Steel rail profile optimization design method based on polishing index

Abstract

The invention provides a steel rail profile optimization design method based on a polishing index, and belongs to the technical field of railway engineering. The steel rail profile optimization design method comprises the following steps: collecting a plurality of actually measured profiles of a steel rail subjected to field polishing and acceptance inspection; calculating deviation values of the actually measured profiles relative to the target profile at different polishing angles; on the basis of the deviation values at the different polishing angles, the profile adjustable degree at the different polishing angles is calculated; and setting optimization constraint conditions according to the profile adjustable degree and the wheel-rail contact geometrical relationship, and calculating the optimized profile of the steel rail. According to the steel rail profile optimization design method, deviation in the polishing process can be effectively recognized and quantified, the profile adjustment degree under different polishing angles is considered, optimization constraint conditions are set in combination with the wheel-rail contact geometrical relationship, the optimized steel rail profile is designed, the negative influence of polishing deviation on the steel rail performance is reduced, and the polishing precision is improved. The wheel-rail contact relation is optimized, and the running stability and riding comfort of a train are improved.

Description

Technical Field

[0001] The invention relates to the technical field of railway engineering, and in particular to a rail profile optimization design method based on a grinding index. Background Art

[0002] As the scale of railways expands, the requirements for railway maintenance will also increase. Rails are the components that directly carry trains. The matching relationship between wheels and rails determines their complex interactions, as well as the development trend of wear and damage. Rail grinding is a direct and effective way to reduce diseases. At the same time, it can improve the wheel-rail relationship and extend the service life of rails.

[0003] The optimization design of the target profile for rail grinding is the key point in the rail grinding process. Whether the target profile is suitable directly affects the train operation effect after rail grinding. Reasonable optimization design of the rail profile can not only improve the wheel-rail contact geometry and improve the train dynamics, but also has great significance for extending the service time and maintenance cycle of the rail and improving the service life of the rail.

[0004] An existing rail grinding technology solution directly uses the 60N profile as the target profile for rail grinding. However, the adaptability of the 60N profile to match different vehicle profiles is insufficient. If the 60N profile is used on site for grinding, the grinding amount is too large, the grinding cost is high, and the effect may not be ideal.

[0005] Another existing rail grinding technology solution optimizes the original profile, first determines the rail profile optimization area based on the wheel-rail contact position, takes rail wear as the objective function, and generates alternative rail profiles based on the points to be optimized and the constraints in the rail profile optimization area. Input and output data sets are generated based on the coordinates of the alternative profiles and the objective function, and a neural network model is constructed. A rail profile optimization model is then built based on a genetic algorithm, and the optimized rail profile is calculated and formed through a neural network. However, this solution focuses on reducing wheel-rail wear. According to actual optimization cases, the optimized profile can reduce wheel-rail wear. However, judging from actual service conditions, the current high-speed railway rail wear rate is relatively low, the profile variable is not large, and the impact on wheel-rail matching is also relatively small. Therefore, for the high-speed railway rail profile, the impact of grinding is still relatively large,

[0006] In view of this, the inventor, based on many years of experience in production design in this field and related fields, has designed a rail profile optimization design method based on grinding index through repeated experiments, in order to solve the problems existing in the prior art. Summary of the invention

[0007] The purpose of the present invention is to provide a rail profile optimization design method based on grinding index, which can reduce the influence of grinding negative deviation, optimize the wheel-rail contact relationship, and improve the stability of the vehicle during operation.

[0008] To achieve the above-mentioned purpose of the invention, the present invention proposes a rail profile optimization design method based on grinding index, wherein a plurality of measured profiles of the rails after on-site grinding and acceptance are collected; the deviation value of each measured profile relative to the target profile at different grinding angles is calculated; the adjustable degree of the profile at different grinding angles is calculated based on the deviation value at different grinding angles; and optimization constraints are set according to the adjustable degree of the profile and the wheel-rail contact geometry, and the optimized rail profile is calculated.

