CN115510735A - Rail grinding target profile optimization design method based on wheel-rail contact parameters - Google Patents

Rail grinding target profile optimization design method based on wheel-rail contact parameters Download PDF

Info

Publication number
CN115510735A
CN115510735A CN202211006886.1A CN202211006886A CN115510735A CN 115510735 A CN115510735 A CN 115510735A CN 202211006886 A CN202211006886 A CN 202211006886A CN 115510735 A CN115510735 A CN 115510735A
Authority
CN
China
Prior art keywords
profile
rail
steel rail
wheel
curve
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202211006886.1A
Other languages
Chinese (zh)
Inventor
樊文刚
杨家庭
李建勇
刘月明
聂蒙
徐康
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Beijing Jiaotong University
Original Assignee
Beijing Jiaotong University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Beijing Jiaotong University filed Critical Beijing Jiaotong University
Priority to CN202211006886.1A priority Critical patent/CN115510735A/en
Publication of CN115510735A publication Critical patent/CN115510735A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/17—Mechanical parametric or variational design
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2111/00—Details relating to CAD techniques
    • G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/14—Force analysis or force optimisation, e.g. static or dynamic forces

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Geometry (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Constituent Portions Of Griding Lathes, Driving, Sensing And Control (AREA)

Abstract

The invention belongs to the technical field of steel rail grinding, and particularly relates to a steel rail grinding target profile optimization design method based on wheel rail contact parameters, wherein the method uses a local weighted regression method to carry out smooth filtering on a plurality of groups of measured steel rail sample profile curves; fitting the measured multiple groups of sample profile curves into a steel rail representative profile comprehensively representing the line wear characteristics by using an arithmetic mean method; performing numerical modeling on the steel rail representative profile curve by using a 3-time NURBS curve interpolation fitting method; the method comprises the steps of taking a control weight factor of a NURBS curve as a design variable of a grinding profile, taking wheel-rail contact stress, wheel-rail contact geometric parameters and wheel-rail contact grinding power as optimization indexes, establishing a multi-objective optimization model of the steel rail grinding profile in Isight software, and carrying out multi-objective optimization on representative profiles of straight line segments and curved line segments to obtain the optimal grinding target profile. The invention can optimize the grinding target profile in the process of grinding the steel rail by wearing the steel rail at the straight line section or the curve section of the railway line, thereby improving the quality of the grinding target profile of the steel rail, effectively reducing the wear among the wheel rails and prolonging the service life of the wheel rails.

