The Research Of Intelligent Scheduling For The Power System Containing Large-scale Renewable Energy

Coal, oil and other fossil fuels are not renewable, and its large-scale development and use has been cause serious damage to the environment. Governments around the world are paid more and more attention to the development of renewable energy. By2015, China’s wind turbine installed capacity will reach100million KW and photovoltaic power installed capacity will reach21million KW. Grid wind power installed capacity of solar power growth rapidly. With the rapid growth of grid connected of wind power and photovoltaic power plant capacity, adverse factors of power grid dispatch is becoming more and more important. Wind power and solar power greatly influenced by environmental factors. With the change of wind speed and light intensity, wind generator’s and photovoltaic array’s output will fluctuate. These features will have a greater impact on the grid and threaten the safety of the grid.This paper deeply studied containing large-scale renewable energy power system scheduling problem. Studied the dispatching model and full understanding the negative of the grid connected large-scale renewable energy. First, a careful study of the power system scheduling model and according to the results of previous studies decomposition the grid scheduling process to units commitment problem and load dispatching problem. Load distribution model and unit commitment model is established, the positive and negative spinning reserve capacity and the positive and negative ramp rate of the power system is taken into account. In order to balance the impact of grid connected renewable energy. In the unit commitment problem, the objective is minimizing the operation costs of thermal power units. For the model of the load dispatching problem, the objective are minimizing the running costs of the unit and minimizing pollutant emissions of thermal power units.Then, for each of the unit commitment problem and load distribution problem in power system scheduling model, use different algorithm to solve the problems in order to find suitable algorithm. During the study, proposed some method to improved algorithm and solve constraints. Through case studies to demonstrate the validity of those proposed method. In order to compare to the algorithm, this paper select two different methods to solve the unit commitment problem and load dispatch problem. MMAS algorithm and LINGO software solution are used in the unit commitment model, analyze the results and the performance of the algorithm. The analysis result showed that the effect of LINGO software to solve unit commitment problem is slightly better than the MMAS algorithm. MMAS algorithm also has advantage of fast solving speed. In the load dispatching model use MOPSO algorithm and the NSGA2algorithm to solve the problem, and then analyzed the calculated results. The results shows, NSGA2solve load dispatch problem of multi-objective optimization problem better. Compare to NSGA2, MOPSO even easy to fall into local optimal solution.Based on the above research results, we developed a software system for the scheduling operation. To establish a database of the scheduling system, can displaying real-time output, scheduling data and other information about each unit. Provide system management, report download and print functions.