[0009] Compared with the prior art, the present invention has the following characteristics and advantages:

[0010] The grinding index-based rail profile optimization design method proposed in the present invention quantifies the deviation distribution characteristics between the profile after grinding and the target profile by statistically analyzing the profile data of on-site grinding acceptance, and designs an optimized rail profile in combination with the wheel-rail contact geometry relationship, so as to improve the accuracy and reliability of rail profile design, enhance the wheel-rail matching performance, improve the stability and safety of train operation, improve ride comfort, reduce maintenance costs, and extend the service life of rails. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] The accompanying drawings described herein are only for explanation purposes and are not intended to limit the scope of the present invention in any way. In addition, the shapes and proportional dimensions of the various components in the figures are only schematic, used to help understand the present invention, and are not specifically limited to the shapes and proportional dimensions of the various components of the present invention. Those skilled in the art can select various possible shapes and proportional dimensions to implement the present invention according to the teachings of the present invention.

[0012] Figure 1 It is a schematic diagram of the rail profile optimization design method of the present invention;

[0013] Figure 2 It is a schematic diagram of calculating the deviation values ​​of each measured profile relative to the target profile at different grinding angles according to the present invention;

[0014] Figure 3 A schematic diagram of a method for calculating the adjustable degree of the profile at different grinding angles by using the deviation values ​​at different grinding angles of the present invention;

[0015] Figure 4 It is a schematic diagram of the profile angle of a certain point on the profile of the present invention;

[0016] Figure 5 It is a schematic diagram of the normal equation and the tangent point at different angles of the profile of the present invention;

[0017] Figure 6 It is a schematic diagram of the deviation of the profile grinding of the rail head of a 60N rail of the present invention;

[0018] Figure 7 The figure is a schematic diagram of the tangent relationship between the arc segments analyzed for the rail head profile of the 60N rail of the present invention. DETAILED DESCRIPTION

[0019] The details of the present invention can be more clearly understood with reference to the accompanying drawings and the description of the specific embodiments of the present invention. However, the specific embodiments of the present invention described herein are only used for the purpose of explaining the present invention and cannot be understood as limiting the present invention in any way. Under the guidance of the present invention, technicians can conceive of any possible variations based on the present invention, which should all be considered to belong to the scope of the present invention.

[0020] The present invention proposes a rail profile optimization design method based on a grinding index, wherein a plurality of measured profiles of rails after on-site grinding and acceptance are collected; the deviation value of each measured profile relative to a target profile at different grinding angles is calculated; the adjustable degree of the profile at different grinding angles is calculated based on the deviation value at different grinding angles; and optimization constraints are set according to the adjustable degree of the profile and the wheel-rail contact geometry relationship, and the optimized rail profile is calculated.

[0021] The rail profile optimization design method based on grinding index proposed in the present invention can effectively identify and quantify the deviation in the grinding process by accurately collecting the measured rail profile data after on-site grinding and acceptance, calculating the deviation value from the target profile, and considering the adjustable degree of the profile under different grinding angles, setting optimization constraints in combination with the wheel-rail contact geometry to design the optimized rail profile, thereby reducing the negative impact of grinding deviation on rail performance, optimizing the wheel-rail contact relationship, and improving the stability of train operation and ride comfort.

[0022] In an optional embodiment of the present invention, Figure 2 As shown, the deviation value of each measured profile relative to the target profile at different polishing angles is calculated, including: calculating the normal equation of the normal line of each polishing angle; recording the coordinates of the first intersection point of each normal line with the target profile; recording the coordinates of the second intersection point of each normal line with the measured profile; calculating the distance between the first intersection point and the second intersection point on the same normal line.

[0023] like Figure 4 As shown in the figure, the measured rail profile after on-site grinding is collected, the measured profile is aligned with the standard 60N profile (i.e., the target profile), and the deviation values ​​of the profile under different grinding angles are calculated and counted. The specific calculation process is as follows: first, the normal line equation of each angle is calculated, and the angle θ between the tangent line at a certain point on the profile and the x-axis is the profile angle at that point.

