Parameter optimization and reliability evaluation method for active suspension control system of high-speed train

Abstract

The invention discloses a parameter optimization and reliability evaluation method for an active suspension control system of a high-speed train, and the method comprises the steps: combining a kinetic model with a multi-objective optimization algorithm; an active suspension secondary vertical stiffness parameter, an active suspension secondary transverse stiffness parameter, an active suspension primary radial stiffness parameter and an anti-snaking shock absorber stiffness parameter of the high-speed train are respectively used as design variables, and curve passing performance, a sperling index and a critical speed in the operation of the high-speed train are optimized; and when the optimization reaches the maximum iteration parameter and the obtained pareto optimal parameter combination meets the constraint condition, verifying the safety, stability and stability of the high-speed train until reaching the standard, thereby verifying the reliability of the active suspension control system, and finally evaluating the reliability of the active suspension control system of the high-speed train. Therefore, the running performance and the line adaptability of the train are improved.

Description

Technical Field

[0001] This invention belongs to the field of evaluation technology for active suspension systems of trains, specifically relating to a method for optimizing parameters and evaluating the reliability of active suspension control systems for high-speed trains. Background Technology

[0002] By the end of 2022, my country's railway operating mileage reached 155,000 kilometers, of which high-speed rail accounted for 42,000 kilometers, forming the world's largest high-speed rail network. With the long-term, large-scale operation of high-speed trains, many typical dynamic problems have emerged. For example, there is the issue of train stability during cross-line operation. Vehicles on high-speed, intercity, and existing railway lines sometimes cross lines, with high-speed trains operating at reduced speeds on existing lines. Vehicle-track mismatch leads to vehicle swaying failure; wheel-rail mismatch results in a significant dominant frequency of lateral vibration acceleration around 1Hz, and the vehicle's lateral stability exceeds the standard. Different railway lines have different characteristics, and the relevant dynamic performance of railway vehicles varies when operating on different lines. Secondly, my country's high-speed trains face very complex geographical and climatic environments during operation. The dynamic problems exposed by existing high-speed trains reveal that the unchanging suspension parameters and structure of traditional railway vehicles result in a lack of proactive adaptation to track conditions, while new bogies with proactively variable structures and adaptive adjustment capabilities to track changes can effectively solve the aforementioned cross-line operation challenges.

[0003] Active suspension technology is primarily used to further improve train performance and track adaptability: enhancing ride comfort, improving train homing stability, improving curve clearance performance, and reducing wheel-rail wear. Multi-objective optimization and collaborative control logic and algorithm design, by controlling the parameter values ​​of active suspension components, can improve the dynamic performance of the running system. Currently, evaluating the reliability of high-speed train active suspension control systems has significant strategic importance and engineering practical value. However, there is no reliable evaluation method for high-speed train active suspension control systems; therefore, this paper proposes a parameter optimization and reliability evaluation method for high-speed train active suspension control systems. Summary of the Invention

[0004] The purpose of this invention is to provide a method for optimizing parameters and assessing the reliability of an active suspension control system for high-speed trains. This method combines a dynamic model with a multi-objective optimization algorithm, using the vertical stiffness parameters of the secondary active suspension system, the lateral stiffness parameters of the secondary active suspension system, the radial stiffness parameters of the primary active suspension system, and the stiffness parameters of the anti-hunting damper as design variables. It optimizes the high-speed train's curve-passing performance, sperling index, and critical speed during operation. When the optimization reaches the maximum iterative parameters and the obtained Pareto optimal parameter combination satisfies the constraints, it verifies whether the safety, smoothness, and stability of the high-speed train meet the standards. If not, it changes the design variable parameters until the safety, smoothness, and stability of the high-speed train meet the standards, thereby verifying the reliability of its active suspension control system. This method improves the train's operating performance and track adaptability through the reliability assessment of the active suspension control system of high-speed trains.

[0005] This invention is achieved through the following technical solution: The method for parameter optimization and reliability assessment of the active suspension control system of high-speed trains includes the following steps: S1: Establish a vehicle-track coupled dynamics model for active suspension control; S2: Select design variables and determine optimization objectives based on the vehicle-track coupled dynamics model of active suspension control; S3: Perform multi-objective optimization on the active suspension control vehicle-track coupled dynamics model to obtain the final optimized solution; S4: Substitute the final optimized solution into the active suspension control vehicle-track coupled dynamics model to determine the train's operating status.

[0006] Preferably, in step S2, the active suspension secondary vertical stiffness parameter, the active suspension secondary lateral stiffness parameter, the active suspension primary radial stiffness parameter, and the anti-hunting damper stiffness parameter are selected from the existing high-speed train suspension parameter library as design variables.