Description

Rail grinding target profile optimization design method based on wheel-rail contact parameters
Technical Field
The invention belongs to the technical field of steel rail grinding, and particularly relates to a steel rail grinding target profile optimization design method based on wheel-rail contact parameters.
Background
The problem of wheel-rail contact is a complex nonlinear problem, the factors involved in optimizing the grinding profile of the steel rail are more, and the optimized evaluation indexes are different. At present, the optimization design of the grinding profile of the steel rail mainly aims at the standard grinding profile, single or less target optimization is realized under specific working conditions, and the research object lacks general adaptability. The optimization design of the steel rail grinding profile needs to be combined with actual line characteristics and vehicle running conditions, a complete system flow is required from profile measurement, profile data processing, profile optimization design and performance inspection, and a perfect optimization design scheme is rarely given through current research. The method for optimizing and designing the steel rail grinding target profile is provided aiming at the problems that the optimization of the current steel rail grinding target profile is high in design cost, long in optimization period, weak in optimization target pertinence, low in solving efficiency and the like, and based on the wheel-rail contact parameters and according to the operation conditions of the rails and vehicles of the actual railway line, and has certain engineering value significance for optimizing and designing the current steel rail grinding target profile and implementing steel rail grinding work.
Chinese patent with an authorization publication number of CN106951657A discloses a method for rapidly designing a grinding target profile of a worn steel rail, which comprises the steps of constructing a parameterized model of the wear profile by using a cubic NURBS curve; establishing a Kriging model of the grinding target profile performance of the worn steel rail by taking an adjustable weight factor of a NURBS parameterized model of the profile of the worn steel rail as a design variable and taking the contact stress of the wheel rail and the dynamic performance index of the wheel rail as dependent variables; and establishing an optimal design model of the rail grinding target profile by taking the adjustable weight factors as design variables and a Kriging model of the rail grinding target profile performance as a target function, so as to realize the optimal design of the worn rail grinding target profile. This method has some problems: firstly, the method does not relate to the pretreatment process of steel rail measurement data, does not provide selection or calculation standard of measurement profile original curve data, and when noise appears in the measured data or the difference between the used wear steel rail profile and the wear state of the whole grinding line steel rail is large, the optimized grinding profile has the condition of large error, and the quality of the ground steel rail can be reduced in serious cases; secondly, the method does not give concrete details about how the weight factors that the NURBS curve needs to be optimized are determined; moreover, the method uses a Kriging model, is a combined prediction method, needs to use other simulation software in optimization calculation, and causes that the optimization process is complex, and the integration level and the integration force of an optimization means are low.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to provide a rail grinding target profile optimization design method based on wheel rail contact parameters, which can optimize the grinding target profile in the rail grinding process of a rail worn by a straight section or a curved section of a railway line, thereby improving the quality of the rail grinding target profile, effectively reducing the wear among wheel rails, prolonging the service life of the wheel rails and improving the capability of a locomotive vehicle passing through the curved track.
In order to achieve the above purposes, the technical scheme adopted by the invention is as follows:
a rail grinding target profile optimization design method based on wheel-rail contact parameters comprises the following steps:
s1, measuring coordinate point data of a plurality of groups of steel rail section profile curves needing to be polished on a railway line, and performing smooth filtering on the profile curves by using a local weighted regression method;
s2, sequentially taking arithmetic mean values of corresponding coordinate points of the plurality of groups of filtered and smoothed sample profile curves to reconstruct a steel rail profile curve, and taking the steel rail profile curve as a representative profile curve of the group of steel rail sample profile curves;
s3, taking the coordinate points of the representative profile curve of the steel rail in the step S2 as model value points of the NURBS curve, fitting the representative profile by using a method of interpolation fitting of the NURBS curve for not less than 3 times, and establishing a numerical calculation model of the representative profile of the steel rail;
s4, solving two-dimensional wheel-rail contact geometric parameters by using the steel rail representative profile obtained by fitting in the step S3 and a trace method to obtain the positions of wheel-rail contact points of the wheels under different transverse displacement amounts and a distribution probability density curve of the contact points on the steel rail representative profile;
s5, determining an optimized region of the steel rail representative profile according to the distribution positions of the wheel rail contact points by using the control weight factors of the NURBS curve of the steel rail representative profile in the steps S3 and S4, and adjusting n control weight factors omega of the NURBS curve of the region i I belongs to {1,2,3, \8230;, n }, and a plurality of groups of steel rail profile curves are generated;
