Route optimization method for new railway parallel to existing lines based on reliability model

Abstract

The present invention relates to the technical field of line optimization, and provides a line optimization method for a new railway parallel to an existing line based on a reliability model, which includes: 1. establishing a system reliability model; 2. finite element analysis; 3. reliability and probability analysis; setting the allowable settlement limit of the bridge and the roadbed according to the engineering specification; using the reliability analysis module of the software, combined with the Monte Carlo simulation method, identifying the key influencing factors causing the settlement of the adjacent bridge and roadbed, and performing reliability analysis; applying the BNN proxy model to perform probability analysis on the finite element model; based on the analysis results, evaluating the reliability of the adjacent existing railway bridge and roadbed; 4. optimizing the railway line in combination with the reliability analysis results. The present invention can better optimize the railway line.

Description

Technical Field

[0001] The invention relates to the technical field of line optimization, and in particular to a line optimization method for a new railway parallel to an existing line based on a reliability model. Background Art

[0002] In addition to the construction risks, the parallel construction of new railways and existing lines also presents a series of other problems. The construction process of the new line, such as foundation excavation, piling and filling, will disturb the existing line foundation adjacent to it, thereby reducing its strength and weakening its stability, which has a direct impact on the driving comfort and safety of the train. During the construction of the new line adjacent to the existing line railway, the deadweight of the bridge, piers and road foundation will cause the soil around the pile foundation to sink. This additional settlement will have an adverse effect on the existing line railway, especially the uneven deformation of the ballastless track caused by local settlement, which seriously threatens the operational safety of the existing railway. It is worth noting that this additional settlement value gradually decreases with the increase of the distance between the two lines. However, increasing the line spacing will lead to an increase in the sandwiched land between the two lines and aggravate the demolition problem. Therefore, research is needed to determine the minimum safe distance between the two lines under the premise of being affected by settlement.

[0003] Research shows that railway subgrades are different from structural engineering that is susceptible to material aging. Under the long-term action of trains, the foundation is consolidated and the subgrade fill is compacted, making the subgrade in good working condition. However, disturbances to the subgrade by the external environment may cause adverse changes to the subgrade. The impact of adjacent construction may cause deformation of the operating railway beyond the limit. Although the subgrade has not become unstable or damaged, it also threatens driving safety. Therefore, it is necessary to study the service status of the railway under the special condition of construction disturbance.

[0004] There are relatively few studies on the impact of different construction stages on the existing line foundation in the process of new line construction. It is difficult to simulate infinite foundation dynamic damping using model tests, and field tests are expensive and difficult to monitor operating lines. Therefore, numerical analysis methods are very important. Summary of the invention

[0005] The present invention provides a line optimization method for a new railway parallel to an existing line based on a reliability model, which can overcome certain or some defects of the prior art.

[0006] According to the present invention, a line optimization method for a new railway parallel to an existing line based on a reliability model comprises the following steps:

[0007] 1. Establish a system reliability model;

[0008] Determine and evaluate the settlement and horizontal displacement of the existing railway bridge and roadbed nearby; build a system reliability model based on engineering specifications and actual conditions;

[0009] 2. Finite element analysis;

[0010] Use engineering software (Abaqus) to build finite element models of existing railway bridges and roadbeds; simulate the construction process of new railways through the model and analyze its impact on adjacent structures; build a two-dimensional cross-section model to determine the settlement load effect caused by the construction of new railways; calculate the maximum allowable settlement R of its structure 1 and horizontal displacement R 2 ;

[0011] 3. Reliability and probability analysis;

[0012] Set the allowable settlement limits of bridges and roadbeds according to engineering specifications; use the reliability analysis module of Abaqus software, combined with the Monte Carlo simulation method, to identify the key factors that cause the settlement of adjacent bridges and roadbeds, and conduct reliability analysis; use the BNN proxy model to perform probabilistic analysis on the finite element model; based on the analysis results, evaluate the reliability of the adjacent existing railway bridges and roadbeds;

[0013] 4. Optimize railway lines based on reliability analysis results.

[0014] Preferably, in step one, specifically:

[0015] For the two parameters of settlement and horizontal displacement, the limit state functions are defined as:

[0016] o 1 (x) = R 1 -S 1

[0017] o 2 (x) = R 2 -S 2

[0018] Where:

[0019] o 1 (x), o 2 (x)—limit state functions of settlement and horizontal displacement, respectively;

[0020] R 1 , R 2 — respectively, the safe allowable values ​​of settlement and horizontal displacement;

[0021] S 1 , S 2 —are the actual settlement and horizontal displacement under specific conditions, respectively.

