Prediction Of Wind Pressure On CAARC Standard High-rise Building Using POD

In recent years,the typhoon has always invaded the coast of our country and brought tremendous losses to our country.Due to the frequent structural damage caused by strong winds,the experimental data measured by the wind tunnel test can not meet the requirements of structural analys is.Therefore,wind pressure prediction is often performed in places where no measuring points are arranged so as to obtain enough wind pressure data.There are many researches on wind pressure prediction at this stage,such as interpolation function metho d,artificial neural network method and POD method.The accuracy of POD prediction depends on the interpolation function used.In the previous study,no interpolation function could accurately predict the fluctuating wind pressure in the turbulent flow region.In this paper,different interpolation functions are selected to study POD pulsatile wind pressure prediction,which complements previous research.Considering the complex and diverse structural forms at present,it is difficult to accurately predict the fluctuating wind pressure using the ordinary POD method.BP neural network is a tool that can predict the nonlinear relationship well.Therefore,combining the BP neural network with the POD method,Complicated area for pulsating wind pressure predicti on.The main content of this paper is divided into the following sections:Rigid manometry experiments were carried out on the CAARC high building standard model and the raised-surface high-rise building model to obtain the wind pressure data of the buildi ng surface.The wind tunnel experiment was carried out in the atmospheric boundary layer wind tunnel in Shenzhen Wind Environmental Technology Engineering Laboratory of Harbin Institute of Technology(Shenzhen).The experimental results of the standard model of high-rise building are in good agreement with the experimental results of authoritative wind tunnel institutions both at home and abroad.The performance meets the requirements of the actual project.By comparison,it is found that the convex surface has a great influence on the distribution of fluctuating wind pressure on the surface.The POD method based on different interpolation functions is used to predict the fluctuating wind pressure at unplanned measuring points.Based on the experimental data of CAARC high-rise building standard model,the POD method is used to decompose the fluctuating wind pressure field on the building surface and analyze the characteristics of pulsating wind pressure field on the standard high-rise building.The fluctuatin g wind pressure field is rebuilt and analyzed,The difference between the reconstruction results and the optimal reconstruction mode order of each plane is obtained;then the POD method based on different interpolation functions is predicted for the fluctu ating wind pressure at the unplanned points,and the predicted fluctuating wind pressure field The time-domain and frequency-domain properties of the proposed method are obtained.The influence of different interpolation functions on the prediction results is obtained.Combined with the respective advantages of BP neural network and POD method,the more complicated area of fluctuating wind pressure is predicted.POD-BPNN method is used to forecast the fluctuating wind pressure.The POD-BPNN method can accurately predict the fluctuating wind pressure in this area.Whether in the time domain or the frequency domain,Which are in good agreement with the measured results.POD-BPNN method is used to predict the fluctuating wind pressure in the presence of raised building surfaces.The effect of the height of the projections on the prediction results is also given.The prediction effect of this method when the turbule nce is complex is obtained.