A high-speed railway fine-tuning optimization method adapted to multiple working conditions

Abstract

The present invention discloses a high-speed railway fine-tuning optimization method suitable for multiple working conditions, including S1, obtaining basic data of high-speed railway ballastless track fine-tuning; S2, establishing a high-speed railway ballastless track double-track multi-objective optimization model; S3, solving the double-track adjustment amount based on the interior point method; S4, obtaining the fastener adjustment level. The method of the present invention takes the fastener adjustment amount as a decision variable, and takes the fastener adjustable margin, the long-medium-short wave unevenness of the double track, the track gauge, the horizontal and other unevenness indicators as constraints. On this basis, the objective function is established by linearly weighting the three objectives of "minimum adjustment amount, linear regression, and track smoothness", so as to achieve the optimization of track adjustment amount suitable for multiple working conditions, reduce the use of fastener accessories during the fine-tuning process, achieve the double-track fine-tuning of the track in one go, and simplify the track fine-tuning process.

Description

Technical Field

[0001] The present invention relates to the field of high-speed railway fine-tuning, and in particular to a high-speed railway fine-tuning optimization method adapting to multiple working conditions. Background Art

[0002] High-speed rail ballastless track fine-tuning is an important part of the opening of new lines and the daily maintenance of operating lines. Its effect directly affects the safety and comfort of train driving. At present, ballastless track fine-tuning mainly relies on manual fine-tuning scheme design based on human experience. Based on the premise of obtaining the lateral and elevation deviations of the track by the difference between the actual track coordinates obtained by track measurement and the linear coordinates of the designed track, the fine-tuning work first carries out the first round of adjustment with the goal of eliminating the 30m and 300m long wave unevenness of the reference rail. Subsequently, at least three rounds of adjustment are carried out in turn to control the unevenness of the reference rail and non-reference rail, gauge and level parameters. During the adjustment process, it is easy to encounter situations such as the adjustment amount exceeding the adjustable margin of the fastener and the static TQI not meeting the acceptance standard, which requires rework, resulting in a large waste of fastener accessories during the fine-tuning process and delaying the construction period and skylight time.

[0003] During the track fine-tuning process, it is necessary to control the static unevenness index of the track in accordance with the specifications. However, for different working conditions and different track structures, the fine-tuning strategies are different. In the track fine-tuning work after the stress release of the long rails of the newly built line, since the fasteners used are all standard parts, the adjustable margin of each fastener is the same, which is conducive to the optimization of the line shape and smoothness. Therefore, the main consideration is to control the overall adjustment amount of the fasteners to reduce the types of fastener accessories adjusted and the use and consumption of fasteners; the operating lines mainly carry out track fine-tuning during the overhaul process. Due to years of track fine-tuning maintenance, the adjustable margin range of each fastener is inconsistent and has great volatility. Therefore, the adjustment is mainly based on line shape and smoothness control, and the control of the fastener adjustment amount is relaxed. At this stage, the main problems of track fine-tuning are as follows:

[0004] (1) Manual fine-tuning scheme design cannot provide the fine-tuning amount of the double tracks at one time, resulting in more overall steps for fine-tuning high-speed rail ballastless tracks and a complicated process.

[0005] (2) During the design of the fine-tuning scheme, only the track smoothness requirements were considered, and the constraints of the adjustable margin of the fasteners were not considered.

[0006] (3) The design of the fine-tuning scheme only considers the principle of minimum adjustable amount, and does not consider the differences in design objectives under different working conditions.

[0007] Track fine-tuning requires comprehensive consideration of multiple objective factors, so an automated track fine-tuning algorithm that can adapt to a variety of working conditions is needed. Summary of the invention

[0008] In order to solve the problems existing in the prior art, the present invention provides a high-speed railway fine-tuning optimization method that is adaptable to multiple working conditions. Through this optimization method, the high-speed railway ballastless track fine-tuning solution can be implemented at one time, thereby improving the fine-tuning efficiency.

