| 1. |
- Kulkarni, Rohan, 1991-, et al.
(author)
-
iVRIDA: intelligent Vehicle Running Instability Detection Algorithm for high-speed rail vehicles using Temporal Convolution Network : – A pilot study
- 2022
-
In: Proceedings of the 7th European Conference of the Prognostics and Health Management Society 2022. - : PHM Society. ; , s. 269-277
-
Conference paper (peer-reviewed)abstract
- Intelligent fault identification of rail vehicles from onboard measurements is of utmost importance to reduce the operating and maintenance cost of high-speed vehicles. Early identification of vehicle faults responsible for an unsafe situation, such as the instable running of highspeed vehicles, is very important to ensure the safety of operating rail vehicles. However, this task is challenging because of the nonlinear dynamics associated with multiple subsystems of the rail vehicle. The task becomes more challenging with only accelerations recorded in the carbody where, nevertheless, sensor maintenance is significantly lower compared to axlebox accelerometers. This paper proposes a Temporal Convolution Network (TCN)-based intelligent fault detection algorithm to detect rail vehicle faults. In this investigation, the classifiers are trained and tested with the results of numerical simulations of a high-speed vehicle (200 km/h). The TCN based fault classification algorithm identifies the rail vehicle faults with 98.7% accuracy. The proposed method contributes towards digitalization of rail vehicle maintenance through condition-based and predictive maintenance.
|
|
| 2. |
- Kulkarni, Rohan, 1991-, et al.
(author)
-
Fault detection and isolation method for vehicle running instability from vehicle dynamics response using machine learning
- 2019
-
In: Proceedings of 11th International Conference on Railway Bogies and Running Gears (BOGIE'19). - Budapest.
-
Conference paper (peer-reviewed)abstract
- In this paper, a Fault Detection and Isolation (FDI) method is proposed for monitoring the vehicle running stability in a high-speed railway bogie. The objective is to detect and isolate the different faults of bogie components which are critical to vehicle stability, especially degraded yaw dampers and high equivalent conicity caused by wheel wear. The proposed method has two steps; firstly, signal features sensitive to the characteristics of running instability are extracted based on frequency domain and time domain analysis of lateral accelerations of bogie frame and axlebox; then these features along with vehicle speed are fed into machine learning based fault classifiers. The supervised machine learning based fault classifier are trained to identify the cause of observed running instability among yaw damper degradation and wheel-rail profile pair with high equivalent conicity. The Support Vector Machine (SVM) classifier with Linear and Gaussian kernels are trained by k-fold crossvalidation method and the hyperparameters are optimized with a bayesian optimization algorithm to minimize the classification error. These fault classifiers are trained and tested with an extensive database generated from numerical experiments performed by multibody simulation (MBS) software. The performance of Linear and Gaussian SVM fault classifiers is compared with each other to identify the best performing classifier. The results underline the ability of machine learning based fault classifiers to be used for FDI of vehicle running instability and outline the possibility of detecting and isolating bogie faults critical to the vehicle stability based on onboard measurement of vehicle dynamic response.
|
|
| 3. |
- Kulkarni, Rohan, 1991-, et al.
(author)
-
Identification of vehicle response features for onboard diagnosis of vehicle running instability
- 2022
-
In: 2022 IEEE International Conference on Prognostics and Health Management (ICPHM). - : Institute of Electrical and Electronics Engineers (IEEE). ; , s. 52-57
-
Conference paper (peer-reviewed)abstract
- Condition Monitoring (CM) of dynamic vehicle track interaction is an important research topic in rail vehicle dynamics. The most cost-effective method for CM is through carbody floor mounted accelerometers because this is most safe and reliable location for onboard accelerometers onboard inservice train. However, the dynamic response of carbody is influenced not only by excitations coming from track but also by various nonlinearities such as wheel-rail interface and vehicle suspension elements. Thus, it is very challenging to accurately monitor track subsystems via carbody floor accelerations. In this article, two feature extraction algorithms are proposed with the objective of obtaining crucial information on the stability of vehicle using carbody floor accelerations. The first algorithm is based on spectral analysis and the latter is on adaptive signal processing technique. The first algorithm calculates transfer function between track irregularities and carbody floor acceleration using Multiple Input Multiple Output (MIMO) system identification method. The later method analyses the carbody floor accelerations with Empirical Mode Decomposition followed by Singular Value Decomposition (EMD+SVD). These algorithms are evaluated on simulated carbody floor accelerations obtained with vehicle dynamic simulations. In this investigation, it is observed that the first method extracts more crucial information from carbody floor acceleration in comparison to EMD+SVD method. These features are planned to be used in future research to develop machine learning based intelligent fault identification algorithm for identification of root cause of vehicle running instability occurrence.