[0007] Preferably, in step S2, the optimization objectives include one or more of the following: curve passing performance, sperling index, critical speed, and vehicle vibration acceleration.

[0008] Preferably, the curve passing performance includes wheel-rail vertical force, wheel-axle lateral force, derailment coefficient, and wheel load reduction rate.

[0009] Preferably, in step S3, a multi-objective optimization algorithm is used for multi-objective optimization. The multi-objective optimization algorithm is one of the following: second-generation fast non-dominated sorting genetic algorithm, intensity Pareto evolution algorithm, or optimization algorithm for high-dimensional objectives.

[0010] Preferably, in step S3, a second-generation fast non-dominated sorting genetic algorithm is used for multi-objective optimization. The second-generation fast non-dominated sorting genetic algorithm is designed based on genetic algorithms and non-dominated sorting genetic algorithms.

[0011] Preferably, in step S3, the algorithm parameters of the multi-objective optimization algorithm are set, the chromosomes of the first generation population in the multi-objective optimization algorithm are encoded, and fast non-dominated sorting and virtual crowding are calculated and genetic operations are performed until the last generation population is obtained, and the chromosomes of the last generation population are decoded.

[0012] Preferably, the optimal Pareto parameter combination is obtained by decoding the chromosomes of the last generation population, and the constraint conditions of the optimal Pareto parameter combination are determined.

[0013] Preferably, the constraints include constraints on the design variables and the objective function; the optimal Pareto parameter combination that satisfies the constraints is used as the final optimization scheme and substituted into the active suspension control vehicle track coupling dynamics model to determine the train operation status.

[0014] Preferably, if the optimal Pareto parameter combination does not meet the constraints, the algorithm returns to the parameter setting step.

[0015] Compared with the prior art, the present invention has the following advantages and beneficial effects: 1) In this invention, by combining the vehicle-track coupled dynamics model of active suspension control with a multi-objective optimization algorithm, the vertical stiffness parameters of the active suspension secondary system, the lateral stiffness parameters of the active suspension secondary system, the radial stiffness parameters of the active suspension primary system, and the stiffness parameters of the anti-hunting damper of the high-speed train are used as design variables to optimize the curve passing performance, sperling index, and critical speed of the high-speed train. When the optimization reaches the maximum iteration parameters and the obtained Pareto optimal parameter combination satisfies the constraints, the safety, smoothness, and stability of the high-speed train are verified until they meet the standards, thereby verifying the reliability of its active suspension control system. Finally, through the reliability evaluation of the active suspension control system of the high-speed train, the train's operating performance and track adaptability are improved.

[0016] 2) In this invention, three key performance objectives, namely curve passage, Sperling index and critical speed, are simultaneously used as optimization objectives. A multi-objective evolutionary algorithm is used to find the Pareto optimal solution set, so that the trade-off relationship between different objectives can be clearly seen. The parameter combination that best meets the specific line or operation requirements can be selected according to different situations, which fundamentally avoids the overall performance imbalance caused by single objective optimization.

[0017] 3) In this invention, the reliability assessment process is embedded into a target performance-oriented active optimization loop algorithm. The optimization algorithm actively explores the parameter space to find Pareto parameter combinations that can simultaneously improve the performance of multiple targets. Reliability verification is then performed based on these solution sets. This greatly improves the efficiency and probability of finding reliable parameter combinations, and the evaluation results are based on optimized performance, making them more convincing.

[0018] 4) In this invention, a clear closed-loop verification process is constructed. If the reliability assessment performed on the optimal solution set obtained by optimization still fails to meet the standard, the process will automatically feed back to the starting point, reselect the design variables, and then re-optimize and verify. This closed-loop mechanism forcibly ensures that the final output evaluation result is based on meeting all core performance standards, which significantly improves the reliability and practicality of the evaluation conclusion.

[0019] 5) In this invention, the second-generation fast non-dominated sorting genetic algorithm can efficiently explore the parameter space and find a series of robust Pareto optimal solutions. Based on these solutions, reliability assessment is carried out, which means that the finally selected reliable parameter combination naturally has the ability to cope with multi-objective conflicts and a certain degree of parameter change, thus improving the robustness of the entire active suspension control system. At the same time, this method greatly reduces the trial and error process that relies on experience and accelerates the design iteration and reliability verification cycle. Attached Figure Description

[0020] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings in the embodiments will be briefly described below. It should be understood that the following drawings only show some embodiments of the present invention and should not be regarded as a limitation on the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a flowchart of the parameter optimization and reliability evaluation method for the active suspension control system of high-speed trains in this invention.