s6, writing the coordinate point data of the steel rail profile curve generated in the step S5 into a steel rail profile file available for multi-body dynamics software, establishing a three-dimensional simulation model of a vehicle and a track in the multi-body dynamics software according to an actual track line and vehicle parameters, and importing the steel rail profile file into the model to establish a vehicle track dynamics simulation calculation model;
s7, calculating and obtaining the contact stress P of the wheel-rail contact in the running process of the vehicle, the rolling circle radius difference delta r between the guide wheels of the vehicle and the abrasion work W of the wheel-rail contact by using the simulation model in the step S6 t ；
S8, enabling the steel rails in the step S5 to represent NURBS curve control weight factors omega of the to-be-optimized region of the profile i I e {1,2,3, \8230;, n } as design variables, the contact stress P of the wheel-rail contact in step S7, the rolling circle radius difference Deltar between the guide wheels of the vehicle, and the abrasion work W of the wheel-rail contact t As an optimization target, building a multi-target optimization model of the steel rail representative profile;
s9, calculating to obtain a control weight factor omega of the optimized steel rail profile NURBS curve by using the multi-objective optimization model in the step S8 i I ∈ {1,2,3, \8230;, n }, and constructing an optimized profile curve of the rail using step S3, therebyAnd obtaining the final rail grinding target profile.
On the basis of the above technical solution, step S8 specifically includes the following steps:
s81, controlling the weight factor omega of the n NURBS curves of the area to be optimized i I ∈ {1,2,3, \8230;, n } as design variables, and ω ∈ ω i ∈[0,1]The weight factors of other areas on the curve default to 1;
s82, sampling and generating m groups of weight factor combinations by using an optimal Latin hypercube experimental design method, reconstructing m steel rail profile curves with different shapes by using the generated weight factor combinations and weight factors of other areas as weight factors of the steel rail profile NURBS curve, and generating a plurality of groups of steel rail profile curves in the step S5, wherein m is more than or equal to 2n +1;
s83, the weight factor omega of the step S5 i I belongs to {1,2,3, \8230;, n } and the contact stress P, the rolling circle radius difference delta r and the abrasion power W which are obtained by corresponding calculation in the step S7 t Forming m groups of sample data, using RBF radial basis function neural network proxy model technology, taking the m groups of sample data as training samples of the proxy model, and establishing a proxy model for designing variables and optimizing indexes;
s84, minimizing the contact stress P of the wheel-rail contact, maximizing the rolling circle radius difference Delta r between the guide wheels of the vehicle and maximizing the abrasion work W of the wheel-rail contact t Minimum as optimization target, with weight factor ω i I belongs to {1,2,3, \8230;, n } is used as a design variable, a constructed proxy model is used for carrying out multi-target optimization calculation by using an NSGA-II algorithm to obtain an optimal weight factor combination omega i I ∈ {1,2,3, \8230;, n } and its corresponding optimal wheel-rail contact parameters.
On the basis of the above technical solution, step S3 specifically includes the following steps:
s31, taking the starting point of the straight line edge of the rail representative profile curve close to the left side 1;
and S32, calculating a node vector of the rail representative profile NURBS curve by adopting an accumulated chord length parameterization method, determining a boundary condition of the rail representative profile NURBS curve by using a tangent vector boundary condition, calculating a control vertex of the rail representative profile NURBS curve reversely, and finally determining a NURBS interpolation fitting curve of the rail representative profile.
S33, setting the initial values of the weight factors of the NURBS curve to be 1, establishing a numerical calculation model of the steel rail representative profile curve, and adjusting the values of the weight factors to obtain the steel rail profile curves in different shapes.
The steel rail grinding target profile optimization design method based on the wheel-rail contact parameters has the following beneficial effects:
the method solves the problem that when noise appears in the measured data or the difference between the used worn steel rail profile and the wear state of the whole polished line steel rail is large, the error of the optimized polished profile is large;
the method provides a method for determining the weight factor of the NURBS curve to be optimized;
the method has high optimization integration level and integration force, and the optimization process is relatively simple.
Drawings
The invention has the following drawings:
FIG. 1 is a flow chart of the method of the present invention;
FIG. 2 is a graph of a sample profile measured on a railway;
FIG. 3 is a NURBS interpolation fit plot of a representative profile of a rail;
FIG. 4 is a graph showing a contact point position and a contact point distribution probability density of the profile in contact with the wheel;
FIG. 5 is a RBF radial basis function neural network proxy model diagram of the design variables of the rail grinding target profile and the contact stress of the wheel rail;
FIG. 6 is a RBF radial basis function neural network proxy model diagram of the difference between the design variable of the rail grinding target profile and the radius of the wheel rail contact rolling circle;
FIG. 7 is a RBF radial basis function neural network proxy model diagram of the contact wear power of the rail grinding target profile design variable and the wheel rail;
and (5) optimally designing a calculated rail grinding target profile diagram.
Detailed Description
The present invention will be described in further detail with reference to the accompanying drawings.