[0022] The series system reliability model is used because the existing adjacent railway is considered non-compliant as long as any one of the settlement or horizontal displacement exceeds the standard. Therefore, the system non-compliance probability P nc It can be expressed as:

[0023] P nc =P(o 1 (x)≤0∪o 2 (x)≤0)

[0024] P nc =P nc1 +P nc2 -P nc1nc2

[0025] Where:

[0026] P nc — is the probability of non-compliance of the system, indicating that when any limit state function (settlement or horizontal displacement) does not meet the conditions, the system is considered to be non-compliant. This model allows the assessment of the overall safety of the system under given conditions.

[0027] P nc1 — is the probability of non-compliance of the settlement pattern;

[0028] P nc2 — is the non-compliance probability of horizontal displacement;

[0029] P nc1nc2 — is the joint non-compliance probability of settlement and horizontal displacement occurring simultaneously.

[0030] Preferably, in the two-dimensional cross-section model of step 2, the following are set:

[0031] (1) The soil is an elastic material and the Mohr-Coulomb model is used;

[0032] (2) The initial stress field of the roadbed and foundation is generated by the deadweight load;

[0033] (3) The new roadbed is reinforced with piles, the pile-soil contact condition is partial sliding, and the contact interface is simulated by the slip coefficient;

[0034] (4) The new roadbed is filled in layers, and each layer has good continuity;

[0035] (5) The left and right boundaries of the foundation are impermeable, and the bottom and top are permeable boundaries; the bottom is completely fixed and constrained, and sliding constraints are imposed on the vertical boundaries on both sides.

[0036] Preferably, in step three, in the finite element model, the soil parameters are subjected to Monte Carlo simulation to construct a high-fidelity finite element model.

[0037] Preferably, in step 3, the Monte Carlo simulation method is:

[0038] First, multiple simulation scenarios were created, each containing a different combination of existing railway lines and construction conditions. Next, a large number of random simulations were performed on each scenario in the Abaqus finite element analysis software to generate settlement and horizontal displacement data for the existing railway lines under different conditions. These data were then analyzed to determine the frequency with which the existing railway lines remained within the safe allowable range in each scenario, i.e., their reliability.

[0039] Preferably, in step 3, the reliability analysis method combining the Monte Carlo simulation method with the finite element analysis is specifically as follows:

[0040] 1) Initialize reliability analysis variables;

[0041] 2) Establish a solid model and divide the unit, including setting the unit type, real constant, and material. Parametric modeling must be performed during the process;

[0042] 3) Define the analysis type and corresponding options, apply loads, determine load step options, and call the function for initial solution;

[0043] 4) Simulate the construction process and solve the problem step by step;

[0044] 5) After the solution is completed, the corresponding calculation results are extracted in post-processing and the values ​​are assigned to the parameters designated as input variables and output result variables.

[0045] 6) Conduct reliability analysis, including: selecting and defining input variables and the correlation coefficients between input variables; determining the distribution type, distribution function and its parameters of each input variable; specifying the output result variable; selecting analysis tools and methods; and executing the analysis cycle.

[0046] Preferably, in step 3, the specific method of applying the BNN proxy model to perform probability analysis on the finite element model is:

[0047] (1) Initialize model parameters;

[0048] (2) Define the data preprocessing function PreprocessData; load and standardize the data, split it into training set and test set, and convert it into tensor format;

[0049] (3) Define the neural network model NeuralNetwork and implement the attention mechanism and multi-layer Dropout;

[0050] (4) Define the uncertainty estimation function PredictUncertainty and perform multiple forward propagations to estimate the uncertainty of the prediction;

[0051] (5) Start the training process, initialize the optimizer, perform forward propagation for each epoch, calculate the loss, perform backpropagation, update the weights, and output periodic loss information;

[0052] (6) Call the PredictUncertainty function to evaluate the prediction uncertainty of the test set;

[0053] (7) Use actual data for model prediction;

[0054] (8) Plot the prediction results and their uncertainties.

[0055] Preferably, in step 4, the constraint boundary line of the minimum distance between the new railway and the existing railway is obtained by calculating the probability of non-compliance, and the railway line is optimized in combination with the reliability evaluation result.

