Application Research On Optimization And Control For Boiler Steam Pressure Of Power Station

Boiler-steam turbine system is a typical multi-input multi-output system. In engineering practices, control design methods based on single-input single-output or local linearization are always used.In this paper, we consider the control of boiler steam pressure in the boiler-steam turbine system. In the controller design procedure, problems including overcoming of system disturbances and control input constraints, measurement of differential signal and time delay of boiler steam pressure, analysis of uncertainties and modeling are studied, which is of profound theoretical significance and practical engineering value. Using data mining technology, massive historical data of power stations which reflects the equipment running status is analyzed, and the deep process and integration of the power station information are achieved, which has important practical significance on reducing consumption of the whole electrical industry.The main contributions of this dissertation are summarized as follows:(1) A new method called dynamic surface control is proposed for fuel-steam pressure system. System states are estimated by constructing a state observer, and dynamic surface control method is applied to overcome the problem of "explosion of terms" existing in traditional backstepping. Since only boiler steam pressure signal in the control system is needed, the method has good practical value. Lyapunov stability analysis method proves that the closed-loop system is semi-globally uniformly stable and the tracking error can be made arbitrarily small.(2) In order to improve the performance of sliding mode control, sliding mode control method is combined with different kinds of observers. By designing disturbance observer, effective compensation of the system disturbance is achieved and the capability of disturbance rejection is improved. By designing extended high gain observer, the uncertainties and disturbances are all observed. By designing time-delay observer, the problem of output measurement delay is solved. From these observers, high-precision performance of the fuel-steam pressure system is realized and the robustness performance of the system is improved.(3) RBF neural network control method is proposed for boiler steam pressure of fuel-steam pressure system. Combined with differentiator of the signal, sliding mode controller is designed without model information and speed signal. Differentiator is utilized to estimate system states and the observation results are served as the inputs of RBF neural network. Considering control input constrain, RBF network is used to compensate the controller signal and a sliding mode controller is designed for the anti-windup closed-loop system. (4) The operation conditions are classified by using history data of thermal power unit, and the improved K-means method is used to get the K-clusters of data considering the specific operation conditions. Finally, fuzzy association rule is used to get the linearization balance operating point from K-clusters, which provides basis for the design of energy-saving multi-model coordinated control system.