|
|
| 4. |
- Kulkarni, Rohan, 1991-, et al.
(author)
-
Investigating the effect of the equivalent conicity function's nonlinearity on the dynamic behaviour of a rail vehicle under typical service conditions
- 2022
-
In: Vehicle System Dynamics. - : Taylor & Francis. - 0042-3114 .- 1744-5159. ; 60:10, s. 3484-3503
-
Journal article (peer-reviewed)abstract
- Generally, the equivalent conicity function (ECF) is denoted by equivalent conicity at 3mm (λ3mm) and a Nonlinearity Parameter (NP). NP describes the nonlinearity of the ECF and its influence on a vehicle design is explored thoroughly, however, NP’s role in vehicle and track maintenance is not researched yet. This paper investigates the influence of track maintenance actions on vehicle dynamics with help of NP vs λ3mm scatter plots of ECF database. The ECF database is constructed by combining measured worn wheel and rail profile pairs of the Swedish high-speed vehicle and rail network, respectively. The ECF database revealed an inverse relationship between λ3mm and NP, i.e., NP is negative for larger λ3mm values. The combination of negative NP and high λ3mm causes reduction in the vehicle’s nonlinear critical speed and vehicle often exhibit the unstable running on the Swedish rail network. Thus, the occurrence of ECF with negative NP and high λ3mm is undersirable and the undesirable ECF can be converted into desirable ECF by grinding the rail, which converts ECF’s into positive NP and low λ3mm combinations. Thus, the NP parameter along with the λ3mm must be considered in track maintenance decisions.
|
|
| 5. |
- Kulkarni, Rohan, 1991-, et al.
(author)
-
iVRIDA-fleet: Unsupervised rail vehicle runninginstability detection algorithm for passenger vehicle fleet
-
Other publication (other academic/artistic)abstract
- Identifying faults contributing to unsafe conditions, such as a high-speed railvehicle running instability, is crucial to ensuring operational safety. But the occurrence of vehicle running instability during regular operation across the whole vehicle fleet is a rare anomaly. An unsupervised anomaly detection (AD) based iVRIDA-fleet framework is therefore proposed to detect vehicle running instability and identify its root cause. The performance of Principal Component Analysis (PCA-AD, baseline model), Sparse Autoencoder (SAE-AD),and LSTM Encoder Decoder (LSTMEncDec-AD) models are evaluated to detect the occurrence of vehicle running instability. A k-means algorithm is then applied to latent space representations to identify various clusters associated with different root causes of observed vehicle running instability.The effectiveness of the proposed iVRIDA-fleet framework is demonstrated using onboard accelerations measured on a Swedish X2000 vehicle fleet. The probability of vehicle running instability occurrence is observed to be only 0.35% of onboard accelerations corresponding to 827,467 km travel distance.Furthermore, the root causes identified by the iVRIDA-fleet framework are validated by investigating the maintenance records of the vehicles and track. It is identified that heavily worn wheels were the primary root cause of observed vehicle running instability, but the track (actual gauge and rail profiles) was also a contributing factor. The proposed algorithm contributes towards the digitalisation of vehicle and track maintenance by intelligently identifying anomalous events of the vehicle-track dynamic interaction.
|
|
| 6. |
- Kulkarni, Rohan, 1991-, et al.