[0009] To this end, the present invention adopts the following technical solutions:

[0010] A high-speed railway fine-tuning optimization method adapted to multiple working conditions specifically comprises the following steps:

[0011] S1, obtain the basic data for fine-tuning the high-speed rail ballastless track:

[0012] S11, obtaining the deviation value of the segment to be fine-tuned Wherein, k = L indicates the left rail, k = R indicates the right rail, w = x indicates the horizontal direction, and w = y indicates the elevation; i is the sleeper number, 0≤i≤n, and n is the total number of sleepers in the section to be fine-tuned;

[0013] S12, determine the decision variable vector

[0014] First, obtain the fastener adjustment amount and adjustable margin; the fastener adjustment amount on sleeper i is The upper and lower limits of the adjustable allowance of the fastener are l represents the lower limit of the adjustable margin, and u represents the upper limit of the adjustable margin;

[0015] Then, construct the decision variable vector There is the following formula:

[0016]

[0017] S2, establish a multi-objective optimization model for high-speed railway ballastless double-track:

[0018] S21, constructing constraint conditions for the adjustable margin of the fastener;

[0019] S22, construct the constraints of the track acceptance criteria:

[0020] When fine-tuning the transverse direction of the section to be fine-tuned, the constraints of the track acceptance standard are specifically the track direction, 30m sag difference, 300m sag difference and track gauge;

[0021] When fine-tuning the elevation of the section to be fine-tuned, the constraints of the track acceptance standard are specifically height, 30m sag difference, 300m sag difference and level;

[0022] The track direction and height belong to the 10m midpoint chord;

[0023] S23, determine the objective function

[0024] S3, solve the dual-track adjustment based on the interior point method:

[0025] S31, using the interior point method to transform all the constraints and objective functions constructed in S2 To unconstrained repair

[0026] Positive objective function

[0027] S32, for the Construct optimization constraints and optimization objective function

[0028] S33, using the interior point method to transform the optimization constraints and optimization objective function in S32 into an unconstrained modified optimization objective function

[0029] S34, using Newton's method to solve the The obtained results are taken as The initial solution of

[0030] S35, using the initial solution obtained in S34 and Newton's method to solve the The optimal solution

[0031] S4, obtain the fastener adjustment level:

[0032] The fastener grade is calculated by the following formula

[0033]

[0034] Among them, μ is the level difference; after the value of the fastener level is obtained, it is used to fine-tune the railway in the section to be fine-tuned.

[0035] Constraints on the adjustable allowance of fasteners in S21 As shown below:

[0036]

[0037] The constraint condition of the 10m midpoint chord in S22 is as follows:

[0038]

[0039] Among them, i∈[8,n-8], is the value of the 10m midpoint chord corresponding to sleeper i, σ 10 is the limit value of the 10m midpoint chord of sleeper i in the design document;

[0040] The constraint condition of the 30m vector difference in S22 is as follows:

[0041]

[0042] in, is the deviation value of the starting point of the 30m chord (0≤m <i<i+8<m+48≤n), is the 30m vector difference corresponding to sleeper node i, σ 30 It is the limit value of 30m vector difference in the design document;

[0043] The constraint condition of 300m vector difference in S22 is as follows:

[0044]

[0045] in, is the deviation value of the starting point of the 300m chord (0≤j <i<i+240<j+480≤n), is the 300m vector difference corresponding to sleeper i, σ 300 It is the limit value of 300m vector difference in the design document;

[0046] The track gauge or level constraint in S22 is as follows:

[0047]

[0048] in, is the gauge or horizontal constraint function of sleeper i, σ val is the limit value of track gauge or level in the design document; The result has no effect, so no matter whether the left or right track is calculated, The values ​​are the same.

