Research On Boiler-turbine Unit Coordinated Control Technology Of Marine Power Plant

Marine power plant usually adopts operation way of "Boiler with turbine" ( thermal output of boiler will change with the load of turbine). With the problem of time-delay and large-inertia to the boiler, it's difficult to get the energy-equivalence between the quicker control link for the steamer and the slower one for the boiler. It is important for both to be coordinated better during the operation of the boiler and the turbine, especially off-design working condition. On this background, the dissertation pays attention to study dynamic characteristics of marine boiler-turbine unit.This paper comes up with the mathematical model of marine boiler-turbine unit. Based on utilizing related designing data, we can simulate dynamically several transient processes, such as increasing process, decreasing process, full astern etc. In this paper dynamic change of each important variable is monitored in these processes, and some dynamic simulation results of some important issues are given such as the mass flow of the fuel, entrance pressure of the turbine, rotate speed. Based on these, the results of simulation are analyzed; the trend of dynamic change meet the demand.This paper focus on decoupling control based on Neural Network. Direct against the strong coupling characteristic of control system of coordinated control system in the marine stereotypical power plant, a decoupling control system composed of self-adaptive PID controller and dispersive decoupling network is presented which is applied in control system of coordinated control system to decouple this system.The decoupling control system based on PID Neural Network and the coordinated control system with direct-energy-equivalence are designed. Under the two control systems, dynamic simulations of accelerating process and moderating process for the ship are done. The simulation results of the two different control systems for the boiler-turbine unit, under different typical load disturbances, are carried on the comparison. The result shows that the decoupling control system based on Neural Network achieves better tracing.