(author)
-
Monitoring of Alignment Level (AL)and Cross Level (CL) track geometry irregularities from onboard vehicle dynamics measurements using probabilistic fault classifier
- 2021
-
In: Lecture Notes in Mechanical Engineering. - Budapest : Springer Science and Business Media Deutschland GmbH. ; , s. 479-487
-
Conference paper (peer-reviewed)abstract
- Condition monitoring of track geometry irregularities from onboard measurements of vehicle response is a cost-effective method for surveilling qual-ity of track irregularities on daily basis. The monitoring of Alignment Level (AL)and Cross Level (CL) track irregularities is challenging due to the nonline-arities of the contact between wheels and rails. Recently, the authors proposed a signal-based method in combination with a machine learning (ML) fault classi-fier to monitor AL and CL track irregularities based on bogie frame accelerations. The authors concluded that the Support Vector Machine (SVM) fault classifier outperformed other traditional ML classifiers. Thus, an important question arises: Is the previously reported decision boundary an optimal boundary? The objective of this research investigation is to obtain an optimal decision boundary according to theory of probabilistic classification and compare the same against the SVM decision boundary. In this investigation, the classifiers are trained with results of numerical simulations and validated with measurements acquired by a diagnostic vehicle on straight track sections of a high-speed line (300 km/h). A fault classi-fier based on Maximum A Posterior Naïve Bayes (MAP-NB) classification is developed. It is shown that the MAP-NB classifier generates an optimal decision boundary and outperforms other classifiers in the validation phase with classifi-cation accuracy of 95.9±0.2% and kappa value of 80.4±0.6%. Moreover, the Lin-ear SVM (L SVM) and Gaussian-SVM (G SVM) classifiers give similar perfor-mance with slightly lower accuracy and kappa value. The decision boundaries of previously reported SVM based fault classifiers are very close to the optimal MAP-NB decision boundary. Thus, this further strengthens the idea of imple-menting statistical fault classifiers to monitor the track irregularities based on dynamics in the lateral plane via in-service vehicles. The proposed method con-tributes towards digitalization of rail networks through condition-based and pre-dictive maintenance.
|
|
| 7. |
- Kulkarni, Rohan, 1991-
(author)
-
Onboard condition monitoring of vehicle-track dynamic interaction using machine learning : Enabling the railway industry’s digital transformation
- 2023
-
Doctoral thesis (other academic/artistic)abstract
- The railway sector’s reliability, availability, maintainability, and safety (RAMS) can significantly improve by adopting condition based maintenance (CBM). In the CBM regime, maintenance decisions are driven by condition monitoring (CM) of the asset. This thesis proposes machine learning (ML) based onboard CM (OCM) algorithms for CM of vehicle-track dynamic interaction via vehicle response (VR). More specifically, the algorithms are developed to monitor track irregularities (TI) and vehicle running instability incidences (VRII) via VR.CM of TI from onboard accelerations is a cost-effective method for daily surveillance of tracks. Most of the latest research is focused on monitoring vertical irregularity via vertical accelerations. Less attention is given to monitoring alignment level (AL) and cross level (CL) track irregularities. The PhD thesis proposes an ML based OCM algorithm to identify track sections with AL and CL track irregularities exceeding maintenance thresholds via bogie frame accelerations (BFAs). In this thesis, the OCM algorithm’s supervised ML models are trained on BFAs’ datasets synthesized with multibody simulation (MBS) of a high-speed diagnostic vehicle. Furthermore, the trained ML models and OCM algorithm are validated with measurements acquired by the same high-speed vehicle. The proposed OCM algorithm shows excellent performance in track quality surveillance only from BFAs. OCM of vehicle running instability (VRI) is important to ensure safety and onboard ride comfort. The latest research focuses on designing OCM algorithms for detecting VRI, but these OCM algorithms lack fault diagnosis (FD) of detected VRII. The PhD thesis proposes various OCM algorithms under an "intelligent vehicle running instability detection algorithm" (iVRIDA) umbrella to detect VRII and diagnose corresponding root causes via carbody accelerations. The occurrence of VRI during regular operation across a whole train fleet is an anomaly. Thus, an unsupervised anomaly detection (AD) based iVRIDA algorithm is proposed and later extended as iVRIDA-fleet for vehicle fleetwide application. The proposed OCM algorithms iVRIDA and iVRIDA-fleet are verified by onboard measurements of a European high-speed vehicle and the Swedish X2000 vehicle fleet.The thesis contributes towards the digitalization of vehicle and track maintenance by enabling adaptation of the CBM regime.
|
|
| 8. |
- Kulkarni, Rohan, 1991-, et al.