[0049] Objective function in S23 As shown below:

[0050]

[0051] in, is a linear regression subfunction, which is used to control the fine-tuned linear shape not to deviate from the original design linear shape; for

[0052] The fastener adjustment amount sum sub-function is used to control the fastener adjustment amount; is the linear smoothness subfunction, which is used to control the smoothness of the trajectory line after fine-tuning; α c is the weight coefficient, which is set according to the working conditions of the section to be fine-tuned.

[0053] Unconstrained modified objective function in S31 As shown below:

[0054]

[0055] in, is the value of the 10m midpoint chord corresponding to sleeper i, is the 30m vector difference corresponding to sleeper node i, is the 300m vector difference corresponding to sleeper i, λ is the first barrier multiplier; the λ is used to

[0056] Correct the above mentioned

[0057] The optimization objective function is The z is a construction optimization variable; the optimization constraint condition is as follows:

[0058]

[0059] Unconstrained Modified Optimization Objective Function in S33 As shown below:

[0060]

[0061] Where γ is the second obstacle multiplier; the γ is used to correct the

[0062] S34 specifically includes the following steps:

[0063] S341, construct solution conditions:

[0064] First, set S33 The solution is Where t is the second iteration number. When t = 0, it is the initial solution without iteration and is set to:

[0065]

[0066] Then, set the barrier multiplier γ0 = 10 and the second iteration error threshold and the second barrier multiplier error threshold τ2 = 10 -3 ;

[0067] S342, calculate and obtain the second search direction vector

[0068]

[0069] in, is the Laplace operator, is the Hamiltonian operator;

[0070] S343, yes The norm of is used to judge:

[0071] like Then proceed to S344, otherwise update the second barrier multiplier so that γ t+1 =γ t / 10, at this time, if γ t+1 When ≤τ2, the solution is terminated, indicating that there is no solution that satisfies the constraint conditions. The constraint conditions in S2 are adjusted and the method is executed again; otherwise, the method returns to S342;

[0072] S344. Calculate by the following formula

[0073]

[0074] If t+1 ≤0, then the solution ends, and z t+1 Corresponding Then it is the The initial solution

[0075] Otherwise, use Return to step S342 and continue iterative calculation until the initial solution is obtained.

[0076] S35 includes the following steps:

[0077] S351, construct solution conditions:

[0078] Newton's method is used to iteratively calculate the Solution The value of S34 is obtained As of

[0079] Initial value, s is the first iteration number, is the second search direction vector; the first obstacle multiplier λ is set to 10, and the first iteration error threshold is The first barrier multiplier error threshold θ2 = 10 -3 ;

[0080] S352, calculate and obtain the second search direction vector

[0081]

[0082] S353, for the The norm of is used to judge:

[0083] if Then calculate and use Return to S352 to continue iteration, otherwise update the first barrier multiplier so that λ s+1 =λ s / 10, if λ s+1 >θ2, then use λ s+1 Return to S352 to continue iteration; if λ s+1 ≤θ2, then is the optimal solution and is marked as

[0084] Preferably, the The optimal solution For the The minimum value in the solution.

[0085] Preferably, the step difference μ in S4 is 0.5 mm or 1 mm.

[0086] Compared with the prior art, the present invention has the following beneficial effects:

[0087] 1. The optimization method of the present invention takes into account the correlation between the two rails during the calculation process. The optimization result can give the optimal adjustment amount of the reference rail and the non-reference rail at one time, simplifying the existing fine-tuning process of "reference rail first, then non-reference rail", and can achieve one-time fine-tuning, effectively shortening the project construction period and skylight time.

[0088] 2. The optimization method of the present invention takes into account the three optimization goals of minimum adjustment amount, linear regression, and track smoothness. It can be applied to track fine-tuning work under working conditions such as construction of new high-speed railway lines and overhaul of operating lines. It has strong adaptability and provides a high degree of flexibility for the design of fine-tuning schemes.

