Research Of Chaos Genetic Algorithm On AGC Units Deployment

The most important control operation of energy management system (EMS) is automatic generation control (AGC), and it's also an important ancillary service in electric power market. In planned economy pattern, the problem of AGC participating in the frequency and power control was only one simple dispatch issue in ignorance of the economy factor. Since power market has built up, dispersive decisions replace the centralized regulation and the security of the grid is influenced. So only economic methods can inspirit power plants to supply the ancillary service. First, as a merchandise, AGC ancillary service is no longer free, grid company should pay for the service to the plants who supply it. Second, power plants must compete in AGC ancillary service market and win the pact by high quality and preferential price. In other words, AGC has changed to unfree service which is compensated with money or even focus on price competition.At present, AGC ancillary service has became a hot research spot. The problem of AGC units dispatch in generation market is addressed theoretically and practically in this paper. The purpose of this thesis is try to use CGA in units selection, and the calculating result proves that CGA is a advanced algorithm adapting AGC units selection.First of all, this paper summarizes AGC ancillary service in power market and introduces the actuality and development of AGC ancillary service in and out of China. Next, we present a model on the deployment of AGC units and give a particular analysis. Strict conditions are the keys of resolving AGC units selection, so we introduce some methods to resolve these two strict values. Then, we use conventional GA and CGA on the same units in order to gain different optimal selections of the units. The results prove that CGA is not only applicable to the AGC units deployment, but also with higher efficiency and lower costs. Under the premise of affirmation on CGA, we add the constraint condition of network flows into the model, and simulate the optimization dispatch. Finally, we give a conclusion of this paper, and point out some topics for the further research on the deployment of AGC units.