[0024] like Figure 5 As shown, record the normal equations at different angles and the coordinates of the tangent point. The tangent point is the intersection of the normal line and the target profile. Record the horizontal and vertical shifts in the profile alignment, and perform the same horizontal and vertical shift operations on the obtained normal line and tangent point to obtain the normal equations and tangent points at different angles of the profile after alignment. The calculation method of the profile deviation value is as follows:

[0025] like Figure 6 As shown in the figure, the tangent points after alignment obtained in the above steps are the intersection points of each normal line and the target profile. The profile is translated to obtain the intersection points of the measured profile and each normal line. The distance between the two points is the deviation value between the measured profile of the rail head for grinding and acceptance and the target profile (here 60N).

[0026] For example, when θ = 60°, the intersection of the normal and the target silhouette is p, and the coordinates are marked as (x p ,y p ), the intersection of the normal line and the measured profile is q, and the coordinate is (x q ,y q ). When θ=60°, the error between the two surfaces is

[0027] Furthermore, the calculated deviation values ​​of the rail profile can be listed in the following Table 1:

[0028] Table 1:

[0029] Mileage(Share) Profile deviation / mm Mileage(Share) Profile deviation / mm XXX XXX

[0030] In the present invention, the rail profile optimization design method based on the grinding index can accurately evaluate the rail grinding effect and optimize the design. The deviation under each grinding angle is quantified, so as to more accurately identify and adjust the rail profile to ensure its consistency with the target profile, which not only improves the quality and efficiency of rail grinding, but also helps to reduce rail damage caused by improper grinding and extend the service life of the rail.

[0031] In an optional embodiment of the present invention, Figure 3 As shown, the degree of profile adjustability at different grinding angles is calculated based on the deviation values ​​at different grinding angles, including: performing statistical analysis on the deviation values ​​at different grinding angles and obtaining profile deviation distribution characteristics; calculating the sensitivity of profile deviation at each grinding angle according to the profile deviation distribution characteristics; calculating the degree of profile shell adjustment at each grinding angle according to the sensitivity of profile deviation at each grinding angle.

[0032] In this embodiment, the measured profile deviation values ​​at different grinding angles are statistically analyzed to obtain the distribution characteristics of the profile deviation. By performing mathematical and statistical processing on the collected measured data, the distribution law of the deviation values ​​at different grinding angles is revealed. Next, based on the distribution characteristics of the deviation values ​​at different grinding angles, the sensitivity of the profile deviation at each grinding angle is calculated, that is, the degree of influence of the profile deviation on the final wheel-rail contact geometry at different grinding angles is evaluated. Finally, based on the sensitivity analysis results of the deviation values ​​at different grinding angles, the adjustable degree of the profile at each grinding angle is calculated, which provides an accurate adjustment basis for the subsequent optimization of the rail profile. Through the above method, the present invention can provide a scientific and quantitative adjustment plan for rail grinding, ensuring that the deviation between the optimization of the rail profile in actual application and the target profile is minimized.

[0033] Specifically, the deviation values ​​of all measured profiles are calculated, and statistical eigenvalue calculation and kernel density estimation (KDE) calculation are performed to obtain the maximum positive deviation and the minimum negative deviation of the measured profile at each angle, that is, the allowable deviation change limit at each angle, and the deviation value distribution characteristics of the actual acceptance profile at different angles are obtained.

[0034] The statistical results can be listed in Table 2:

[0035] Table 2:

[0036] Grinding angle / ° Mean / mm Standard Deviation

[0037] The calculation method of kernel density estimation is as follows:

[0038] (1) Select an appropriate kernel function K(x). Common kernel functions include Gaussian kernel, uniform kernel, triangular kernel, etc. In an optional example, a Gaussian kernel is used, and the formula is as follows:

[0039]

[0040] (2) For a given sample data {x1,x2,x3...x n}, the formula for kernel density estimation is:

[0041]

[0042] This formula means that at each data point x i A kernel function with a bandwidth of h is placed at the point x, and then the values ​​of all kernel functions are added and normalized to obtain the density estimate at point x.