[0089] 3. The optimization method of the present invention comprehensively considers constraints such as the adjustable margin of fasteners and the acceptance conditions for track smoothness. While effectively improving the smoothness of the track, it avoids rework of the design solution due to the adjustment amount exceeding the adjustable margin of fasteners. BRIEF DESCRIPTION OF THE DRAWINGS

[0090] Figure 1 is a flow chart of the method of the present invention;

[0091] Figure 2 is a flow chart of step S34 in the method of the present invention;

[0092] Figure 3 is a flow chart of step S35 in the method of the present invention;

[0093] Figure 4 is a diagram showing the relationship between the left and right adjustment amounts and the fastener constraint amounts after optimization in an embodiment of the present invention;

[0094] Figure 5 This is a diagram comparing the height values ​​of the left and right rails before and after optimization in an embodiment of the present invention;

[0095] Figure 6 This is a comparison diagram of the 30m vector distance difference between the front and rear left and right tracks optimized in an embodiment of the present invention;

[0096] Figure 7 This is a comparison diagram of the 300m vector distance difference between the front and rear left and right tracks optimized in an embodiment of the present invention. DETAILED DESCRIPTION

[0097] S1, obtain the basic data for fine-tuning the high-speed rail ballastless track:

[0098] S11, obtain the track deviation value:

[0099] High-speed rail ballastless track measurement mainly uses the track inspection instrument to perform joint measurement on the CPIII points of the track section to be fine-tuned to obtain the three-dimensional coordinates of the track section to be fine-tuned, and then obtains the deviation value of the sleeper i by comparing the coordinates with the center line of the designed line. k = L indicates the left rail, k = R indicates the right rail, w = x indicates the horizontal direction, and w = y indicates the elevation; i is the sleeper number, 0≤i≤n, n is the total number of sleepers in the section to be fine-tuned; the design line centerline is the line centerline in the design file;

[0100] For the deviation value The total lateral deviation of the left rail of sleeper i is Right rail lateral deviation Left track elevation deviation And the right track elevation deviation value Four situations.

[0101] S12, determine the decision variable vector

[0102] Since the fine-tuning of high-speed railway ballastless tracks is mainly achieved by adjusting fasteners, the adjustable range of fasteners is limited. In particular, for high-speed railway ballastless tracks that have been in operation for many years, the fasteners have been adjusted repeatedly over the years, and the adjustable range of each fastener is different. With the development of hardware technologies related to track gauges, by integrating structured light scanning modules on track gauges, it is possible to measure the relative position relationship between the rails and the rail supports or sleepers, thereby indirectly obtaining the adjustable margin of the fasteners. By establishing a "one sleeper, one file" fastener accessory file, the remaining adjustable margin of the fasteners can also be obtained.

[0103] Get the fastener adjustment amount v and adjustable margin σ; the fastener adjustment amount on sleeper i is The upper and lower limits of the adjustable allowance of each fastener are l represents the lower limit of the adjustable margin, and u represents the upper limit of the adjustable margin.

[0104] Constructing the decision variable vector There is the following formula:

[0105]

[0106] Therefore, the horizontal decision variable vector is The decision variable vector for elevation is

[0107] S2, establish a multi-objective optimization model for high-speed railway ballastless double-track:

[0108] Since the lateral and elevation fine-tuning models of the ballastless track are consistent, then:

[0109] When w=x, the transverse fine adjustment mainly considers the track direction (10m midpoint chord), 30m vector difference, 300m vector difference and track gauge;

[0110] When w=y, the elevation fine-tuning mainly considers the height (10m midpoint chord), 30m vector distance difference, 300m vector distance difference and level. Moreover, the calculation method of the level is consistent with the calculation method of the track gauge. Since there is no correlation between the fine-tuning of the lateral direction and the elevation, it is possible to decouple and separate the fine-tuning of the lateral direction and the elevation to reduce the number of decision variables and improve the speed of calculation convergence. Specifically, the following steps are included:

[0111] S21, construct the constraint conditions for the adjustable allowance of fasteners

