Research On Single-shaft Gas Turbines' Dynamic Model Of Combined Cycle Power Station

With the recognition of environmental problems and the further utilization of the natural gas in the world, combined cycle (CC) power station will occupy more and more important position in the power production field. Single-shaft gas turbine is the key equipment of combined cycle power station, on which modeling research has much more meaning. Based on the dynamic characteristics analysis of single-shaft gas turbine, this paper introduces some modeling methods of gas turbine and analyzes each advantage and disadvantage. On account of that the turbine is a complicated dynamic system with strong nonlinear, strong coupling and frequent disturbance, it is difficult to build its accurate dynamic model with the traditional methods. However neural network has the characters of strong nonlinear-study ability, self-study ability and self-organize ability, which can build up a much more accurate model of gas turbine.First of all, this paper proposes a linear gas turbine model with the least-squares method. Secondly dynamic models of single-shaft gas turbine built by BP neural network and RBF neural network respectively are established, and compared with the linear model. The simulation results show the validity.