[0043] The distribution of the profile deviation obtained by statistics is shown in Table 3:

[0044] Table 3:

[0045]

[0046] The sensitivity of the deviation index of all measured profiles to the profile deviation of each area is calculated, and the influence of different areas on the size of the profile deviation index is obtained. That is, the influence of the profile deviation in different areas of the rail head on the profile quality is determined, and the key control areas for grinding distributed on the rail head profile are analyzed.

[0047] Regarding the sensitivity calculation method, in an optional example, the Sobol sensitivity calculation method is used, as follows:

[0048] (1) Define the model. Define the input variable X = (X1, X2, X3...X n ) and the model relationship Y=f(X) of the output variable Y. In this example, the input variable is the deviation value at different angles, and the output variable is the rail profile index. The model relationship between the two is determined through the following process.

[0049] First, divide the rail profile into different areas according to the grinding angle, as shown in Table 4:

[0050] Table 4:

[0051] Partition 1 2 3 4 angle -10°～0° 0°～5° 5°～15° 15°～55° Deviation value +0.2 / -0.2 +0.2 / -0.2 +0.2 / -0.2 +0.2 / -0.6

[0052] Among them, the weight coefficient w of each region is determined by expert consultation method. qm (m=1,2,…,4), the values ​​of region 1 to region 4 are 0.22, 0.41, 0.20, and 0.17 respectively.

[0053] For a point k on the measured rail profile i The absolute value of the deviation between the point and the target profile (standard profile) in the normal direction is Δk i , then the calculation process of the profile deviation index is as follows:

[0054]

[0055] Where P di Δk is the calculation result of the deviation value of a single scattered point in the profile; i is the actual deviation value of a certain (i-th) angle of the measured rail profile; Δk 标 It is the upper and lower deviation limit of the corresponding angle.

[0056]

[0057] In the formula, w qm (m=1,2,…,4) are the weight values ​​of the rail profiles in different areas; n m(m=1,2,…,4) respectively represent the total number of grinding angles in the mth rail surface area.

[0058] (2) Construct samples. Generate independent and identically distributed input variable samples. In an optional example, a Monte Carlo method is used to generate the sample set.

[0059] (3) Calculate the Sobol sensitivity index. Calculate the Sobol sensitivity index of the input variable. The first-order sensitivity index S i , the calculation formula is:

[0060]

[0061] Among them, V i =Var[E(Y|X i )], V(Y) is the total variance of the output Y.

[0062] The calculated sensitivity at different grinding angles is shown in Table 5:

[0063] Table 5:

[0064] Grinding angle / ° Grinding index sensitivity

[0065] According to the sensitivity calculated from the rail profile index, we can clearly know the influence of the profile change under different grinding angles on the profile index. Based on this, we can find the inverse of the sensitivity at each angle and perform normalization to get the adjustable degree of the profile change at each grinding angle:

[0066]

[0067] The degree of adjustability of the profile at different angles can be recorded in Table 6 below:

[0068] Table 6:

[0069] Grinding / ° Adjustable degree

[0070] In an optional example of this embodiment, the deviation value distribution characteristics of the rail profile are characterized by kernel density.

[0071] In the rail profile optimization design method based on the grinding index of the present invention, the deviation distribution characteristics of the rail profile are characterized by using kernel density estimation. The measured profile data collected at each grinding angle are deeply analyzed, and the distribution characteristics of the rail profile deviation can be accurately captured and described through kernel density estimation. In the rail profile optimization design method based on the grinding index of the present invention, the distribution of deviation values ​​between the measured profile and the target profile is depicted by kernel density estimation to obtain a continuous probability density function, which can reveal the distribution trend and concentrated area of ​​the deviation value. The method of characterizing the distribution characteristics of rail profile deviation through this kernel density estimation can provide a more accurate and intuitive way to understand and analyze the deviation of the rail profile, and can quickly identify which areas have the most significant deviations, so as to make targeted adjustments.

