Research And Application Of Model Identification For Main Steam System Of Thermal Power Plant

While the thermal power generating units are developing in the direction of larger capacity and higher parameters, the increase of single-machine capacity and the improvement of steam parameters, require the system with higher reliability and automation level, which makes the status of the automatic thermal control much more important in large thermal power plant units, becoming a reliable assurance of the safety and economic operation in the large units. The main steam temperature system of thermal power plant units has the characteristics such as nonlinear, time-varying, large delay and large inertia, which is always one of the difficult problems for thermal power plant control, and the main steam temperature control strategy based on DCS systems has a wide range of consistency. Consequently, this article selects the main steam temperature control system as a research object.Analyzing the dynamic characteristics of main steam temperature control system, the recursive least squares method and the model transformation method are combined to apply to closed-loop identification of the cascade control system, the controller parameters can be set by automatic optimization, while ITAE error performance index is made for the objective function. It can reduce the workload of the controller parameter tuning, further improve the accuracy of the control parameters, and also can overcome the reducing of the system performance cased by environmental change or disturbance action.A solution is proposed about the pretreatment for feedforward loop and the feedforward controller design in the closed-loop identification of the main steam temperature control system. It improves the feedforward control engineering tuning, it's easy to calculate parameters of the feedforward controller, which has good robustness, and can reduce the dynamic error effectively.A field test proposal is developed for model acquisition in main steam temperature system, which is consistent with the actual power plant operation. A software is developed based on DCS actual project for the model identification and optimization of the main steam temperature control system. The results of the study are applied to the practical 600MW units, completing model acquisition, controller parameters optimization and feedforward controller design of the main steam temperature control system, which improves the performance of the system control significantly.