Research on Rapid Prediction Method of Thermal Process Based on Reduced Order Model
In order to design the thermal process control,the computational fluid dynamics(CFD)method with"heat flow coupling"is often used to simulate the thermal process.In order to obtain a reasonable thermal control scheme,repeated CFD analysis is often required,the cost is high and it is not convenient to adjust the control scheme in real time.To solve this problem,a method for rapid analysis and prediction of thermal process is proposed based on Volterra series reduced-order model,it provides a instrument for the formulation of thermal process con-trol strategy and the determination of control parameters.According to the response of thermal process system under step temperature heating condition,the kernel function of Volterra series reduced-order model is identified,the reduced-order model of the relationship between system temperature and control conditions is constructed,then the temperature response of the system under different control processes can be calcu-lated quickly.The reduced-order model can be used to predict the heating process of the system in real time and optimize the thermal process control method.The result of numerical example shows that the temperature response obtained by Volterra series reduced-order model is con-sistent with that obtained by CFD method,and the efficiency of reduced-order model method for thermal process prediction and control optimi-zation is very high.
万方数据
维普