[0072] In an optional example of this embodiment, the rail profile deviation distribution characteristics are characterized by probability density or quantile.

[0073] In an optional example of this implementation, statistical analysis is performed on the deviation values ​​at different grinding angles, including: calculating the mean and standard value of the deviation values ​​at each grinding angle.

[0074] Specifically, in the rail profile optimization design method based on the grinding index of the present invention, the profile deviation under different grinding angles is statistically analyzed to obtain the mean, standard deviation and other data of the grinding deviation. The calculation of the mean reflects the average level of deviation under each grinding angle, and the calculation of the standard deviation reveals the discrete degree of the deviation value, that is, the fluctuation of the deviation value around the mean. The purpose of calculating these two statistics is to clarify the correlation between the profile deviation and the grinding angle, to quantify and evaluate the influence of different grinding angles on the rail profile, and to provide an important reference for the subsequent optimization of the rail profile. The calculation of these two statistics can also be calculated by other calculation methods to obtain the probability density of different grinding angles or different percentage quantiles, etc. The rail profile optimization design method based on the grinding index of the present invention uses probability density or quantiles to characterize the distribution characteristics of rail profile deviations, and describes and analyzes the deviation of the rail profile by quantitative means, which can more accurately understand and predict the actual performance of the rail profile, and can identify the key areas and extreme cases of the rail profile deviation.

[0075] In an optional example of this embodiment, the sensitivity of the profile deviation of each grinding angle is calculated by using the Sobol sensitivity analysis method, the Parzen window method or the PAWN method. In particular, in the rail profile optimization design method based on the grinding index of the present invention, the sensitivity of the profile index to the profile deviation at different angles is calculated, and the Sobol sensitivity analysis method is used. The purpose is to calculate the unit deviation of the profile under different grinding angles. It is a global sensitivity analysis technology for evaluating the degree of influence of input parameter changes on the output results, and can quantify the influence of each grinding angle on the profile deviation. Therefore, the sensitivity calculation method used in this step can also be performed using methods such as the Parzen window method or the PAWN method, and can also be adjusted according to actual needs. The Parzen window method is a non-parametric probability density estimation method that can be used to evaluate the influence of the grinding angle on the profile deviation distribution. The PAWN method is a sensitivity analysis technology based on particle swarm optimization, which identifies key grinding parameters by simulating the behavior of particle swarms. The application of the above method enables the present invention to accurately identify which grinding angles have the most significant impact on the rail profile. First, the sensitivity analysis method provides a quantitative means to evaluate the impact of different grinding angles on rail profile deviation, allowing designers to more accurately understand and predict the actual performance of rail profiles. Second, by identifying key grinding parameters, it helps to optimize the grinding process, reduce rail damage caused by improper grinding, and extend the service life of rails.

[0076] In an optional embodiment of the present invention, the wheel-rail contact geometry is the wheel-rail equivalent taper, contact force, wheel-rail contact point distribution or rolling radius difference value. In particular, in the rail profile optimization design method based on the grinding index of the present invention, different rail profiles with different optimization purposes can be obtained by setting different constraints. The set constraints are that when the optimized rail profile matches the LMB10 wheel, the wheel-rail equivalent taper is in a reasonable range. The wheel-rail equivalent taper refers to the cone angle at the wheel-rail contact point, which has a direct impact on the stability and comfort of the train. When the actual purpose and requirements are different, other constraints can also be set, such as wheel-rail contact force, wheel-rail contact point distribution, rolling radius difference value, etc. The wheel-rail contact force involves the interaction force between the wheel and the rail, which is crucial to the wear of the rail and the safety of the train. The wheel-rail contact point distribution describes the position of the contact point on the rail, which affects the matching effect of the wheel and the rail. The rolling radius difference value is related to the dynamic response between the wheel and the rail. Wheel-rail contact geometry is a key factor in rail optimization design. Optimizing rail profiles can better adapt to wheel-rail contact geometry and reduce abnormal contact and wear between wheels and rails. By precisely controlling the equivalent conicity of wheels and rails, the stability and ride comfort of trains can be improved, and the shaking phenomenon can be reduced; optimized wheel-rail contact force and contact point distribution can help reduce noise and vibration and improve the train operating environment; optimized rolling radius difference values ​​can help reduce dynamic impact between wheels and rails.

