Distribution Network Reconfiguration With Distributed Generation Based On Short-term Load Forecasting

In recent years, the smart grid is considered as the latest trends and development model of global power system in the 21st century. Due to historical reasons, distribution automation is at a lower level, nearly half of the net loss and more than 95%of the outage time generated in the distribution network, therefore distribution network reconfiguration is so necessary in order to promote better and faster smart distribution grid construction. Short-term load forecasting can provide the necessary data for state estimation and flow calculation. At the same time, as more and more distributed generation (DG) incorporated into the grid, it will have a huge impact on distribution network. In summary, this thesis made research on distribution network reconfiguration including DG based on short-term load forecasting under the smart distribution network framework. The main work and research results are as follows:1. This thesis used the least square support vector machines(LS-SVM) to predict IEEE 33 node system load and used particle swarm algorithm to optimize key parameters automatically, making forecasting results more accurate.2. Study of the distribution network dynamic reconfiguration based on the predicted load. In this thesis, dynamic reconfiguration problem is decomposed into multiple static reconfiguration optimization problems. Firstly determined the optimization function, then made topology recognition, analyse system constraints, wrote flow calculation program, finally gave dynamic reconfiguration results using comprehensive learning quantum-inspired evolutionary algorithm(CLQIEA).3. Reaearch on the distribution network reconfiguration including DG based on the predicted load. Firstly, analysed the impact of DG on power loss and voltage from theoretical and simulation aspects, focusing on different installation location and capacity of DG, then combined with traditional reconfiguration methods, finally gave the results of reconfiguration, which may provide some basis for distribution network optimal operation.