Search

Lithium-ion batteries are ubiquitous in modern day applications ranging from portable electronics to electric vehicles. Irrespective of the application, reliable real-time estimation of battery state of health (SOH) by on-board computers is crucial to the safe operation of the battery, ultimately safeguarding asset integrity. In this paper, we design and evaluate a machine learning pipeline for estimation of battery capacity fade - a metric of battery health - on 179 cells cycled under various conditions. The pipeline estimates battery SOH with an associated confidence interval by using two parametric and two non-parametric algorithms. Using segments of charge voltage and current curves, the pipeline engineers 30 features, performs automatic feature selection and calibrates the algorithms. When deployed on cells operated under the fast-charging protocol, the best model achieves a root mean squared percent error of 0.45\%. This work provides insights into the design of scalable data-driven models for battery SOH estimation, emphasising the value of confidence bounds around the prediction. The pipeline methodology combines experimental data with machine learning modelling and can be generalized to other critical components that require real-time estimation of SOH., Comment: Peer review, pre-print to be published in Nature Machine Intelligence - 32 pages and 24 figures (including supplementary material)

Wheelset intermittent over-creeps (WIOs), i.e., slips or slides, can decrease the overall traction and braking performance of Electric Multiple Units (EMUs). However, they are difficult to detect and isolate due to their small magnitude and short duration. This paper presents a new index called variable-to-minimum difference (VMD) and a new technique called weighted moving average (WMA). Their combination, i.e., the WMA-VMD index, is used to detect and isolate WIOs in real time. Different from the existing moving average (MA) technique that puts an equal weight on samples within a time window, WMA uses correlation information to find an optimal weight vector (OWV), so as to better improve the index's robustness and sensitivity. The uniqueness of the OWV for the WMA-VMD index is proven, and the properties of the OWV are revealed. The OWV possesses a symmetrical structure, and the equally weighted scheme is optimal when data are independent. This explains the rationale of existing MA-based methods. WIO detectability and isolability conditions of the WMA-VMD index are provided, leading to an analysis of the properties of two nonlinear, discontinuous operators, $\min$ and $\textrm{VMD}_i$. Experimental studies are conducted based on practical running data and a hardware-in-the-loop platform of an EMU to show that the developed methods are effective.

So far, problems of intermittent fault (IF) detection and detectability have not been fully investigated in the multivariate statistics framework. The characteristics of IFs are small magnitudes and short durations, and consequently traditional multivariate statistical methods using only a single observation are no longer effective. Thus in this paper, moving average T^2 control charts (MA-TCCs) with multiple window lengths, which simultaneously employ a bank of MA-TCCs with different window lengths, are proposed to address the IF detection problem. Methods to reduce false/missing alarms and infer the IFs' appearing and disappearing time instances are presented. In order to analyze the detection capability for IFs, definitions of guaranteed detectability are introduced, which is an extension and generalization of the original fault detectability concept focused on permanent faults (PFs). Then, necessary and sufficient conditions are derived for the detectability of IFs, which may appear and disappear several times with different magnitudes and durations. Based on these conditions, some optimal properties of two important window lengths are further discussed. In this way, a theoretical framework for the analysis of IFs' detectability is established as well as extended discussions on how the theoretical results can be adapted to real-world applications. Finally, simulation studies on a numerical example and the continuous stirred tank reactor (CSTR) process are carried out to show the effectiveness of the developed methods.

The moving average (MA)-type scheme, also known as the smoothing method, has been well established within the multivariate statistical process monitoring (MSPM) framework since the 1990s. However, its theoretical basis is still limited to smoothing independent data, and the optimality of its equally or exponentially weighted scheme remains unproven. This paper aims to weaken the independence assumption in the existing MA method, and then extend it to a broader area of dealing with autocorrelated weakly stationary processes. With the discovery of the non-optimality of the equally and exponentially weighted schemes used for fault detection when data have autocorrelation, the essence that they do not effectively utilize the correlation information of samples is revealed, giving birth to an optimally weighted moving average (OWMA) theory. The OWMA method is combined with the Hotelling's $T^2$ statistic to form an OWMA $T^2$ control chart (OWMA-TCC), in order to detect a more challenging type of fault, i.e., intermittent fault (IF). Different from the MA scheme that puts an equal weight on samples within a time window, OWMA-TCC uses correlation (autocorrelation and cross-correlation) information to find an optimal weight vector (OWV) for the purpose of IF detection (IFD). In order to achieve a best IFD performance, the concept of IF detectability is defined and corresponding detectability conditions are provided, which further serve as selection criteria of the OWV. Then, the OWV is given in the form of a solution to nonlinear equations, whose existence is proven with the aid of the Brouwer fixed-point theory. Moreover, symmetrical structure of the OWV is revealed, and the optimality of the MA scheme for any IF directions when data exhibit no autocorrelation is proven., Comment: Automatica, Published online, 2020

This paper develops an explainable deep learning model that estimates the remaining useful lives of rotating machinery. The model extracts high-level features from Fourier transform using an autoencoder. The features are used as input to a feedforward neural network to estimate the remaining useful lives. The paper explains the model's behavior by analyzing the composition of the features and the relationships between the features and the estimation results. In order to make the model explainable, the paper introduces octave-band filtering. The filtering reduces the input size of the autoencoder and simplifies the model. A case study demonstrates the methods to explain the model. The study also shows that the octave band-filtering in the model imitates the functionality of low-level convolutional layers. This result supports the validity of using the filtering to reduce the depth of the model.