[0077] In an optional embodiment of the present invention, the rail profile optimization design method based on the grinding index also includes: analyzing the rail head profile of the rail, analyzing the rail head profile into multiple arcs based on different centers, and obtaining parameters that can characterize the rail head profile based on the tangent relationship between each arc, and characterizing the optimized profile of the rail through the parameters. First, the rail head profile of the rail is geometrically analyzed and decomposed into multiple arcs based on different centers. This step involves accurate measurement and mathematical modeling of the rail head profile to determine the center position, radius size, and starting and ending points of each arc. Then, based on the tangent relationship between these arcs, the geometric connection between the arcs is determined, thereby obtaining a set of key parameters that can characterize the rail head profile. These parameters include not only the geometric dimensions of the arcs, but also the relative positions and angle relationships between them. Through these parameters, the rail head profile of the rail can be accurately described and reproduced, and the optimized rail profile can be designed accordingly. During the rail profile optimization design process, these parameters are used as the basis for adjustment and optimization to ensure that the final rail profile can meet the predetermined performance requirements and wheel-rail matching relationship.

[0078] like Figure 7As shown, the rail head profile of the 60N rail is analyzed, and the entire rail head profile including the fully polished area of ​​the working edge and the non-working edge is analyzed into multiple arcs based on different circle centers, and according to the tangent relationship between the arcs, they are characterized using the least parameters.

[0079] Set the constraints for parameter optimization calculation. The first constraint is that the optimized profile should be as adjustable as possible at each grinding angle. The second constraint is to determine the equivalent taper range of the optimized profile matching the new wheel according to the new wheel application standard.

[0080] The obtained parameters are optimized using particle swarm optimization (PSO).

[0081] By setting the initial number of particles N and the upper and lower limits of each profile parameter, the algorithm can randomly generate a group of particles in the parameter space, each particle representing a possible solution. The speed and position of the initial particles are randomly set within the range of the parameter values. The fitness value of each particle is calculated according to the constraints set in step 6.

[0082] Then the velocity and position of the particle are updated, and the update strategy is as follows:

[0083]

[0084] x i (t+1)=x i (t)+v i (t+1)

[0085] Among them, v i (t) and x i (t) represents the velocity and position of particle i at time t, w is the inertia weight; c1 and c2 are learning factors; r1 and r2 are random numbers between [0,1]; is the best historical position of particle i; g best is the global best position.

[0086] After each iteration, update the historical best position of each particle and the global optimal position g best Finally, the fitness evaluation and position and speed update steps are repeated until the preset convergence conditions are met. A set of parameter combinations that meet the constraints can be obtained, and then the final optimized profile can be obtained.

[0087] The parametric rail profile design method helps to improve the efficiency and quality of rail production, and can also quickly manufacture rails that meet design requirements through automation and digitalization.

[0088] In an optional example of this embodiment, calculating the optimized rail profile includes: optimizing the parameters by a particle swarm optimization algorithm, a genetic algorithm or an ant colony algorithm to achieve the purpose of the present invention.