A rail grinding target profile optimization design method based on wheel-rail contact parameters uses a local weighted regression method to perform smooth filtering on a plurality of groups of measured rail sample profile curves; fitting the measured multiple groups of sample profile curves into a steel rail representative profile comprehensively representing the line wear characteristics by using an arithmetic mean method; the method comprises the steps of carrying out numerical modeling on a representative profile curve of a steel rail by using a 3-time NURBS curve interpolation fitting method, establishing a multi-objective optimization model of the grinding profile of the steel rail in Isight multidisciplinary optimization software by using a control weight factor of the NURBS curve as a design variable of the grinding profile and using wheel-rail contact stress, wheel-rail contact geometric parameters and wheel-rail contact grinding power as optimization indexes, wherein the process from the processing of measured data of the profile of the steel rail to the calculation of the wheel-rail contact parameters and the building and solving processes of the multi-objective optimization model form a complete optimization system model, and carrying out multi-objective optimization on the representative profiles of the steel rail of straight line segments and curve segments so as to obtain the optimal grinding target profile.
As shown in fig. 1, the method for optimally designing a rail grinding target profile based on wheel-rail contact parameters comprises the following steps:
(1) As shown in fig. 2, coordinate point data of a plurality of groups of rail section profile curves to be polished on a railway line are measured by using a profile measuring instrument, and a local weighted regression method (loess) is used in MATLAB software to carry out smoothing filtering on the profile curves.
(2) And sequentially taking the arithmetic mean value of the corresponding coordinate points of the plurality of groups of filtered and smoothed sample profile curves to reconstruct a steel rail profile curve as a representative profile curve of the group of steel rail sample profile curves.
(3) And (3) establishing a 3-time NURBS interpolation fitting curve of the representative profile of the steel rail as shown in FIG. 3, taking the coordinate points of the representative profile curve of the steel rail in the step (2) as the model value points of the NURBS curve, fitting the representative profile by using a 3-time NURBS curve interpolation fitting method, and establishing a numerical calculation model of the representative profile of the steel rail.
The step (3) specifically comprises the following steps:
and (3.1) taking the starting point of the straight line edge of the steel rail representative profile curve close to the slope of the left side 1 as the origin, and uniformly selecting k coordinate points from the coordinate points on the steel rail representative profile curve in the step (2) as model value points for 3 times of NURBS curve interpolation fitting.
And (3.2) calculating node vectors of the 3-time NURBS curve of the steel rail representative profile by adopting an accumulated chord length parameterization method, determining the boundary conditions of the 3-time NURBS curve of the steel rail representative profile by using tangent vector boundary conditions, inversely calculating the control vertexes of the 3-time NURBS curve of the steel rail representative profile, and finally determining the 3-time NURBS interpolation fitting curve of the steel rail representative profile.
And (3.3) setting initial values of the weight factors omega of the NURBS curves for 3 times to be 1, establishing a numerical calculation model of the profile curve represented by the steel rail, and adjusting the values of the weight factors to obtain the profile curves of the steel rails with different shapes.
(4) And (4) solving two-dimensional wheel-rail contact geometric parameters by using the steel rail representative profile obtained by fitting in the step (3) and a trace method, as shown in fig. 4, to obtain the positions of the wheel-rail contact points under different transverse displacement amounts (+/-12 mm) and the distribution probability density curve of the contact points on the steel rail representative profile.
(5) Determining an optimized region of the representative profile of the steel rail according to the distribution position of the contact points of the rail by using the control weight factors of the NURBS curve of the representative profile of the steel rail in the step (3) and the step (4), and adjusting the control weight factors (omega) of n NURBS curves in the region i I belongs to {1,2,3, \8230;, n }), and a plurality of groups of steel rail profile curves are generated.
(6) And (4) writing the coordinate point data of the steel rail profile curve generated in the step (5) into a steel rail profile file available for multi-body dynamics software, establishing a three-dimensional simulation model of the vehicle and the track in the multi-body dynamics software according to the actual track line and the vehicle parameters, and importing the steel rail profile file into the model to establish a vehicle-track dynamics simulation calculation model.
(7) Calculating and obtaining the contact stress P of the wheel-rail contact in the running process of the vehicle, the rolling circle radius difference delta r between the guide wheels of the vehicle and the abrasion work W of the wheel-rail contact by using the simulation model in the step (6) t 。
(8) Representing the steel rail in the step (5) as a NURBS curve control weight factor (omega) of the region to be optimized of the profile i I ∈ {1,2,3, \ 8230;, n }) as design variables, the contact stress P of the wheel-rail contact in step (7), the rolling circle radius difference Δ r between the guide wheels of the vehicle, and the abrasion work W of the wheel-rail contact t And (3) as an optimization index, building a multi-objective optimization model of the steel rail representative profile in Isight software.