(author)
-
Unsupervised rail vehicle running instability detection algorithm for passenger trains (iVRIDA)
- 2023
-
In: Measurement. - : Elsevier. - 0263-2241 .- 1873-412X. ; 216, s. 112894-112894
-
Journal article (peer-reviewed)abstract
- Intelligently identifying rail vehicle faults instigating running instability from carbody floor acceleration is essential to ensure operational safety and reduce maintenance costs. However, the vehicle-track interaction's nonlinearities and scarcity of running instability occurrences complicate the task. The running instability is an anomaly in the vehicle-track interaction. Thus, we propose unsupervised anomaly detection and clustering algorithms based iVRIDA framework to detect and identify running instability and corresponding root cause. We deploy and compare the performance of the PCA-AD (baseline), Sparse Autoencoder (SAE-AD), and LSTM-Encoder-Decoder (LSTMEncDec-AD) model to detect the running instability occurrences.Furthermore, we deploy a k-means algorithm on latent space to identify clusters associated with root causes instigating instability. We deployed the iVRIDA framework on simulated and measured accelerations of European high-speed rail vehicles where SAE-AD and LSTMEncDec-AD models showed 97% accuracy. The proposed method contributes to smart maintenance by intelligently identifying anomalous vehicle-track interaction events.
|
|
| 9. |
- Kulkarni, Rohan, 1991-, et al.
(author)
-
Vehicle running instability detection algorithm (VRIDA): A signal based onboard diagnostic method for detecting hunting instability of rail vehicles
- 2021
-
In: Proceedings of the Institution of mechanical engineers. Part F, journal of rail and rapid transit. - : Sage Publications. - 0954-4097 .- 2041-3017. ; , s. 095440972110205-
-
Journal article (peer-reviewed)abstract
- In recent years, significant research transpired on onboard monitoring of various phenomena arising in dynamic vehicle-track interaction. One key issue being monitoring of vehicle hunting instability. Current hunting detection standards are appropriate for certification tests of vehicles, but incapable to monitor the health of the vehicle and track subsystems influencing the hunting instability. This paper proposes a signal based procedure for accurately triggering Hunting/No-Hunting alarm by conforming to requirements of onboard monitoring. A new method is conceived to reveal coherence among lateral and longitudinal accelerations during vehicle hunting. Furthermore, an index which amalgamates phase and amplitude information of lateral and longitudinal axlebox accelerations is introduced to detect coupled modes in lateral and yaw directions, i.e. hunting modes. Several simulations based pragmatic case studies are performed to assess the efficacy of the proposed procedure. The proposed method outperforms traditional hunting detection procedures by detecting more Hunting/No-Hunting occurrences. The proposed method contributes towards digitalization of rail vehicles through condition-based and predictive maintenance.
|
|
| 10. |
- Rosa, Anna De, et al.
(author)
-
Monitoring of lateral and cross level track geometry irregularities through onboard vehicle dynamics measurements using machine learning classification algorithms
- 2020
-
In: Proceedings of the Institution of mechanical engineers. Part F, journal of rail and rapid transit. - : SAGE Publications. - 0954-4097 .- 2041-3017.
-
Journal article (peer-reviewed)abstract
- In recent years, significant studies have focused on monitoring the track geometry irregularities through measurements of vehicle dynamics acquired onboard. Most of these studies analyse the vertical irregularity and the vertical vehicle dynamics since the lateral direction is much more challenging due to the non-linearities caused by the contact between the wheels and the rails. In the present work, a machine learning-based fault classifier for the condition monitoring of track irregularities in the lateral direction is proposed. The classifiers are trained with a dataset composed of numerical simulation results and validated with a dataset of measurements acquired by a diagnostic vehicle on the straight track sections of a high-speed line (300 km/h). Classifiers based on decision tree, linear and Gaussian support vector machine algorithms are developed and compared in terms of performance: good results are achieved with the three algorithms, especially with the Gaussian support vector machine. Even though classifiers are data driven, they retain the essence of lateral dynamics.
|
|