[0056] The present invention focuses on the reliability analysis and refined line selection technology of railway lines. By constructing a system reliability model, a finite element model that reproduces the existing line, and a BNN proxy model, the present invention deeply explores the reliability problem of railway lines under various conditions, and proposes a set of efficient and refined line selection methods. The system reliability model can accurately predict the reliability performance of railway lines in various complex environments, providing an important basis for refined line selection. The finite element model not only improves the analysis accuracy, but also provides solid data support for subsequent optimization work. The BNN proxy model realizes efficient prediction of unknown points, reduces calculation complexity, and improves optimization efficiency. This discovery is of great significance for dealing with large-scale, high-dimensional railway line optimization problems. In terms of railway line optimization based on the system reliability model, the present invention takes system reliability as one of the important constraints, calculates the constraint boundary line of the minimum distance between the new railway and the existing railway by specifying the system's non-compliance probability, and forms a continuously adjusted and improved linear optimization process by combining the reliability evaluation results and the railway line optimization algorithm. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] Figure 1 The present invention is a flow chart of a line optimization method for a new railway parallel to an existing line based on a reliability model in an embodiment. DETAILED DESCRIPTION

[0058] In order to further understand the content of the present invention, the present invention is described in detail in conjunction with the accompanying drawings and embodiments. It should be understood that the embodiments are only for explaining the present invention and are not intended to limit it.

[0059] Example

[0060] like Figure 1As shown, this embodiment provides a line optimization method for a new railway parallel to an existing line based on a reliability model, which includes the following steps:

[0061] 1. Establish a system reliability model;

[0062] Determine and evaluate the settlement and horizontal displacement of the existing railway bridge and roadbed nearby; build a system reliability model based on engineering specifications and actual conditions;

[0063] Specifically:

[0064] For the two parameters of settlement and horizontal displacement, the limit state functions are defined as:

[0065] o 1 (x) = R 1 -S 1

[0066] o 2 (x) = R 2 -S 2

[0067] Where:

[0068] o 1 (x), o 2 (x)—limit state functions of settlement and horizontal displacement, respectively;

[0069] R 1 , R 2 — respectively, the safe allowable values ​​of settlement and horizontal displacement;

[0070] S 1 , S 2 —are the actual settlement and horizontal displacement under specific conditions, respectively.

[0071] The series system reliability model is used because the existing adjacent railway is considered non-compliant as long as any one of the settlement or horizontal displacement exceeds the standard. Therefore, the system non-compliance probability P nc It can be expressed as:

[0072] P nc =P(o 1 (x)≤0∪o 2 (x)≤0)

[0073] P nc =P nc1 +P nc2 -P nc1nc2

[0074] Where:

[0075] P nc— is the probability of non-compliance of the system, indicating that when any limit state function (settlement or horizontal displacement) does not meet the conditions, the system is considered to be non-compliant. This model allows the assessment of the overall safety of the system under given conditions.

[0076] P nc1 — is the probability of non-compliance of the settlement pattern;

[0077] P nc2 — is the non-compliance probability of horizontal displacement;

[0078] P nc1nc2 — is the joint non-compliance probability of settlement and horizontal displacement occurring simultaneously.

[0079] 2. Finite element analysis;

[0080] Use engineering software (Abaqus) to build finite element models of existing railway bridges and roadbeds; simulate the construction process of new railways through the model and analyze its impact on adjacent structures; build a two-dimensional cross-section model to determine the settlement load effect caused by the construction of new railways; calculate the maximum allowable settlement R of its structure 1 and horizontal displacement R 2 ;

[0081] In the 2D cross-section model, the following settings are made:

[0082] (1) The soil is an elastic material and the Mohr-Coulomb model is used;

[0083] (2) The initial stress field of the roadbed and foundation is generated by the deadweight load;

[0084] (3) The new roadbed is reinforced with piles, the pile-soil contact condition is partial sliding, and the contact interface is simulated by the slip coefficient;

[0085] (4) The new roadbed is filled in layers, and each layer has good continuity;

[0086] (5) The left and right boundaries of the foundation are impermeable, and the bottom and top are permeable boundaries; the bottom is completely fixed and constrained, and sliding constraints are imposed on the vertical boundaries on both sides.

[0087] 3. Reliability and probability analysis;

[0088] Set the allowable settlement limits of bridges and roadbeds according to engineering specifications; use the reliability analysis module of Abaqus software, combined with the Monte Carlo simulation method, to identify the key factors that cause the settlement of adjacent bridges and roadbeds, and conduct reliability analysis; use the BNN proxy model to perform probabilistic analysis on the finite element model; based on the analysis results, evaluate the reliability of the adjacent existing railway bridges and roadbeds;

[0089] In the finite element model, the soil parameters are subjected to Monte Carlo simulation to construct a high-fidelity finite element model.

[0090] The Monte Carlo simulation method is:

[0091] First, multiple simulation scenarios were created, each containing a different combination of existing railway lines and construction conditions. Next, a large number of random simulations were performed on each scenario in the Abaqus finite element analysis software to generate settlement and horizontal displacement data for the existing railway lines under different conditions. These data were then analyzed to determine the frequency with which the existing railway lines remained within the safe allowable range in each scenario, i.e., their reliability.