[0112] Through S12 Adjustment of fasteners on sleeper i To constrain, we have the following formula:

[0113]

[0114] S22, construct the constraints of the track acceptance criteria:

[0115] The track acceptance standard constraints are specifically the constraints on the 10m midpoint chord (track direction or height), 30m sag difference, 300m sag difference and track gauge (horizontal), including the following steps:

[0116] S221, construct the constraint condition of the 10m midpoint chord, and adjust the fastener on the sleeper i There is the following formula:

[0117]

[0118] Among them, i∈[8,n-8], the 10m midpoint chord is an indicator of track direction or height during static detection, is the value of the 10m midpoint chord corresponding to sleeper i, σ10 It is the limit value of the 10m midpoint chord of sleeper i in the design document.

[0119] S222, construct the constraint condition of 30m vector difference, and adjust the fastener on sleeper i There is the following formula:

[0120]

[0121] in, is the deviation value of the starting point of the 30m chord (0≤m <i<i+8<m+48≤n), is the 30m vector difference corresponding to sleeper node i, σ 30 This is the limit value of the 30m vector distance difference in the design document.

[0122] S223, construct the constraint condition of 300m vector difference, and adjust the fastener on sleeper i There is the following formula:

[0123]

[0124] in, is the deviation value of the starting point of the 300m chord (0≤j <i<i+240<j+480≤n), is the 300m vector difference corresponding to sleeper i, σ 300 This is the limit value of the 300m vector distance difference in the design document.

[0125] S224, construct the constraint condition of track gauge (horizontal), which is as follows:

[0126]

[0127] in, is the gauge or horizontal constraint function of sleeper i, σ val is the limit value of track gauge or level in the design document; The result has no effect, so whether calculating the left or right track, The values ​​are the same.

[0128] S23, determine the objective function:

[0129] The objectives of track fine-tuning are different under different working conditions, so different optimization objective sub-functions The multi-objective value optimization is realized by linear weighting. Considering the requirements for the optimal solution under different working conditions, the objective function is obtained There is the following formula:

[0130]

[0131] in:

[0132] f1( (w) v0 is a linear regression subfunction, which represents the degree of convergence between the fine-tuned track shape and the track shape in the design file, and is used to control the fine-tuned track shape not to deviate from the original design shape;

[0133] is the total sub-function of the fastener adjustment amount, which represents the size of the total amount of fastener adjustment after fine-tuning and is used to control the size of the fastener adjustment amount;

[0134] It is a linear smoothness sub-function, which characterizes the curvature change of the track line after fine-tuning. It is used to control the smoothness of the track line after fine-tuning to avoid the distortion of the track line.

[0135] α c is the weight coefficient, which can be set according to different working conditions.

[0136] In actual situations, when the lateral adjustment of the railway line is made, the corresponding objective function is When the elevation of the railway line is adjusted, the corresponding objective function is

[0137] S3, solve the dual-track adjustment based on the interior point method:

[0138] S31, constructing an unconstrained modified objective function based on the interior point method

[0139] Use the interior point method to convert the constraints and objective function constructed in S2 into an unconstrained modified objective function And introduce the first barrier multiplier λ, we have the following formula:

[0140]

[0141] S32, construct optimization constraints and optimization objective function:

[0142] Due to too many constraints, it is impossible to directly solve the initial solution of equation (8) that satisfies all the constraints. Therefore, it is necessary to construct new optimization constraints and optimization objective functions to solve:

[0143] Optimization objective function:

[0144] Optimization constraints:

[0145] Among them, z is the construction optimization variable;

[0146] S33, using the interior point method, transforms equations (9) and (10) in S32 into an unconstrained modified optimization objective function And introduce the second barrier multiplier γ:

[0147]

[0148] Among them, due to the value of k for The result has no effect, so there is no need to use the left and right rails in equations (8) and (11). The results are summed up;

[0149] S34, such as Figure 2 As shown, Newton's method is used to solve The initial solution The following steps are involved:

[0150] S341, construct solution conditions:

[0151] First, assume that the solution of equation (11) is Where t is the second iteration number. When t = 0, it is the initial solution without iteration.

