Research On Reduced-order Extrapolation Model Of Air-cooled System Based On Proper Orthogonal Decomposition Theory

The air-cooled system of a thermal power plant is a complex partial differential equation system.Experiments or numerical simulations are usually used to study its flow and heat transfer characteristics.However,due to the complex coupling characteristics and high dimensional freedom of the system model,both methods require a large amount of experimental resources and calculation time,and cannot be applied to the prediction and control of the actual system.Therefore,it is important to establish a low-order extrapolation model for air-cooled systems that can guarantee sufficient accuracy and has fast calculation characteristics.The proper orthogonal decomposition(POD)is a multivariate statistical method that aims at obtaining a compact representation of the data.And it is widely used in model reduction of partial differential equations.In this paper,the mathematical model of the air-cooled system is reduced and a finite-difference-reduced-order extrapolation model based on the POD method is established.Firstly,a finite difference reduction model based on POD method for 2D Rayleigh-Benard model is established.The fastness of the reduced-order method is verified by an example analysis.Afterwards,according to the air-cooled condenser in the 600MW air-cooled system of an actual power plant,a flow model based on N-S equations is established.The finite difference method was used for the discretization.The combination of staggered grid and Poisson equation was used to solve the velocity field at the initial time,and the sample data was extracted to constitute the snapshots.Then,based on the singular value decomposition technique and the proper orthogonal decomposition theory,the POD basis is obtained.On this basis,the reduced-order extrapolation model of the two-dimensional air-cooled condenser is established.Under the condition of a given initial boundary value,the numerical analysis of the model before and after the reduction is compared.It is concluded that the calculation time of the reduced model is short and the result error is small,which verifies the high efficiency of the reduced model.At the same time,comparing the numerical simulation results under different operating conditions,it is found that the calculation time is not related to the working conditions.This is because for the partial differential equations numerical calculation,changing the working conditions only changes the calculated initial boundary conditions,and does not change the internal iterative solution in the calculation process.highlighting the flexibility of the reduced-order model.The reduced-order extrapolation model based on the POD method in this paper improves the calculation efficiency,saves the calculation time and ensures the accuracy.It provides a reliable basis for the flow field analysis of the following air-cooled system,and has a wide application prospect and practical value.