[0092] The reliability analysis method combining Monte Carlo simulation method with finite element analysis is as follows:

[0093] 1) Initialize reliability analysis variables;

[0094] 2) Establish a solid model and divide the unit, including setting the unit type, real constant, and material. Parametric modeling must be performed during the process;

[0095] 3) Define the analysis type and corresponding options, apply loads, determine load step options, and call the function for initial solution;

[0096] 4) Simulate the construction process and solve the problem step by step;

[0097] 5) After the solution is completed, the corresponding calculation results are extracted in post-processing and the values ​​are assigned to the parameters designated as input variables and output result variables.

[0098] 6) Conduct reliability analysis, including: selecting and defining input variables and the correlation coefficients between input variables; determining the distribution type, distribution function and its parameters of each input variable; specifying the output result variable; selecting analysis tools and methods; and executing the analysis cycle.

[0099] The specific method of using the BNN agent model to perform probabilistic analysis on the finite element model is:

[0100] (1) Initialize model parameters;

[0101] (2) Define the data preprocessing function PreprocessData; load and standardize the data, split it into training set and test set, and convert it into tensor format;

[0102] (3) Define the neural network model NeuralNetwork and implement the attention mechanism and multi-layer Dropout;

[0103] (4) Define the uncertainty estimation function PredictUncertainty and perform multiple forward propagations to estimate the uncertainty of the prediction;

[0104] (5) Start the training process, initialize the optimizer, perform forward propagation for each epoch, calculate the loss, perform backpropagation, update the weights, and output periodic loss information;

[0105] (6) Call the PredictUncertainty function to evaluate the prediction uncertainty of the test set;

[0106] (7) Use actual data for model prediction;

[0107] (8) Plot the prediction results and their uncertainties.

[0108] 4. Optimize railway lines based on reliability analysis results.

[0109] In step 4, by calculating the probability of non-compliance, the constraint boundary line of the minimum distance between the new railway and the existing railway is obtained, and the railway line is optimized based on the reliability evaluation results.

[0110] This embodiment is carried out around the reliability analysis and refined line selection technology of railway lines. By constructing a system reliability model, a finite element model of existing lines, and a BNN agent model, the reliability problem of railway lines under various conditions is deeply discussed, and a set of efficient refined line selection methods are proposed; the system reliability model can accurately predict the reliability performance of railway lines in various complex environments, providing an important basis for refined line selection; the finite element model not only improves the analysis accuracy, but also provides solid data support for subsequent optimization work; the BNN agent model realizes efficient prediction of unknown points, reduces calculation complexity, and improves optimization efficiency. This discovery is of great significance for dealing with large-scale, high-dimensional railway line optimization problems; in terms of railway line optimization based on the system reliability model, this embodiment takes system reliability as one of the important constraints, calculates the constraint boundary line of the minimum distance between the new railway and the existing railway by specifying the non-compliance probability of the system, and combines the reliability evaluation results and the railway line optimization algorithm to form a continuously adjusted and improved line optimization process.

[0111] The present invention and its embodiments are described schematically above, and the description is not restrictive. The drawings show only one embodiment of the present invention, and the actual structure is not limited thereto. Therefore, if a person skilled in the art is inspired by it and designs a structural method and an embodiment similar to the technical solution without creativity without departing from the purpose of the invention, they shall all fall within the protection scope of the present invention.

Claims

1. A line optimization method for a new railway parallel to an existing line based on a reliability model, characterized by: The following steps are involved:

1. Establish a system reliability model; Determine and evaluate the settlement and horizontal displacement of the existing railway bridge and roadbed nearby; build a system reliability model based on engineering specifications and actual conditions; In step one, specifically: For the two parameters of settlement and horizontal displacement, the limit state functions are defined as: o1(x)=R1-S1 o2(x)=R2-S2 Where: o1(x), o2(x)—limit state functions of settlement and horizontal displacement, respectively; R1, R2—the safe allowable values ​​of settlement and horizontal displacement respectively; S1, S2—actual settlement and horizontal displacement under specific conditions, respectively; Using the tandem system reliability model, the existing adjacent railway is considered non-compliant as long as any one of the settlement or horizontal displacement exceeds the standard; therefore, the system non-compliance probability P nc It is expressed as: P nc =P(o1(x)≤0∪o2(x)≤0) P nc =P nc1 +P nc2 -P nc1nc2 Where: P nc — is the probability of non-compliance of the system, indicating that when any limit state function does not meet the conditions, the system is considered to be non-compliant; this model allows the assessment of the overall safety of the system under given conditions; P nc1 — is the probability of non-compliance of the settlement pattern; P nc2 — is the non-compliance probability of horizontal displacement; P nc1nc2 —is the joint non-compliance probability of settlement and horizontal displacement occurring simultaneously; 2. Finite element analysis; Use the engineering software Abaqus to build a finite element model of the existing railway bridge and roadbed; simulate the construction process of the new railway through the model and analyze its impact on the adjacent structures; build a two-dimensional cross-section model to determine the settlement load effect caused by the construction of the new railway; calculate the maximum safe allowable value R1 of the settlement and the safe allowable value R2 of the horizontal displacement of the structure; 3. Reliability and probability analysis; Set the allowable settlement limits of bridges and roadbeds according to engineering specifications; use the reliability analysis module of Abaqus software, combined with the Monte Carlo simulation method, to identify the key factors that cause the settlement of adjacent bridges and roadbeds, and conduct reliability analysis; use the BNN proxy model to perform probabilistic analysis on the finite element model; based on the analysis results, evaluate the reliability of the adjacent existing railway bridges and roadbeds; 4. Optimize railway lines based on reliability analysis results; In step 4, by calculating the probability of non-compliance, the constraint boundary line of the minimum distance between the new railway and the existing railway is obtained, and the railway line is optimized based on the reliability evaluation results.

2. The line optimization method for a new railway parallel to an existing line based on a reliability model according to claim 1 is characterized by: In the two-dimensional cross-section model of step 2, the following settings are made: (1) The soil is an elastic material and the Mohr-Coulomb model is used; (2) The initial stress field of the roadbed and foundation is generated by the deadweight load; (3) The new roadbed is reinforced with piles, the pile-soil contact condition is partial sliding, and the contact interface is simulated by the slip coefficient; (4) The new roadbed is filled in layers, and each layer has good continuity; (5) The left and right boundaries of the foundation are impermeable, and the bottom and top are permeable boundaries; the bottom is completely fixed and constrained, and sliding constraints are imposed on the vertical boundaries on both sides.

3. The line optimization method for a new railway parallel to an existing line based on a reliability model according to claim 2 is characterized by: In step three, the soil parameters are subjected to Monte Carlo simulation in the finite element model to construct a high-fidelity finite element model.

4. The line optimization method for a new railway parallel to an existing line based on a reliability model according to claim 3 is characterized by: In step 3, the Monte Carlo simulation method is: First, multiple simulation scenarios were created, each containing a different combination of existing railway lines and construction conditions. Next, a large number of random simulations were performed on each scenario in the Abaqus finite element analysis software to generate settlement and horizontal displacement data for the existing railway lines under different conditions. These data were then analyzed to determine the frequency with which the existing railway lines remained within the safe allowable range in each scenario, i.e., their reliability.

5. The line optimization method for a new railway parallel to an existing line based on a reliability model according to claim 4 is characterized by: In step 3, the reliability analysis method of the Monte Carlo simulation method combined with the finite element analysis is as follows: 1) Initialize reliability analysis variables; 2) Establish a solid model and divide the unit, including setting the unit type, real constant, and material. Parametric modeling must be performed during the process; 3) Define the analysis type and corresponding options, apply loads, determine load step options, and call the function for initial solution; 4) Simulate the construction process and solve the problem step by step; 5) After the solution is completed, the corresponding calculation results are extracted in post-processing, and the values ​​are assigned to the parameters designated as input variables and output result variables; 6) Conduct reliability analysis, including: selecting and defining input variables and the correlation coefficients between input variables; determining the distribution type, distribution function and its parameters of each input variable; specifying the output result variable; selecting analysis tools and methods; and executing the analysis cycle.

6. The line optimization method for a new railway parallel to an existing line based on a reliability model according to claim 5 is characterized by: In step 3, the specific method of applying the BNN proxy model to perform probability analysis on the finite element model is: (1) Initialize model parameters; (2) Define the data preprocessing function PreprocessData; load and standardize the data, split it into training set and test set, and convert it into tensor format; (3) Define the neural network model NeuralNetwork and implement the attention mechanism and multi-layer Dropout; (4) Define the uncertainty estimation function PredictUncertainty and perform multiple forward propagations to estimate the uncertainty of the prediction; (5) Start the training process, initialize the optimizer, perform forward propagation for each epoch, calculate the loss, perform backpropagation, update the weights, and output periodic loss information; (6) Call the PredictUncertainty function to evaluate the prediction uncertainty of the test set; (7) Use actual data for model prediction; (8) Plot the prediction results and their uncertainties.

Images