[0152] Then, set the barrier multiplier γ0 = 10 and the second iteration error threshold and the second barrier multiplier error threshold τ2 = 10 -3 ;

[0153] S342, calculate and obtain the second search direction vector

[0154]

[0155] in, Used for iterative solution; is the Laplace operator, is the Hamiltonian operator.

[0156] S343, yes Make a judgment:

[0157] like Then enter S344; otherwise update the second barrier multiplier so that γ t+1 =γ t / 10, at this time, if γ t+1 ≤τ2, the solution ends, indicating that there is no solution that satisfies the constraints. The constraints in S2 need to be adjusted. Otherwise, return to S342 until Enter S344;

[0158] S344 is calculated by the following formula

[0159]

[0160] After getting the result, if z t+1 ≤0, then the solution ends, and z t+1 Corresponding Then it is the initial solution of the modified objective function

[0161] Otherwise, use Return to step S342 and continue iterative calculation until the initial solution is obtained.

[0162] S35, using Newton's method The minimum value of the solution includes the following steps:

[0163] S351, construct solution conditions:

[0164] Newton's method is also used to iteratively calculate the corrected objective function Solution The value of S34 is obtained As The initial value of , s is the first iteration number, is the first search direction vector.

[0165] like Figure 3 As shown in the figure, considering that the deviation and the fastener adjustment are both in mm, integer optimization is required in the future, that is, the optimal result can be retained to one decimal place, the first obstacle multiplier λ is set to 10, and the first iteration error threshold is The first barrier multiplier error threshold θ2 = 10 -3 ;

[0166] S352, calculate and obtain the first search direction vector

[0167]

[0168] S353, yes The norm of is used to judge:

[0169] calculate norm, if Then calculate and use Return to S352 to continue iteration, otherwise update the first barrier multiplier so that λ s+1 =λ s / 10, if λ s+1 >θ2, then use λ s+1 Return to S352 to continue iteration; if λ s+1 ≤θ2, then is the optimal solution and is marked as The optimal solution at this time for The minimum value of the solution.

[0170] From the above steps, we can see that the horizontal optimal solution is The optimal solution for elevation is

[0171] S4, obtain the fastener adjustment level:

[0172] The fine adjustment of the railway is achieved by adjusting the fasteners, and the adjustment of the fasteners is mainly adjusted by the gauge block, height adjustment pad and other accessories of the fasteners. These adjustment parts can only be adjusted according to the step difference, and the step difference μ is generally 0.5mm or 1mm. The final fastener adjustment level is obtained by rounding off the optimal solution calculated by S353. Specifically, the horizontal fastener adjustment level is The elevation fastener adjustment level is

[0173] Example 1

[0174] This embodiment performs fine adjustment on the horizontal direction of the section to be fine-tuned, as follows:

[0175] The section of a high-speed railway ballastless track to be fine-tuned is about 1.9 km long, with a total of 3,000 sleepers. The elevation adjustment of the track is optimized. In the objective function of S23, the linear regression weight coefficient α1 is 1×10 -4 , the weight coefficient α2 of the adjustment function is 1×10 -5 , the weight coefficient of the smoothing subfunction α3 is 1. The lateral adjustable margin of the fastener is measured by the structured light track detector, and the track irregularity parameter is σ 10 =0.3,σ 30 =0.3,σ 300 =5,σ val =1, their units are all mm; the step difference μ is selected as 0.5mm.

[0176] After optimization, the adjustment amounts of the left and right rails are within the adjustable margin range, such as Figure 4 shown.

