Besearch On The Coordinated Control System Of The 300MW Unit

This paper researches on the coordinated control system (CCS) of the turbine-generator unit. It sums up the existing problems and analysis the factors that affect control performances of the CCS.The main task of CCS is to satisfy the load instruction requirement of the electric network, to reduce disturbances in different forms, and to ensure the stability of the unit at the same time. Instead of one-degree-of-freedom PID controller, two-degree -of-freedom PID controller is adopted as the main controller of the boiler and turbine, and the latter can somewhat reduce the dependency of CCS on the feed forward loop of the boiler.Benefit from global optimal research ability of Genetic Algorithms (GA), a comprehensive fitness function is constructed to evaluate the control performance of the whole system and it is used to search 12 optimal parameters for boiler and turbine controller of the CCS. The simulation results has shown the two-degree-of-freedom PID controller got by GA has better performance and practicability than the common controller which is tuned by traditional methods.To improve the large-time delay characteristics of the boiler, the rule-based fuzzy predictor is used to predict the process output. The fuzzy predict strategy successfully predict the values of the main parameters and their changing trends in the future by computing them and their changing tendency that have occurred in the system. The simulation results have shown that the system overshot can thus be effectively reduced.Fuzzy control strategy is used to adjust the instructions of boiler and turbine intellectually. The controller is then tuned by the instructions to keep up with the changing situations of the units in different load phases and working conditions. This kind of control scheme facilitates the global optimization of CCS.Fuzzy control is introduced into the feed forward loop of the boiler. According to the static and dynamic energy distribution principle and by applying prediction strategy, the dynamic energy distribution is reasonably adjusted to satisfy the different energy requirement in the different phases, it also to reduce the energy accumulation and deficiency caused by the integralaction of the controller. All these improved the respond rate of the load and the control performance of system.