[0022] Figure 2 This is a diagram of the vehicle-track coupling dynamics model for active suspension control constructed in this invention.

[0023] Figure 3 This is a model diagram of the high-speed train bogie constructed according to the present invention.

[0024] Figure 4 This is a diagram of the multi-objective optimization model constructed in this invention. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are some embodiments of the present invention, but not all embodiments.

[0026] Example 1: Methods for parameter optimization and reliability assessment of active suspension control systems for high-speed trains, such as... Figure 1 , Figure 2 , Figure 3 and Figure 4 As shown, it includes the following steps: S1: Establish a vehicle-track coupled dynamics model for active suspension control; S2: Select design variables and determine optimization objectives based on the vehicle-track coupled dynamics model of active suspension control; S3: Perform multi-objective optimization on the active suspension control vehicle-track coupled dynamics model to obtain the final optimized solution; S4: Substitute the final optimized solution into the active suspension control vehicle-track coupled dynamics model to determine the train's operating status.

[0027] like Figure 1 As shown, an active suspension vehicle-track coupling dynamic model is established using Simpack, which includes four sub-models: active suspension train system dynamic model, track system dynamic model, wheel-rail space dynamic contact model, and train / track coupling interface excitation model.

[0028] like Figure 2 and Figure 3 As shown, a high-speed train dynamics model consisting of three cars was established based on multibody dynamics theory. The train dynamics model is a rigid-flexible coupled vehicle model, comprising seven components: one car body, two frames, and four wheelsets. Figure 3 The modeling methods for the frame and wheelsets in the SIMPACK model are presented in detail: both adopt a 6-DOF rigid body model, possessing full motion capabilities including longitudinal, lateral, and vertical translation, as well as roll, pitching, and yaw rotation. The car body is modeled as a flexible body, combining its 6th rigid body mode and 20th elastic mode characteristics, and an active suspension system is introduced to significantly improve the operating performance and track adaptability of the active suspension train.

[0029] The dynamic model of the track system adopts a slab track structure, including rails, fasteners, track slabs, CA mortar layer, and subgrade. The left and right rails are modeled as Timoshenko beams on a continuous elastic discrete-point support foundation, with lateral, vertical, and torsional vibration effects considered. The track slab is treated as a rectangular plate of uniform thickness on an elastic foundation. The rail fastener system and CA mortar layer are simplified as periodic discrete viscoelastic elements. The vibration response of the subgrade is not considered in the model.

[0030] By coupling the train system dynamics model and the track system dynamics model together through the wheel-rail contact relationship, a spatially coupled dynamics model of vehicle-track interaction under active suspension control is formed, enabling flexible vibration analysis of the vehicle body under the excitation of track geometric irregularities. The wheel-rail spatial contact geometry is determined using the trace method. The wheel-rail normal force is calculated based on Hertz nonlinear elastic contact theory, while the tangential creep force is first estimated using Kalker linear theory and then solved precisely using the FASTSIM algorithm. The system dynamics equations are integrated using a novel fast numerical method based on Zhai's approach.

[0031] When obtaining design variables, parameters are selected from the existing high-speed train suspension parameter library. These design variables include the car body's vertical acceleration, lateral acceleration, roll velocity and angle, pitch velocity and angle, yaw velocity and angle, secondary suspension vertical stiffness, secondary suspension lateral stiffness, primary suspension radial stiffness, anti-hunting damper stiffness, axle box vertical acceleration, and axle box lateral acceleration. The secondary suspension connects the car body and bogie, directly affecting ride comfort and stability; vertical stiffness relates to the car body's vertical vibration, while lateral stiffness affects lateral sway; the primary suspension connects the bogie and wheelsets, controlling the force and guidance performance between the wheel and rail; and the anti-hunting damper is crucial for suppressing hunting motion and preventing instability at high speeds. Therefore, this invention selects the secondary suspension vertical stiffness parameter, secondary suspension lateral stiffness parameter, primary suspension radial stiffness parameter, and anti-hunting damper stiffness parameter as design variables.

[0032] The optimization objectives include curve passing performance, sperling index, and critical speed. Curve passing performance includes wheel-rail vertical force, wheel-axle lateral force, derailment coefficient, and wheel load reduction rate. Curve passing performance corresponds to the safety of the active suspension control system, sperling index corresponds to the smoothness of the active suspension control system, and critical speed corresponds to the stability of the active suspension control system. The reliability of the train's active suspension control system is evaluated through multiple dimensions of safety, smoothness, and stability.

