connect to the grid, indicating that massive energy is wasted. Hence, it is necessary to dive deeply into the research on the enhance ment of the consumption of diverse renewable energy. Base on the background above, this dissertation disscusses the feasiblity to enhance the consumption of renewable energy from the perspective of distribution network reconfiguration.This dissertation has finished the following work:Firstly, the current status of research on distribution network reconfiguration and consumption of renewable energy in distribution network is detailed in this dissertation.Secondly, the stochastic models of renewable energy such as wind and solar energy as well as load are established on basis of statistics.Thirdly, a static distribution network reconfiguration model aiming at minimization in mathematical expectation of overall power loss is proposed, and constraints in power flow, voltage, topology are considered. To improve computational effic iency, undirected spanning tree-based parallel genetic algorithm(PSTGA) was devised.Fourthly, a sampling method concerning correlation between wind velocity and solar radiation respectively is proposed, and impact of the correlation on distribution network reconfiguration is examined in terms of states of switches, nodal voltage as well as overall power loss.Fifthly, a bi-level dynamic distribution network reconfiguration model is proposed, whose objective function is form by weighted sum of the operational expense and the reject rate of renewable energy. Revised undirected spanning tree-based parallel genetic algorithm and trust-region based augmented lagrangian method are used in solution of the upper and lower level problem respectively.IEEE 33 distribution system is employed to test the model and algorithm proposed above in the analysis of distribution network reconfiguration. The validity of the model as well as the performance of the algorithm are well verified by IEEE 33 distribution system.