The step (8) specifically comprises the following steps:
(8.1) n control weight factors (ω) of the NURBS curve of the area to be optimized i I ∈ {1,2,3, \ 8230;, n }) as a design variable, and ω i ∈[0,1]And the weight factors of other areas on the curve default to 1.
(8.2) sampling and generating m groups of weight factor combinations (m is more than or equal to 2n + 1) by using an optimal Latin hypercube experimental design method, using the generated weight factor combinations and weight factors of other areas as weight factors of the steel rail profile NURBS curves, reconstructing m steel rail profile curves with different shapes, and generating a plurality of groups of steel rail profile curves in the step (5).
(8.3) weighting factor (omega) of step (5) i I belongs to {1,2,3, \8230;, n }) and the contact stress P, the rolling circle radius difference Deltar and the abrasion power W obtained by corresponding calculation in the step (7) t And (3) forming m groups of sample data, using RBF radial basis function neural network proxy model technology, taking the m groups of sample data as training samples of the proxy model, and establishing the proxy model of the design variables and the optimization indexes as shown in figures 5, 6 and 7.
(8.4) the contact stress P of the wheel-rail contact is minimum, the rolling circle radius difference Deltar between the vehicle guide wheels is maximum, and the abrasion work W of the wheel-rail contact t Minimum as optimization target, with weight factor (ω) i I belongs to {1,2,3, \8230n, }) as a design variable, and the constructed agent model carries out multi-target optimization calculation by using an NSGA-II algorithm to obtain an optimal weight factor combination (omega) i I ∈ {1,2,3, \8230;, n }) and its corresponding optimal wheel-rail contact parameters.
(9) Calculating and obtaining a control weight factor (omega) of the optimized steel rail profile NURBS curve by using the multi-objective optimization model in the step (8) i ,i∈{1,2,3,…N) and constructing an optimized rail profile curve by using the step (3), thereby obtaining a final rail grinding target profile as shown in fig. 8 (a dotted line part).
The rail required to be polished on a certain railway line is taken as an object, and a rail polishing target profile optimization design method based on wheel rail contact parameters is explained in detail below.
A rail grinding target profile optimization design method based on wheel-rail contact parameters comprises the following steps:
(1) The profile of the rail section is measured at regular intervals on a track to be polished on a railway line, and a group of data composed of a plurality of rail sample profile curves shown in fig. 2 is obtained. Smoothing the profile curve by using a local weighted regression (loess) method in MATLAB software, calculating an arithmetic average value of corresponding coordinate points of the plurality of profile curves to obtain a new group of coordinate points, and reconstructing the coordinate points into a rail profile curve as a representative profile of the group of sample rail profile curves.
(2) And (2) uniformly selecting 29 coordinate points on the representative profile curve in the step (1), wherein the coordinate points are shown in a table 1:
TABLE 1 two-dimensional coordinate values of the profile curve of a steel rail
Figure BDA0003809211510000081
(3) The coordinate points in table 1 are used as model points for constructing a 3-time NURBS curve of the representative profile of the steel rail, and a numerical calculation model of the representative profile curve of the steel rail as shown in fig. 3 is established.
(4) Using a trace method, calculating two-dimensional wheel-rail contact geometric parameters, as shown in fig. 4, obtaining the positions of the wheel-rail contact points under different displacements (± 12 mm) and the distribution probability density curves of the contact points on the representative profile of the steel rail.
(5) Finding a weight factor omega for controlling the region on a 3-time NURBS curve of a steel rail representative profile according to the contact region of the wheel rail 12 、ω 13 、ω 14 、ω 22 、ω 23 、ω 24 Apply these weightsAnd (3) taking the variables as design variables, sampling and generating 60 groups of weight factor combinations by using an optimal Latin hypercube experimental design method in Isight software, and constructing 60 steel rail profile curves with different shapes by using the weight factor combinations.
(6) Writing 60 steel rail profile curves with different shapes into 60 steel rail profile files used by SIMPACK, importing the files into a built vehicle-track multi-body dynamics simulation model, and calculating to obtain corresponding wheel-track contact stress P, rolling circle radius difference delta r and abrasion power W t The calculation results are shown in table 2:
TABLE 2 weight factor design variables and corresponding wheel-track contact optimization parameters
Figure BDA0003809211510000091
(7) Constructing an RBF radial basis function neural network agent model with design variables corresponding to optimization targets shown in figures 5, 6 and 7 in Isight software according to the data in the table 2; the contact stress P of the wheel-rail contact is minimum, the rolling circle radius difference Delta r between the guide wheels of the vehicle is maximum, and the abrasion work W of the wheel-rail contact t Minimum as optimization target, with weight factor ω 12 、ω 13 、ω 14 、ω 22 、ω 23 、ω 24 As a design variable, the constructed agent model uses the NSGA-II algorithm to perform multi-target optimization calculation to obtain the optimal weight factor combination omega 12 、ω 13 、ω 14 、ω 22 、ω 23 、ω 24 And its corresponding optimal wheel-rail contact parameters.
(8) And (4) generating an optimized rail profile curve by using the optimal weight factors obtained in the step (7) by using the NURBS curve construction method in the step (3) as shown in fig. 8 (a dotted line part), thereby obtaining a final rail grinding target profile.
Those not described in detail in this specification are within the skill of the art.

