Data‐Driven Incipient Sensor Fault Estimation with Application in Inverter of High‐Speed Railway

Abstract

Incipient faults in high-speed railway have been rarely considered before developing into faults or failures. In this paper, a new data-driven incipient fault estimate (FE) methodology is proposed under multivariate statistics frame, which incorporates with Kullback-Leibler divergence (KLD) in information domain and neural network approximation in machine learning. By defining one sensitive fault indicator (SFI), the incipient fault amplitude can be precisely estimated. According to the experimental platform of China Railway High-speed 2 (CRH2), the proposed incipient FE algorithm is examined, and the more sensitivity and accuracy to tiny abnormality are demonstrated. Followed by the incipient FE results, several factors on FE performance are further analyzed.

Real‐time diagnosis of high‐speed rail traction transformer in different topologies

“…The circuit models of normal operation and fault operation of high‐speed rail traction transformer are established. By measuring the voltage signals of input and output points during operation and combining with fast ICA algorithm, the real‐time monitoring statistics [ I 2 and squared prediction error (SPE)] of micro‐faults are used as a measure of the normal operation of the process indicators, high‐speed rail traction transformer to achieve real‐time micro‐fault diagnosis, thereby enhancing the safety and reliability of high‐speed rail operation [5].…”

Data-Driven Fault Diagnosis for Electric Drives: A Review

“…For example, in ref. [184] a new datadriven incipient fault detector methodology was proposed via Neural Network algorithms for phase current, speed and DC link voltage sensors. It incorporates preprocessing algorithms such as PCA or Kullback-Leibler divergence (KLD) to extract important information from the acquired data.…”

“…Table 7 summarizes papers which perform a ML based data-driven fault detection and diagnosis methodology. [184] Bias and ramp i abc PCA and KLD ANN Test-bench [193] Ramp, stuck and offset i abc , ω and v dc -DeepPCA Test-bench [187] Generic faults i abc an ω PCA ANN and kNN Simu. [194] Bias and offset i abc , ω and v dc -PCA Test-bench [188] Generic faults Strain gauge -kNN Real data [189] Generic faults Temperature t-domain SVM Simu.…”

Health Assessment of High-Speed Train Running Gear System under Complex Working Conditions Based on Data-Driven Model

“…(2) Structural complexity due to the close coupling of components in the running gear system is difficult to be described by precise mathematical models through mechanism analysis, which will limit the use of analytical models Currently, the common methods for health status assessment of complex electromechanical systems are mainly divided into three categories: the method based on semiquantitative information [7][8][9], data-driven method [10][11][12][13], and model-based method [14][15][16]. With an increase in the number of sensors in the running gear system, it has become very easy to obtain a large amount of data that can reflect the actual status of the system.…”