Machine learning from computer simulations with applications in rail vehicle dynamics

Taheri, Mehdi; Ahmadian, Mehdi

doi:10.1080/00423114.2016.1150497

Cited by 18 publications

(10 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since there could be redundant information among the training data, a data sampling process is recommended. For the reason that a sampling plan has a large effect on the accuracy and efficiency of a data-driven model, the sampling plan developed by Taheri et al was introduced to pick the training samples in this paper [31]. The idea of this sampling plan is to select the optimal samples = � (1) , (2) , … , ( ) � T , ( * ) ∈ ⊂ D from the training data set , such that can mostly spatially fill the data set , where D is the dimension of ( * ) .…”

Section: Training Data Samplingmentioning

confidence: 99%

“…To verify the accessibility of the data-driven dynamics simulation framework, in this section two case studies were carried out. In the first one, the usage of this framework in a common vertical dynamics simulation was elaborated, and a comparison of two data-driven models ( the LPR model and the Kriging model [31]) was made to verify the performance of these models; In the second one, we used the LPR model to build the surrogate element and gave a detailed description in the longitudinal dynamics simulation in a train crash. All the simulations in this section were carried out on the platform of Window 7 64bit with two CPUs (Intel Xeon X5680…”

Section: Case Studiesmentioning

confidence: 99%

“…With these data, the step-wise function was used to fit a force-displacement curve during the deformation period of an end structure, which was applied in the crashworthiness design [29,30]. Taheri et al introduced the machine learning techniques to the vehicle dynamics, where MB simulation data were applied to train the Kriging model in building a surrogate element which was used to replace all the force elements in the second suspension system of a vehicle's MB model [31]. In the aspect of the structure crashworthiness design, Forsberg et al and Xie et al used the data from FE simulation to train the Kriging model and the polynomial response surfaces, which were used as the surrogate models to avoid iterations with using the FE method.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A data-driven dynamics simulation framework for railway vehicles

Nie

Zhao

Liu

et al. 2017

Vehicle System Dynamics

View full text Add to dashboard Cite

Sophisticated finite element (FE) model are usually critical for the research and simulation of vehicle dynamics, especially for train crash cases. However, factors such as the complexity of the meshes, the distortion problems involved in a large deformation, etc. would undermine the calculation efficiency of a FE model. Its alternative, a multi-body (MB) model, shows a satisfying time efficiency though, it only presents a limited simulation accuracy when involving highly nonlinear characteristics in a dynamic process. To maintain the advantages of both the two methods, this paper proposes a data-driven simulation framework to model the dynamic behaviours of railway vehicles. In this framework, by extracting the training data of FE simulation using machine learning techniques, the nonlinear characteristics of structures are formulated into a surrogate element to replace the original mechanical elements, then the dynamics simulation is accomplished by co-simulation via embedding the surrogate element into the MB model. This framework consists of a series of techniques including the data collection and feature extraction, the training data sampling, the surrogate element building, and the model evaluation and selection. To verify the accessibility of this framework, two case studies of a vertical and a longitudinal vehicle dynamics simulation are carried out based on Simulink/Simpack co-simulation. By comparing two data-driven models (the Legendre polynomial regression (LPR) and the Kriging), the result shows that using the LPR model in building surrogate elements can largely cut down the simulation time without much compromising of the accuracy.

show abstract

Section: Training Data Samplingmentioning

confidence: 99%

Section: Case Studiesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A data-driven dynamics simulation framework for railway vehicles

Nie

Zhao

Liu

et al. 2017

Vehicle System Dynamics

View full text Add to dashboard Cite

show abstract

“…e results of simulation and experimental demonstrated that the method can estimate the length of wheel-flats [12]. e stochastic models of single suspension and lateral suspension were trained, which can represent the behavior of the suspension system and investigate the behavior of a multibody dynamics model by machine learning from computer simulations [13].…”

Section: Introductionmentioning

confidence: 99%

Influence of the Dry Friction Suspension System Characteristics on the Stick‐Slip of Vertical Vibration of a Three‐Piece Bogie

Lyu

Zhang

Lyu³

et al. 2021

Shock and Vibration

View full text Add to dashboard Cite

The dry friction structure is a commonly used vibration-damping method for railway vehicles. Insufficient vibration damping performance will cause excessive vibration of the vehicle, which is not conducive to the safety of the vehicle. However, the mechanism of vibration damping and the cause of clamping stagnation have not been well resolved. This paper uses the analytical method, numerical method, and finite element method to analyze the vertical dynamic characteristics of the simple suspension system with dry friction and demonstrates that the numerical method is an effective method to study the dry model. The conditions for the system to produce sticking events were analyzed by the numerical method. The analysis shows that the system's excitation is too small, which causes clamping stagnation to the system. The reduction of the wedge angle and the friction coefficient are conducive to eliminating sticking. A negative side frame angle is conducive to reducing the high-frequency energy of the excitation. Decreasing spring stiffness or increasing system mass to reduce system frequency can reduce sticking events. The mutual verification of different methods confirms the correctness of the analysis method and analyzes the cause of sticking or clamping stagnation from the mechanism, which provides a new idea for the design and improvement of the dry friction damping system of railway vehicle bogies.

show abstract

“…Our aim, in this paper, is to use this model to identify the influence of friction coefficient on the appearance and development of stick-slip states. The exact quantification of the coefficient value is an experimental task, which can be assessed using simulation data, empirical results and model identification techniques such as the machine learning approach presented in [21]. Our intent, therefore, is not related to establish the true value for the friction coefficient, but rather to investigate what would happen to system response when different friction coefficients are applied.…”

Section: Introductionmentioning

confidence: 99%

On the application of linear complementarity-based contact to study the dynamic behavior of friction dampers of railway vehicles

Baruffaldi

Júnior

2018

J Braz. Soc. Mech. Sci. Eng.

View full text Add to dashboard Cite

Secondary suspension of freight cars has a small number of components, is reliable, cheap to maintain, and has a very traditional design. Despite all these characteristics, its dynamic response is subject to a number of interesting phenomena related to friction damping that are not completely explored for new design of heavy vehicles. This paper describes a new method for modeling this type of mechanical system using a linear complementarity approach for contact calculation. Results presented here show that friction coefficient not only can affect the general damping of the dynamic response to sinusoidal excitation, but also impacts the development of nonlinear phenomena related to stick-slip. These should be taken into account the development of the new generation of freight wagons aiming for increased performance and safety and reduced maintenance costs.

show abstract

Machine learning from computer simulations with applications in rail vehicle dynamics

Cited by 18 publications

References 10 publications

A data-driven dynamics simulation framework for railway vehicles

A data-driven dynamics simulation framework for railway vehicles

Influence of the Dry Friction Suspension System Characteristics on the Stick‐Slip of Vertical Vibration of a Three‐Piece Bogie

On the application of linear complementarity-based contact to study the dynamic behavior of friction dampers of railway vehicles

Contact Info

Product

Resources

About