Research On The Optimization Of Microgrid Operation Economy Based On Swarm Intelligence Algorithm

As the world’s energy shortage and environment problems are getting more serious, the research and development of renewable energy have become an irreversible trend. Distributed generation (DG) based on renewable energy as the cleanest form of power production, has received widespread attention. Microgrid is an effective carrier of distributed power supply. Its operation mode is favorable to the distributed generation’s control and management, and it can reduce the influences to the power system due to its output power fluctuation. The research on the economic operation of micro-grid has received wide attention of scholars both at home and abroad.Firstly, this thesis discusses the development of distributed power supply and microgrid, reviews the current research works on microgrid and its economic optimization, and states the research significance of this thesis.Secondly, the thesis analyzes the characteristics and the model of distributed power generation, establishes the model of distributed power generation and cost consumption model based on the operating characteristics of photovoltaics, wind turbines, micro gas turbines and fuel cells. Also, the main performance parameters of batteries, the calculation methods of state of charge, depth of discharge and cycle life are studied. The battery cycle life operational cost model is proposed based on the studies of nonlinear battery charge and discharge control strategies.Then, to provide data support for the economic optimization research, the thesis analyzes two different time-scales of photovoltaic power prediction. On the one hand, the thesis completes the short-term photovoltaic power prediction with support vector machine (SVM) algorithm, combined with the meteorological factors; On the other hand, the accuracy of the super short-term photovoltaic power prediction has been improved by selecting a similar day from historical data to establish the changing rate for the next point calculation.Lastly, to meet the needs of microgrid economic optimization, the thesis proposes a multi-time level microgrid economic optimization method. First, the multi-time level within the functional structure optimization method and model related terminology are explained. For each module (day-ahead optimization module, rolling optimization module, super short-term optimization module and real time optimization module) function, the optimization goal and the related constraint conditions are discussed and analyzed in detail. Finally, the improved particle swarm optimization (pso) algorithm is adopted. The proposed optimization method has been tested on simulation platform, and the feasibility was verified.