Virtual Simulation Experiment Teaching of Non-Gaussian Wind Pressure Prediction Based on Artificial Intelligence
Using limited random wind field time history data to predict the random wind field time history of unknown point positions can achieve the purpose of virtual simulation experiment teaching,and can save experimental costs and resources to a certain extent,reduce the difficulty of experimental testing.A non-Gaussian wind pressure prediction simulation method based on support vector machine(SVM)is established using Matlab in the paper.Simulation results indicate that the choice of kernel function in SVM significantly impacts the simulation performance of non-Gaussian wind pressure prediction.The linear kernel function model demonstrates better simulation effectiveness for non-Gaussian wind pressure prediction compared to Gaussian and exponential kernel functions.Therefore,the SVM linear kernel function model can effectively predict non-Gaussian wind pressures,providing valuable insights for virtual simulation experiments in wind tunnel tests or field measurements.
non-Gaussian wind pressurevirtual simulationteaching reformartificial intelligencesupport vector machinewind field predition
NETL
NSTL
万方数据