Claims (3)

1. A rail grinding target profile optimization design method based on wheel-rail contact parameters is characterized by comprising the following steps:
s1, measuring coordinate point data of a plurality of groups of steel rail section profile curves needing to be polished on a railway line, and performing smooth filtering on the profile curves by using a local weighted regression method;
s2, sequentially taking arithmetic mean values of corresponding coordinate points of the plurality of groups of filtered and smoothed sample profile curves to reconstruct a steel rail profile curve, and taking the steel rail profile curve as a representative profile curve of the group of steel rail sample profile curves;
s3, taking the coordinate points of the representative profile curve of the steel rail in the step S2 as model value points of the NURBS curve, fitting the representative profile by using a method of interpolation fitting of the NURBS curve for not less than 3 times, and establishing a numerical calculation model of the representative profile of the steel rail;
s4, solving two-dimensional wheel-rail contact geometric parameters by using the steel rail representative profile obtained by fitting in the step S3 and a trace method to obtain the positions of wheel-rail contact points of the wheels under different transverse displacement and a distribution probability density curve of the contact points on the steel rail representative profile;
s5, determining an optimized region of the steel rail representative profile according to the distribution position of the wheel rail contact points by using the control weight factors of the steel rail representative profile NURBS curve in the steps S3 and S4, and adjusting n control weight factors omega of the NURBS curve in the region i I belongs to {1,2,3, \8230;, n }, and a plurality of groups of steel rail profile curves are generated;
s6, writing the coordinate point data of the profile curve of the steel rail generated in the step S5 into a steel rail profile file available for multi-body dynamics software, establishing a three-dimensional simulation model of a vehicle and a track in the multi-body dynamics software according to an actual track line and vehicle parameters, and importing the steel rail profile file into the model to establish a vehicle track dynamics simulation calculation model;
s7, calculating and obtaining the contact stress P of the wheel-rail contact in the running process of the vehicle, the rolling circle radius difference delta r between the guide wheels of the vehicle and the abrasion work W of the wheel-rail contact by using the simulation model in the step S6 t ；
S8, enabling the steel rails in the step S5 to represent NURBS curve control weight factors omega of the to-be-optimized region of the profile i ,i∈{1,23, \8230n } as design variables, the contact stress P of the wheel-rail contact in step S7, the rolling circle radius difference Deltar between the guide wheels of the vehicle, and the abrasion work W of the wheel-rail contact t As an optimization target, building a multi-target optimization model of the steel rail representative profile;
s9, calculating and obtaining the control weight factor omega of the optimized steel rail profile NURBS curve by utilizing the multi-objective optimization model in the step S8 i And i belongs to {1,2,3, \8230;, n }, and constructing an optimized steel rail profile curve by using the step S3, so as to obtain a final steel rail grinding target profile.
2. A rail grinding target profile optimization design method based on wheel and rail contact parameters as claimed in claim 1, wherein said step S8 specifically comprises the following steps:
s81, controlling the weight factor omega of the n NURBS curves of the area to be optimized i I ∈ {1,2,3, \8230;, n } as a design variable, and ω i ∈[0,1]The weight factors of other areas on the curve default to 1;
s82, sampling and generating m groups of weight factor combinations by using an optimal Latin hypercube experimental design method, reconstructing m steel rail profile curves with different shapes by using the generated weight factor combinations and weight factors of other areas as weight factors of the steel rail profile NURBS curve, and generating a plurality of groups of steel rail profile curves in the step S5, wherein m is more than or equal to 2n +1;
s83, the weight factor omega of the step S5 is used i I ∈ {1,2,3, \8230;, n } and the contact stress P, rolling circle radius difference Deltar and abrasion power W calculated corresponding to the step S7 t Forming m groups of sample data, using RBF radial basis function neural network proxy model technology, taking the m groups of sample data as training samples of the proxy model, and establishing a proxy model for designing variables and optimizing indexes;
s84, minimizing the contact stress P of the wheel-rail contact, maximizing the rolling circle radius difference Delta r between the guide wheels of the vehicle and maximizing the abrasion work W of the wheel-rail contact t Minimum as optimization target, with weight factor ω i I belongs to {1,2,3, \8230;, n } is used as a design variable, and a constructed agent model uses an NSGA-II algorithm to carry out multi-target optimization calculation to obtainTo the optimal weight factor combination omega i I ∈ {1,2,3, \8230;, n } and its corresponding optimal wheel-rail contact parameters.
3. A rail grinding target profile optimization design method based on wheel rail contact parameters as claimed in claim 1, wherein said step S3 specifically comprises the following steps:
s31, taking the starting point of the straight line edge of the rail representative profile curve close to the left side 1;
and S32, calculating a node vector of the NURBS curve of the steel rail representative profile by adopting an accumulated chord length parameterization method, determining a boundary condition of the NURBS curve of the steel rail representative profile by using a tangent vector boundary condition, inversely calculating a control vertex of the NURBS curve of the steel rail representative profile, and finally determining a NURBS interpolation fitting curve of the steel rail representative profile.
S33, setting the initial values of the weight factors of the NURBS curve to be 1, establishing a numerical calculation model of the steel rail representative profile curve, and adjusting the values of the weight factors to obtain the steel rail profile curves in different shapes.
CN202211006886.1A 2022-08-22 2022-08-22 Rail grinding target profile optimization design method based on wheel-rail contact parameters Pending CN115510735A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211006886.1A CN115510735A (en) 2022-08-22 2022-08-22 Rail grinding target profile optimization design method based on wheel-rail contact parameters

