Search

Fault diagnosis is crucial to guarantee safe operation and extend the operating time while preventing the thermal runaway of the lithium-ion battery. This study presents a data-driven thermal runaway diagnosis framework where Bayesian optimization techniques are applied to optimize the hyperparameter of various machine learning techniques. We use different machine learning models such as support vector machine, naive Bayes, decision tree ensemble, and multi-layer perceptron to estimate a high likelihood of causes of thermal runaway by using the experimental measurements of open-source battery failure data. We analyze different evaluation metrics, including the prediction accuracy, confusion metrics, and receiver operating characteristic curves of different models. An experimental evaluation shows that the classification accuracy of the decision tree ensemble outperforms that of other models. Furthermore, the decision tree ensemble provides robust prediction accuracy even with the strictly limited dataset.

This paper applies the MIMO over-the-air computation (MIMO-AirComp) technique to compute the control signal of a distributed control system, including state estimation. The control-based target function is expressed as a nomographic function and efficiently computed through the wireless channel with suitable function decomposition. The proposed scheme leverages local pre-processing at the individual sensor, aggregation through the multiple access channel, and post-processing at the controller to compute the control-oriented nomographic function. We analyze the performance of our proposed scheme in terms of the time delay and computation error compared to a state-of-the-art MIMO time-division multiple access scheme by means of Monte Carlo simulations. Numerical results indicate that the proposed scheme considerably improves control performance, particularly in noisy networks that rely on a large number of wireless sensors, with low complexity for resource-constrained scenarios.

MIMO over-the-air computation (MIMO-AirComp) is a recently proposed technique that leverages the superposition property of the multiple access channel to compute the target multifunction of various applications. This article presents how the MIMO-AirComp principle can be applied to the state estimation problem using distributed sensing data. The representative target function is explicitly formulated as a nomographic function matched to the structure of the multiple access channel with the proper processing function. The proposed framework efficiently computes the target multifunction by coordinating local preprocessing at each node, aggregation through the wireless channel, and postprocessing at the fusion center. We analyze and demonstrate that the proposed approach significantly improves the computation throughput for the distributed estimation application. Specifically, the proposed MIMO-AirComp framework outperforms the conventional separated communication and computation approach when the network system relies on noisy measurements obtained by the densely deployed sensors.

Although various deep learning techniques have been proposed to diagnose industrial faults, it is still challenging to obtain sufficient training samples to build the fault diagnosis model in practice. This paper presents a framework that combines wavelet transformation and transfer learning (TL) for fault diagnosis with limited target samples. The wavelet transform converts a time-series sample to a time-frequency representative image based on the extracted hidden time and frequency features of various faults. On the other hand, the TL technique leverages the existing neural networks, called GoogLeNet, which were trained using a sufficient source data set for different target tasks. Since the data distributions between the source and the target domains are considerably different in industrial practice, we partially retrain the pre-trained model of the source domain using intermediate samples that are conceptually related to the target domain. We use a reciprocating pump model to generate various combinations of faults with different severity levels and evaluate the effectiveness of the proposed method. The results show that the proposed method provides higher diagnostic accuracy than the support vector machine and the convolutional neural network under wide variations in the training data size and the fault severity. In particular, we show that the severity level of the fault condition heavily affects the diagnostic performance.