[0033] The derailment factor is the ratio (Q / P) of the lateral force Q and the vertical force P acting on the wheel at a certain moment. The limit value of the derailment factor is: The wheel load reduction rate is the ratio of the load reduction of the wheel on the unloaded side to the average static wheel weight of the wheelset. The safety standard for the wheel load reduction rate is: The maximum vertical force on the wheel and rail is 170 kN, and the formula for calculating the allowable limit of the lateral force on the wheel and axle is: In the formula: Pw is the static axle load.

[0034] The formula for calculating the stationarity index is: In the formula: A —Vibration acceleration, measured in meters per second squared (m / s²) 2 ); f —Vibration frequency, measured in Hertz (Hz); F ( f — Frequency correction factor, see Table 1.

[0035] Table 1 Frequency Correction Coefficient Table Each indicator can also be analyzed and calculated using other methods known to those skilled in the art.

[0036] By combining the vehicle-track coupled dynamics model of active suspension control with a multi-objective optimization algorithm, and using the vertical stiffness parameters of the active suspension secondary system, the lateral stiffness parameters of the active suspension secondary system, the radial stiffness parameters of the active suspension primary system, and the stiffness parameters of the anti-hunting damper as design variables, the curve passing performance, sperling index, and critical speed of the high-speed train are optimized. After obtaining the final optimized scheme, the safety, smoothness, and stability of the high-speed train are verified until they meet the standards, thereby verifying the reliability of its active suspension control system. Finally, through the reliability evaluation of the active suspension control system of the high-speed train, the train's operating performance and track adaptability are improved.

[0037] Example 2: This embodiment further defines a multi-objective optimization method based on the above embodiments, such as... Figure 1 and Figure 4 As shown, to achieve data exchange between Isight and Simpack, the Simcode component in Isight is used to write the interface between the two. This component mainly consists of three parts: 1. Input module: Parses the Suspension.subvar substitution variable file to obtain dangling parameters.

[0038] 2. Execution Module: The Runsimpack.bat or Postsimpack.bat batch command is used to drive the Simpack software, thereby completing the calculation and analysis of sample data or the format conversion of output result files.

[0039] 3. Output module: This module parses and processes the response data in the output file.

[0040] The optimization algorithm is set through the Optimization1 module. Specifically, the second-generation fast non-dominated sorting genetic algorithm is configured by setting the population size, generation number, crossover, and mutation index. The Simcode1 and Simcode3 components are responsible for the preprocessing parameterization of the dynamic models and automated calculations for linear and curved operating conditions, respectively. The Simcode2 and Simcode4 components are responsible for the transformation of calculation results and defining key index values ​​in the calculation results as responses, respectively. The Matlab component is responsible for data processing of the operational stability response for the linear operating condition.

[0041] The multi-objective optimization method employs a second-generation fast non-dominated sorting genetic algorithm. This algorithm is designed based on genetic algorithms and non-dominated sorting genetic algorithms. First, the algorithm parameters are set, including the number of generations, population size, and crossover / mutation parameters. Then, chromosome encoding is performed using the second-generation fast non-dominated sorting genetic algorithm to obtain the initial population. The external performance level of each individual reflects whether the optimization objective is better. By combining the external performance level of each individual with fast non-dominated sorting and virtual crowding calculations, a subpopulation Qg with better performance is obtained. According to the elite retention strategy, the M original populations Pg and the subpopulation Qg are mixed to form a new population Rg with a total size of 2M. This newly generated population is then subjected to a second fusion of fast non-dominated sorting and virtual crowding calculations to form the next generation population Pg+1. The above operations are repeated, and after multiple evolutions, a population with performance approaching optimality is finally obtained, i.e., the final generation population. Chromosome decoding is performed on the individual results in the final generation population to obtain the optimal coefficients of the design variable parameters. Next, the Pareto optimal parameter combination is obtained, and the constraints of the Pareto optimal parameter combination are judged. The constraints include constraints on the design variables and the objective function. If the Pareto optimal parameter combination meets the constraints, it is used as the final optimization scheme and substituted into the active suspension control vehicle track couple and dynamics model to judge the train operation. If the Pareto optimal parameter combination does not meet the constraints, chromosome encoding is repeated. Finally, the train operation is judged. If the vehicle runs smoothly, it means that the safety, smoothness and stability of the evaluated system meet the standards. If the vehicle does not run smoothly, the design variables are selected again, and then optimization and evaluation are carried out again.

[0042] In this embodiment, by constructing a clear closed-loop verification process, if the reliability assessment performed on the optimal solution set obtained through optimization still fails to meet the standards, the process will automatically feed back to the starting point, reselect design variables, and then re-optimize and verify. This closed-loop mechanism forcibly ensures that the final output evaluation result is based on meeting all core performance standards, which can significantly improve the reliability and practicality of the evaluation conclusion.