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211006886.1A CN115510735A (en) 2022-08-22 2022-08-22 Rail grinding target profile optimization design method based on wheel-rail contact parameters

Publications (1)

Publication Number Publication Date
CN115510735A true CN115510735A (en) 2022-12-23

Family

ID=84501423

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211006886.1A Pending CN115510735A (en) 2022-08-22 2022-08-22 Rail grinding target profile optimization design method based on wheel-rail contact parameters

Country Status (1)

Country Link
CN (1) CN115510735A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116150918B (en) * 2023-04-21 2023-08-04 北京交通大学 Intelligent rail profile optimization method considering distribution characteristics of contact positions of wheel and rail on rail

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116150918B (en) * 2023-04-21 2023-08-04 北京交通大学 Intelligent rail profile optimization method considering distribution characteristics of contact positions of wheel and rail on rail

Similar Documents

Publication Publication Date Title
CN106951657B (en) Method for quickly designing grinding target profile of worn steel rail
CN109815523B (en) Multi-target differential evolution algorithm for train operation based on decomposition
Pombo et al. Development of a wear prediction tool for steel railway wheels using three alternative wear functions
CN115510735A (en) Rail grinding target profile optimization design method based on wheel-rail contact parameters
CN108589449A (en) A kind of personalization railway steel rail polishing target profile design method
CN104036089B (en) A kind of high speed train wheel Structure optimization method for designing of low abrasion
CN111764208A (en) Steel rail grinding profile rapid prediction method based on interval segmentation
CN105912755B (en) A kind of steel rail grinding profile prediction technique based on NURBS-Kriging
CN110688710A (en) Turning repairing method based on rail transit vehicle wheel pair service life statistical model
CN112528403A (en) Economic optimization turning repair method for wheel tread of railway vehicle
CN102644663A (en) Cylindrical roller multi-circular-arc variable curvature profile engineering simulation method
CN114912159B (en) Method for fitting geometric line shape of rail transit line plane
CN112685942B (en) Rapid partitioning method for finite element grids of complex tread patterns
CN111582656A (en) Quantitative evaluation method for shallow layer state of high-speed railway steel rail
CN110598275A (en) Wheel profile optimization method based on response surface modeling and improved particle swarm optimization
CN109033482A (en) A kind of abrasion rail type face economy polishing process
CN110567661B (en) Bridge damage identification method based on generalized pattern search algorithm and axle coupling
CN102029293B (en) Tandem cold rolling control method for optimizing multi-motor load
CN112836272B (en) High-speed railway steel rail profile optimization design method based on neural network model
CN115310217A (en) Wheel wear prediction method based on non-Hertz wheel-rail rolling contact theory
CN107103136A (en) A kind of rail vehicle wheel track contact point computational methods based on distance field
de Paula Pacheco et al. Optimization of heavy haul railway wheel profile based on rolling contact fatigue and wear performance
CN111353250B (en) Helical gear spiral type induction heating simulation method based on ANSYS simulation platform
CN102360402A (en) Optimized design method for hydraulic cylinder supporting seat of cover carriage based on response surface
Pan et al. An optimization algorithm for locomotive secondary spring load adjustment based on artificial immune

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination