Font Size: a A A

Research On Kriging Surrogate Model For Railway Intelligent Route Selection

Posted on:2023-07-31Degree:MasterType:Thesis
Country:ChinaCandidate:Q H ZhouFull Text:PDF
GTID:2542307070485754Subject:Road and Railway Engineering
Abstract/Summary:PDF Full Text Request
Railway 3D route selection depends on retrieving the digital ground model constructed by the triangular network of millions to tens of millions,and also depends on the evaluation and calculation of the objective function more than hundreds of thousands of times.In the case of a large number of design variables,this will not only cause serious calculation time,but also make the calculation crash.The surrogate model can "replace" the complex and time-consuming real numerical calculation model in the optimization design process,greatly improving the optimization efficiency and reducing the computational complexity.Based on the classical Kriging surrogate model,this paper improves the construction method of Kriging surrogate model,applies this method to railway route selection,and constructs the Kriging surrogate model of railway intelligent route selection.The main research contents are as follows:(1)The maximum and minimum distance algorithm is introduced into Latin hypercube sampling method,and an improved Latin hypercube sampling method is proposed.By comparing with other sampling methods,the uniformity of sampling results is improved.(2)The particle swarm optimization algorithm is used to solve the super parameters of Kriging surrogate model,and the optimal super parameters are determined.(3)The worst interpolation criterion and the mixed multipoint addition criterion are proposed and applied to the dynamic optimization design of Kriging surrogate model respectively;considering that the mixed multipoint rule can significantly speed up the construction of Kriging surrogate model,a dynamic design framework of Kriging surrogate model based on the mixed multipoint rule is constructed.(4)Kriging surrogate model is applied to railway three-dimensional route selection design,and the surrogate model of railway threedimensional route selection design is constructed for the first time;The validity of the surrogate model in the three-dimensional railway alignment design problem is verified by an example test.
Keywords/Search Tags:Kriging surrogate model, Interpolation criteria, Intelligent Route selection, optimal design
PDF Full Text Request
Related items