[0177] After using the method of the present invention, the maximum and minimum values ​​of the height of the left and right rails before optimization are [-4.7mm, 3.6mm]. After optimization with a 0.5mm step difference, the maximum and minimum values ​​of the height converge to [-0.48mm, 0.49mm]. Figure 5 shown.

[0178] Before optimization, the maximum and minimum values ​​of the 30m chord-vector difference of the left and right rails are [-5.0mm.4.6mm]. After optimization with 0.5mm step difference, the maximum and minimum values ​​of the 30m chord-vector difference converge to [-0.58mm, 0.62mm]. Figure 6 shown.

[0179] Before optimization, the maximum and minimum values ​​of the 300m chord-vector difference of the left and right rails are [-6.0mm, 4.8mm]. After optimization with 0.5mm step difference, the maximum and minimum values ​​of the 300m chord-vector difference converge to [-1.5mm, 1.5mm]. Figure 7 shown.

[0180] By comparing the above embodiments, it can be seen that the optimization calculation method of the present invention can significantly improve the track smoothness within the adjustable range of the fasteners.

Claims

1. A high-speed railway fine-tuning optimization method adapted to multiple working conditions, characterized in that: The specific steps include: S1, obtain the basic data for fine-tuning the high-speed rail ballastless track: S11, obtaining the deviation value of the segment to be fine-tuned Wherein, k = L indicates the left rail, k = R indicates the right rail, w = x indicates the horizontal direction, and w = y indicates the elevation; i is the sleeper number, 0≤i≤n, and n is the total number of sleepers in the section to be fine-tuned; S12, determine the decision variable vector First, obtain the fastener adjustment amount and adjustable margin; the fastener adjustment amount on sleeper i is The upper and lower limits of the adjustable allowance of the fastener are l represents the lower limit of the adjustable margin, and u represents the upper limit of the adjustable margin; Then, construct the decision variable vector There is the following formula: S2, establish a multi-objective optimization model for high-speed railway ballastless double-track: S21, constructing constraint conditions for the adjustable margin of the fastener; S22, construct the constraints of the track acceptance criteria: When fine-tuning the transverse direction of the section to be fine-tuned, the constraints of the track acceptance standard are specifically the track direction, 30m sag difference, 300m sag difference and track gauge; When fine-tuning the elevation of the section to be fine-tuned, the constraints of the track acceptance standard are specifically height, 30m sag difference, 300m sag difference and level; The track direction and height belong to the 10m midpoint chord; S23, determine the objective function S3, solve the dual-track adjustment based on the interior point method: S31, use the interior point method to transform the constraints and objective function constructed in S2 Convert to unconstrained modified objective function S32, for the Construct optimization constraints and optimization objective functions; S33, using the interior point method to transform the optimization constraints and optimization objective function in S32 into an unconstrained modified optimization objective function S34, using Newton's method to solve the The obtained results are taken as The initial solution of S35, using the initial solution obtained in S34 and Newton's method to solve the The optimal solution S4, obtain the fastener adjustment level: The fastener grade is calculated by the following formula Among them, μ is the level difference; after the value of the fastener level is obtained, it is used to fine-tune the railway in the section to be fine-tuned.

2. The high-speed railway fine-tuning optimization method adapted to multiple working conditions according to claim 1 is characterized in that: Constraints on the adjustable allowance of fasteners in S21 For adjustment The deviation function from the adjustable margin is shown in the following formula: The constraint condition of the 10m midpoint chord in S22 is as follows: Among them, i∈[8,n-8], is the value of the 10m midpoint chord corresponding to sleeper i, σ 10 is the limit value of the 10m midpoint chord of sleeper i in the design document; The constraint condition of the 30m vector difference in S22 is as follows: in, is the deviation value of the starting point of the 30m chord (0≤m <i<i+8<m+48≤n), is the 30m vector difference corresponding to sleeper node i, σ 30 It is the limit value of 30m vector difference in the design document; The constraint condition of 300m vector difference in S22 is as follows: in, is the deviation value of the starting point of the 300m chord (0≤j <i<i+240<j+480≤n), is the 300m vector difference corresponding to sleeper i, σ 300 It is the limit value of 300m vector difference in the design document; The track gauge or level constraint in S22 is as follows: in, is the gauge or horizontal constraint function of sleeper i, σ val is the limit value of track gauge or level in the design document; The result has no effect, so no matter whether the left or right track is calculated, The values ​​are the same.

