The Application Of Automatic Control Theory In Thermal Process Automation Of Power Plant

With the combination of automatic control theory and practical thermal process automation object in the power station, three main fields including nonlinear calibration of thermal instrument, new control strategies of main steam pressure and advanced control strategies of main steam temperature are studied in detail in this paper.It is well known that the relationship between the input value and output value of thermal instruments in power station is nonlinear. In order to improve measurement precision and display fidelity of the instrument, three new methods of nonlinear calibration of thermal instruments, which are based on intelligent control theory, are presented in this paper, such as nonlinear compensation of Zr02 oxygen measurement instrument using BP NN, nonlinear calibration of temperature measurement sensors using CMAC NN and nonlinear identification of throttle flow meter using GA. These methods prove to be not only simple but also effective. At the same time, the application of these methods in computer monitoring system of the power station is described in this field.Main steam pressure is a very important parameter in power station. However, its character of big-lagged and big inertial makes it an object can' t be easily regulated. Utilizing the radiation signal of the furnace and fuzzy filter, the cascade fuzzy control strategy is advanced to enhance its regulated quality. The radiation signal is taken as an intermediate regulated variable in this strategy. Using the approximate model of main steam pressure and linear quadratic optimal control theory, the LQ sub-optimal control strategy is brought forward. By the way, simulation results demonstrate that two control strategies work well. The first one is fit for steady-pressure unit and another one is suitable for changeable-pressure unit.Main steam temperature is as important as main steam pressure in power plant and its control requirement is very high. According to optimal control theory and state feedback, the optimal feedback control strategy is designed and implemented in order to satisfy the operation requirement of security and economy. Grounded on CMAC learning controller and cascade control system, the cascade CMAC learning control strategy, characterized in good adaptabil ity, is presented, and which can be applied in computer real-time control. Simulation resultstestify that all these strategies are much better than PID.