(Special) PHM for Rolling Stock in Railway Systems
Chair: Prof. Raphael Pfaff ( FH Aachen University of Applied Sciences)
Fault Detection and Diagnosis of Rolling Element Bearing based on Neural Network Ms. Chaeyoung LIM, Seokgoo KIM, HyungJun PARK and Joo-Ho CHOI (Korea Aerospace university, Korea Aerospace Univ)
Railway is one of the public transportation systems along with shipping and aviation. With the recent introduction of high speed train, its proportion is increasing rapidly, which results in the higher risk of catastrophic failures. The wheel bearing to support the train is one of the important components requiring higher reliability and safety in this aspect. Recently, many studies have been made under the name of prognostics and health management (PHM), for the purpose of fault diagnosis and failure prognosis of the bearing under operation. Among them, the most important step is to extract a feature that represents the fault status properly and is useful for accurate remaining life prediction. In this study, a neural network based diagnosis technique is developed to detect early fault and estimate the severity of the spalls at the inner race, outer race, roller and cage. To this end, a bearing testrig is developed, in which the normal and faulted bearings with differing spall size at different locations are operated under accelerated loading conditions. Features are extracted from the bearings, which include the time based indicators such as rms, peak, crest and kurtosis, frequency based indicators obtained by envelope analysis, and time-frequency based ones like wavelet decomposition. Neural network model is constructed using the features for the classification. The model is then applied to diagnose the fault in the new bearings, which includes the identification of the fault type and severity. Particular attention is maid to the study of statistical significance to check the validity and accuracy of the model prediction.
Effective Methods of Database Establishment and Classification for PHM of Korea Express Train (KTX) Dr. JONGSOON IM, JU WON KIM, JUNG HWA HAN, GI DO CHOI and JUNSIK IM (GLOBIZ, KORAIL)
The standard type 19” subrack measurement systems and composite sensors module with ADC and communication topology are developed for a long-term measurement of dynamic behavior of KTX major components and systems. The systems are fitted to be available for the commercial trains on considering of the aspects of safety and maintenance. By a long term autonomous measurement, database could be established and classified with multi-parameter for application to PHM as reference data. Our systems can provide a solution to overcome the unwilling of maintenance department on installing the extra measurement systems to the existing train. This work was supported by Research Program – Development of a early failure detection module for core parts of the rolling stock at onboard[16RTRP-B103888-03] funded by the Korea Agency for Infrastructure Technology Advancement(KAIA).
Design of Service Model & System Architecture for Maintenance Support System based on PHM Dr. Yonghoon CHOI and Hoon JUNG (ETRI)
The technologies developed for rolling stock maintenance can reduce unnecessary maintenance costs and railroad component damages that result in the suspension of railroad services. These technologies were first developed in advanced countries in Europe to support preventive maintenance, which meant that the railroad cars had to undergo inspections periodically. Note, however, that rolling stock maintenance technologies are now evolving to support predictive maintenance. In addition, as systems increase in speed, there are growing demands for fault diagnosis and prognostic health management techniques to address the challenges of both improving reliability and reducing maintenance costs.
This paper describes the function and architecture of a railroad car maintenance support system for effectively managing rolling stock components. The system can automatically detect and predict flaws in components by sensing data collected through a network of sensors placed at onboard and wayside.
A Study on Necessity to Introduce Prognostic Maintenance of Rolling Stock Mr. Ju Won KIM, Chan hoi AN, SEONG HEE LEE and Jung Hwa HAN (Korail(Korea Railroad))
The easiest way to prevent the railway vehicle failure is to heavily invest money and labor into maintenance and advance the vehicle replacement and maintenance cycles. However, the actual operational maintenance resources (cost and time) are limited, making it difficult to predict when and where sudden failures would occur. Therefore, many studies have examined how to build an effective maintenance system that can minimize failures based on limited resource availability. In this paper, I will examine the necessity of predictive maintenance and research trends.
Analysis of Big Data Streams to Obtain Braking Reliability Information for Train Protection Systems Prof. Raphael PFAFF, Parham SHAHIDI and Manfred ENNING (FH Aachen University of Applied Sciences, PARC)
For decades, the technology of freight railcars has not changed significantly, mostly due to little or no incentive for significant investments in rolling stock. Taking into account the disruptive developments anticipated in automotive transportation, this approach appears no longer feasible, especially if regulatory agencies aim to reduce carbon dioxide emissions while maintaining economic growth.
With the advent of telematics, on-board sensing and cloud-based analytics for control and condition based maintenance, high potential for efficiency improvements has become possible. Such technologies are de facto standards in automotive transport, which induces the need for implementation of similar technologies in rail transport as well.
In addition to enabling efficiency gains, telematics, on-board sensing and cloud-based analytics also offer new means to approach pressing problems such as rail noise emission, train integrity and safety against derailment, while at the same time reducing maintenance cost and downtime.
Furthermore, a connected wagon offers a seamless integration into current and future logistics systems, which are driven and controlled by the industrial Internet of Things to support the fourth industrial revolution. An important concept, introduced with the Wagon 4.0, is standardized hardware, together with an open-source operating system based on prognostics and health management principles for predictive analytics. Thus, the Wagon 4.0 paves the way for new operations and maintenance concepts, user interfaces and value proposals. Additional economic advantages will be made possible from the self-organizing features of such vehicles,the ability to achieve mass customization and from a rise in efficiency in operation and maintenance.
This paper describes the basis of such a system including the power supply, intra-train communications, sensing and cloud-based analytics. A study of use cases from railway operation illustrates the approach and highlights the opportunities of this novel system design. The paper concludes with a description of how the implantation enables the railcar operator to practice predictive maintenance and increase operational efficiency.