[0043] Furthermore, the second-generation fast non-dominated sorting genetic algorithm efficiently explores the parameter space, finding a series of robust Pareto optimal solutions. Reliability assessments are then performed based on these solutions, and the final selected reliable parameter combination naturally possesses the ability to cope with multi-objective conflicts and a certain degree of parameter variation, improving the robustness of the entire active suspension control system and significantly reducing the trial-and-error process dependent on experience. Ultimately, the three key performance objectives—curve passage, Sperling index, and critical speed—are simultaneously used as optimization objectives, and a multi-objective evolutionary algorithm is employed to find the Pareto optimal solution set. This allows the trade-offs between different objectives to be clearly seen, and the parameter combination that best meets the specific route or operational requirements can be selected according to different scenarios, fundamentally avoiding overall performance imbalances caused by single-objective optimization. The other parts of this embodiment are the same as those in the above embodiments and will not be repeated here.

[0044] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications or equivalent changes made to the above embodiments based on the technical essence of the present invention shall fall within the protection scope of the present invention.

Claims

1. A method for parameter optimization and reliability assessment of a high-speed train active suspension control system, characterized in that, Includes the following steps: S1: Establish a vehicle-track coupled dynamics model for active suspension control; S2: Select design variables and determine optimization objectives based on the vehicle-track coupled dynamics model of active suspension control; S3: Perform multi-objective optimization on the active suspension control vehicle-track coupled dynamics model to obtain the final optimized solution; S4: Substitute the final optimized solution into the active suspension control vehicle-track coupled dynamics model to determine the train's operating status.

2. The method for parameter optimization and reliability evaluation of the high-speed train active suspension control system as described in claim 1, characterized in that, In step S2, the active suspension secondary vertical stiffness parameter, the active suspension secondary lateral stiffness parameter, the active suspension primary radial stiffness parameter, and the anti-hunting damper stiffness parameter are selected from the existing high-speed train suspension parameter library as design variables.

3. The method for parameter optimization and reliability evaluation of the high-speed train active suspension control system as described in claim 1, characterized in that, In step S2, the optimization objectives include one or more of the following: curve passing performance, sperling index, critical speed, and vehicle vibration acceleration.

4. The method for parameter optimization and reliability evaluation of the high-speed train active suspension control system as described in claim 3, characterized in that, The curve passes through performance parameters including wheel-rail vertical force, wheel-axle lateral force, derailment coefficient, and wheel load reduction rate.

5. The method for parameter optimization and reliability evaluation of the active suspension control system for high-speed trains as described in claim 1, characterized in that, In step S3, a multi-objective optimization algorithm is used to perform multi-objective optimization. The multi-objective optimization algorithm is one of the following: second-generation fast non-dominated sorting genetic algorithm, intensity Pareto evolution algorithm, or optimization algorithm for high-dimensional objectives.

6. The method for parameter optimization and reliability evaluation of the high-speed train active suspension control system as described in claim 5, characterized in that, In step S3, the second-generation fast non-dominated sorting genetic algorithm is used for multi-objective optimization. The second-generation fast non-dominated sorting genetic algorithm is designed based on genetic algorithms and non-dominated sorting genetic algorithms.

7. The method for parameter optimization and reliability evaluation of the high-speed train active suspension control system as described in claim 6, characterized in that, In step S3, the algorithm parameters of the multi-objective optimization algorithm are set, the chromosomes of the first generation population in the multi-objective optimization algorithm are encoded, and fast non-dominated sorting and virtual crowding are calculated and genetic operations are performed until the last generation population is obtained, and the chromosomes of the last generation population are decoded.

8. The method for parameter optimization and reliability evaluation of the active suspension control system for high-speed trains as described in claim 7, characterized in that, The optimal Pareto parameter combination is obtained by decoding the chromosomes of the last generation population, and the constraint conditions of the optimal Pareto parameter combination are determined.

9. The method for parameter optimization and reliability evaluation of the high-speed train active suspension control system as described in claim 8, characterized in that, The constraints include constraints on the design variables and the objective function; the optimal Pareto parameter combination that satisfies the constraints is used as the final optimization scheme and substituted into the active suspension control vehicle-track coupled dynamics model to determine the train operation status.

10. The method for parameter optimization and reliability evaluation of the active suspension control system for high-speed trains as described in claim 8, characterized in that, If the optimal Pareto parameter combination does not meet the constraints, the algorithm returns to the parameter setting step.

Images