[0089] Specifically, the parameters of the rail profile are optimized by particle swarm optimization algorithm, genetic algorithm or ant colony algorithm. After the rail head profile is analyzed and the characterization parameters are obtained, the intelligent optimization algorithm is used to adjust and optimize these parameters. The particle swarm optimization algorithm finds the optimal solution by simulating the social behavior of bird flocks, the genetic algorithm is based on the principles of natural selection and genetics, and the ant colony algorithm simulates the behavior of ants looking for food paths. The above algorithms can handle complex optimization problems and can adapt to multiple constraints. The rail profile parameters are optimized by the above algorithms to ensure that the optimized rail profile meets the wheel-rail contact geometry while also adapting to the on-site grinding deviation. By setting the fitness function and constraints, the algorithm can search for the optimal parameter combination to obtain the best rail optimized profile. The above algorithms can handle complex nonlinear problems, making the optimization design of the rail profile more efficient and accurate, and can handle complex nonlinear problems and find the global optimal solution. The tree-up algorithm has strong adaptability and can adapt to different design constraints and actual working conditions, which improves the flexibility and practicality of rail profile optimization.

[0090] The specific implementation process of the present invention is now described in detail with reference to the embodiments:

[0091] Step 1: Calculation of rail profile deviation

[0092] According to the deviation calculation method of rail profile given in the above steps, the deviation of the measured rail profile on a certain line is calculated, and the deviation values ​​of different angles of each measured profile can be obtained. The profile deviation results at an angle of 15° at some positions are listed, as shown in Table 7 below:

[0093] Table 7:

[0094] Mileage(Share) Profile deviation / mm Mileage(Share) Profile deviation / mm K19+630 left stock 0.296 K19+630 right share 0.034 K19+900 left stock -0.102 K19+900 right shares 0.083 K19+930 left stock -0.103 K19+930 right share 0.219 K20+100 left share -0.253 K20+100 right shares 0.037 K20+200 left stock -0.190 K20+200 right shares 0.035

[0095] Step 2: Statistical analysis of rail profile deviation

[0096] The statistical values ​​of profile deviation under different grinding angles are calculated. Taking the mean and standard deviation of some angles as examples, the statistical analysis results of profile deviation are given, as shown in Table 8:

[0097] Table 8:

[0098] Angle / ° Mean / mm Standard Deviation 0 0.018 0.080 10 0.021 0.159 15 0.092 0.199 20 0.011 0.245 25 -0.046 0.268

[0099] Kernel density is used to estimate the distribution characteristics of rail profile deviation. The specific calculation method is shown in the following formula:

[0100]

[0101] Where h represents bandwidth, which is 0.1 in this example. Here are the kernel density values ​​under some angle ranges and profile deviation ranges, as shown in Table 9:

[0102] Table 9:

[0103]

[0104]

[0105] Step 3: Calculation of Rail Profile Index Sensitivity

[0106] According to the measured profile deviation distribution characteristics and the grinding index calculation method, the sensitivity of the grinding index to the profile deviation at different angles is calculated and analyzed. The sensitivity of the profile index at some angles is shown in Table 10:

[0107] Table 10:

[0108] Grinding angle / ° Grinding index sensitivity -10 0.009 0 0.175 5 0.008 10 0.009 15 3.24E-4 20 8.99E-4

[0109] Step 4: Analyze the degree of profile adjustment at different grinding angles

[0110] The adjustable degree is calculated based on the sensitivity at each grinding angle. The adjustable degrees at some angles are listed here, as shown in Table 11 below:

[0111] Table 11:

[0112] Grinding / ° Adjustable degree -10 20.6 0 1.0 5 21.5 10 19.2 15 558.9

[0113] Step 5: Analyze the steel profile

[0114] After analysis, taking the 60N profile as an example, five parameters can be used to describe the geometric shape, namely

[0115] R1, R2, R3, O 2x ,O 3x

[0116] When performing optimization design, you only need to conduct design experiments on these five parameters, and screen and determine the final parameter combination based on the optimization constraints to obtain the optimized profile.

