DYNAMIC MODEL IDENTIFICATION OF ONCE-THROUGH STEAM GENERATION SYSTEM BASED ON PEM METHOD
He Jinliang,Tao Mo,Wang Wei,Song Feifei
DOI: https://doi.org/10.1115/icone25-66555
2017-01-01
Abstract:Once-Through Steam Generator (OTSG) has the advantages of simple structure, good static characteristics, and can produce superheated steam. In recent years it has been widely used in nuclear power system for its good maneuverability. The working fluid flows once through the heat transfer tube forced by the pressure head of the feedwater pump. The water is heated by the coolant of primary side as it flows along the heat transfer tube. After preheating, evaporating and overheating, steam of required temperature is then produced. In addition, flow instability may occur in OTSG because of the present of two-phase flow. Once-through steam generation system is a typical nonlinear multi-variable coupling system, the water/heat storage capacity of secondary side is small, and the steam pressure is very sensitive to the load fluctuation. Various disturbances such as the change of feed water flow, the change of heat transfer rate from the primary side and the change of the secondary side load, will lead to the variety of the parameters in the heating channel. A more complex and precise water control system is therefore needed. An accurate system model is very important to optimize the control strategy and improve the control quality. Obtaining the dynamic mathematical model of the once-through steam generation system is the basis for its effective control. At present, the dynamic modeling of once-through steam generation system mainly adopts mechanism modeling method. By analyzing the inherent mechanism of steam generation process and using the basic conservation equation to derive the relationship between model variables. However, due to the complexity of OTSG and the feedwater system structure and the two-phase heat transfer mechanism, the mechanism modeling is very difficult and the model is not precise enough and very complex. Many nonlinear equations are included, which makes it difficult to determine an effective numerical method for real-time simulation. System identification is based on the actual measurement of input and output information of the process. The system model can be estimated without having to study its internal mechanism. In this paper, the dynamic model of once-through steam generation system in nuclear power plant is identified. Ensuring the stability of the steam outlet pressure during the operation of the system is very important to the safety of steam turbine. Therefore a three-input, two-output coupling system is obtained by analyzing the influence factor of the once-through steam generation system. The pseudo-random sequences are used as the input signal, and the Prediction-Error Minimization (PEM) method is used to identify the system. Dynamic state space models of the system are obtained. The multi-input and multi-output (MIMO) system are identified at different power levels, and the model verification are carried out by simulate the step response output. The results show that the state space model of the once-through steam generation system identified by PEM method is of high precision. The step response of the model and the output of the actual system are in good agreement with each other. The identification scheme proposed in this paper provides a new method and idea for the modeling of nuclear power plant system. The model can provide foundation and technical support for the research of high precision and high quality control system.