3. The high-speed railway fine-tuning optimization method adapted to multiple working conditions according to claim 1 is characterized in that: Objective function in S23 As shown below: in, is a linear regression subfunction, which is used to control the fine-tuned linear shape not to deviate from the original designed linear shape; It is the sub-function of summing the fastener adjustment amount, which is used to control the size of the fastener adjustment amount; is the linear smoothness subfunction, which is used to control the smoothness of the trajectory line after fine-tuning; α c is the weight coefficient, which is set according to the working conditions of the section to be fine-tuned.

4. The high-speed railway fine-tuning optimization method adapted to multiple working conditions according to claim 1 is characterized in that: Unconstrained modified objective function in S31 As shown below: in, is the value of the 10m midpoint chord corresponding to sleeper i, is the 30m vector difference corresponding to sleeper node i, is the 300m vector difference corresponding to sleeper i, λ is the first obstacle multiplier; the λ is used to correct the 5. The high-speed railway fine-tuning optimization method adapted to multiple working conditions according to claim 4 is characterized in that: The optimization objective function is The z is a construction optimization variable; the optimization constraint condition is as follows: Unconstrained Modified Optimization Objective Function in S33 As shown below: Where γ is the second obstacle multiplier; the γ is used to correct the 6. The high-speed railway fine-tuning optimization method adapted to multiple working conditions according to claim 5 is characterized in that: S34 specifically includes the following steps: S341, construct solution conditions: First, set S33 The solution is Where t is the second iteration number. When t = 0, it is the initial solution without iteration and is set to: Then, set the barrier multiplier γ0 = 10 and the second iteration error threshold and the second barrier multiplier error threshold τ2 = 10 -3 ; S342, calculate and obtain the second search direction vector in, is the Laplace operator, is the Hamiltonian operator; S343, yes The norm of is used to judge: like Then proceed to S344, otherwise update the second barrier multiplier so that γ t+1 =γ t / 10, at this time, if γ t+1 When ≤τ2, the solution is terminated, indicating that there is no solution that satisfies the constraint conditions. The constraint conditions in S2 are adjusted and the method is executed again; otherwise, the method returns to S342; S344. Calculate by the following formula If t+1 ≤0, then the solution ends, and z t+1 Corresponding Then it is the The initial solution Otherwise, use Return to step S342 and continue iterative calculation until the initial solution is obtained.

7. The high-speed railway fine-tuning optimization method adapted to multiple working conditions according to claim 5 is characterized in that: S35 includes the following steps: S351, construct solution conditions: Newton's method is used to iteratively calculate the Solution The value of S34 is obtained As The initial value of , s is the first iteration number, is the second search direction vector; the first obstacle multiplier λ is set to 10, and the first iteration error threshold is The first barrier multiplier error threshold θ2 = 10 -3 ; S352, calculate and obtain the second search direction vector S353, for the The norm of is used to judge: if Then calculate and use Return to S352 to continue iteration, otherwise update the first barrier multiplier so that λ s+1 =λ s / 10, if λ s+1 >θ2, then use λ s+1 Return to S352 to continue iteration; if λ s+1 ≤θ2, then is the optimal solution and is marked as 8. The high-speed railway fine-tuning optimization method adapted to multiple working conditions according to claim 1 is characterized in that: Said The optimal solution For the The minimum value in the solution.

9. The high-speed railway fine-tuning optimization method adapted to multiple working conditions according to claim 1 is characterized in that: The step difference μ in S4 is 0.5mm or 1mm.

Images