[0117] Step 6: Set optimization constraints

[0118] There are two constraints set for profile optimization: (1) the overall profile change at each grinding angle conforms to the profile adjustability calculated in the fourth step; (2) when the optimized profile is matched with the LMB10 wheel, the wheel-rail equivalent taper is improved by 0.11 to 0.13 compared with the 60N wheel.

[0119] Step 7: Calculate the optimized profile

[0120] Using the particle swarm optimization algorithm, the optimal profile parameter combination can be calculated. Based on this set of parameters, the optimized rail profile can be obtained. When the profile matches the LMB10, the nominal wheel-rail equivalent taper is about 0.115.

[0121] The detailed explanation of the above-mentioned embodiments is only intended to explain the present invention so as to facilitate a better understanding of the present invention. However, these descriptions cannot be interpreted as limitations on the present invention for any reason. In particular, the various features described in different embodiments may also be arbitrarily combined with each other to form other embodiments. Unless there is a clear description to the contrary, these features should be understood to be applicable to any embodiment and are not limited to the described embodiments.

Claims

1. A rail profile optimization design method based on grinding index, characterized in that: Collect multiple measured profiles of the rails after on-site grinding and acceptance; calculate the deviation values ​​of each measured profile relative to the target profile at different grinding angles; calculate the adjustable degree of the profile at different grinding angles based on the deviation values ​​at different grinding angles; set optimization constraints according to the adjustable degree of the profile and the wheel-rail contact geometry and calculate the optimized profile of the rail.

2. The rail profile optimization design method based on grinding index according to claim 1, characterized in that: Calculating the deviation value of each of the measured profiles relative to the target profile at different grinding angles includes: Calculate the normal equation of the normal line at each grinding angle; Recording the coordinates of the first intersection point of each normal line with the target silhouette; Recording the coordinates of the second intersection point of each normal line with the measured profile; The distance between the first intersection point and the second intersection point on the same normal line is calculated.

3. The rail profile optimization design method based on grinding index according to claim 1, characterized in that: The adjustable degree of the profile at different grinding angles is calculated based on the deviation values ​​at different grinding angles, including: The deviation values ​​at different grinding angles were statistically analyzed and the profile deviation distribution characteristics were obtained; Calculating the sensitivity of each grinding angle profile deviation according to the profile deviation distribution characteristics; The degree of profile adjustability at each grinding angle is calculated based on the sensitivity of the profile deviation at each grinding angle.

4. The rail profile optimization design method based on grinding index according to claim 3, characterized in that: The rail profile deviation distribution characteristics are characterized by kernel density.

5. The rail profile optimization design method based on grinding index according to claim 3, characterized in that: The distribution characteristics of rail profile deviation are characterized by probability density or quantile.

6. The rail profile optimization design method based on grinding index according to claim 3, characterized in that: The deviation values ​​at different grinding angles are statistically analyzed, including: calculating the mean and standard value of the deviation values ​​at each grinding angle.

7. The rail profile optimization design method based on grinding index according to claim 3, characterized in that: The sensitivity of each grinding angle profile deviation is calculated using the Sobol sensitivity analysis method, the Parzen window method or the PAWN method.

8. The rail profile optimization design method based on grinding index according to claim 1, characterized in that: The wheel-rail contact geometric relationship is the wheel-rail equivalent taper, wheel-rail contact force, wheel-rail contact point distribution or rolling radius difference value.

9. The rail profile optimization design method based on grinding index according to claim 1, characterized in that: The rail profile optimization design method based on the grinding index also includes: The rail head profile of the rail is analyzed to be a plurality of arcs based on different centers of circles, and parameters capable of characterizing the rail head profile are obtained according to the tangent relationship between the arcs, and the optimized profile of the rail is characterized by the parameters.

10. The rail profile optimization design method based on grinding index according to claim 9, characterized in that: Calculation of optimized rail profile, including: The parameters are optimized by particle swarm optimization algorithm, genetic algorithm or ant colony algorithm.

Images