Research On Economic Dispatch Of Electric System Contained Wind Farms Based On Coevolutionary Algorithm

The key of economic dispatch of electric system is to distribute the load efficiently in order to minimize the whole cost of the system. The traditional economic dispatch problems make a little effect on the power system since these problems do not involve the new energy such as wind power、solar energy. And the output power changes slightly. The wind industry has been developed rapidly in China recently, while the uncertainty and volatility of the wind make the economic dispatch problems become much more complicated than before. In view of the economic dispatch problems of electric system contained wind farm, the main work was carried out as follows.Firstly, the effects of large-scale wind power integration on the economic dispatch of the electric system were studied in this thesis. Based on the net economic value that the wind farms produce for electric system, how the value is changing with different wind farm configurations, wind power prediction, wind power penetration was given in this thesis. According to that, the effects that wind farms on the economic dispatch and the whole cost of the electric system were studied.Secondly, the traditional economic dispatch of electric system was analyzed. Based on the empirical mode decomposition, the time series method was used to forecast the wind power output and improve the prediction accuracy of wind power output. The mathematical model of economic dispatch for electric systems contained wind farms was established. And the environmental factors and the extra cost of spinning reserve caused by wind farms were considered.Thirdly, the traditional multi-objective coevolutionary algorithm(MOCEA) was improved. Three cooperative operators and a Pareto crossover operator are designed. And the crowded distance was also introduced to reduce the size of the external set. Three standard functions were used to test the performance of the algorithm, which includes the common characters of multi-objective optimization problems. The results of MOCEA were compared with the NSGA to verify its feasibility.Finally, the improved MOCEA algorithm was used to solve the established mathematic model. The particle swarm optimization was used to solve the single objective model. The results of single objective optimization was analysed and its weakness was explained.The improved MOCEA based on the biology system was used to solve the multi-objective model, which take the two objectives into consideration. The increased cost caused by the wind farm was added into the whole cost of the electrical system. The parameters and steps were designed. The results of the MOCEA were compared with the results of single objective optimization and the NSGA-II. The simulation results show that the proposed model and the algorithm are feasible. The results of electrical system with or without wind farms were also compared to verify the value of wind energy.