Sensors and Measurements in Machine-Learning-Based Battery Management Systems

Dear Colleagues,

In the last few decades, electrochemical accumulators have seen an enormous diffusion in many applications, e.g., consumer electronics, hybrid and electric vehicles and energy storage systems. In particular, lithium-ion batteries show the best performance in terms of high energy and power density, high energy and coulombic efficiency, low self-discharge and a negligible memory effect.

However, damage or improper management of the battery due to the presence of a flammable electrolyte can lead to explosions and fires. Moreover, these batteries experience degradation over their lifetime and are dramatically affected by different operating conditions.

Therefore, an appropriate Battery Management System (BMS) is required to guarantee the proper functioning of such complex devices. Among the different tasks undertaken by a BMS, those most frequently considered in the literature are state estimation, fault diagnosis, safety management and fast charging.

In the last few decades, several approaches have been proposed that rely on a first-principle mathematical model of batteries to accomplish the aforementioned tasks. The foundations of . To enhance the accuracy, an identification process can be exploited based on experimentally collected input and output data.

Note that the choice of an appropriate physics-based model and its level of detail are also crucial for achieving high battery performance. However, there is much debate in the literature on which electrochemical processes should be considered for modeling and which can be neglected. Data-driven methodology can be exploited to overcome such issues. In particular, machine learning algorithms can be used to fit a black box representation of the battery directly on measured data, by conducting a model selection process based on the performance achieved in validation.  Finally, learning-based control strategies can be applied to achieve optimal management policy for batteries by relying only on experimentally collected measurements. Within this context, model-free reinforcement learning and deep model predictive control seem to be promising solutions.

Notably, all the alternatives discussed here share a key element: the use of data and measurements collected through sensors installed on the battery pack.

This Special Issue focuses on battery management policies that are based on an appropriate and innovative exploitation of sensor measurements with the aim of enhancing battery performance. Original research, even that presenting opposing results, is welcome.

We will consider high-quality articles presenting original research results and review articles exploring technologies related to the potentiality of lithium-ion batteries and their management systems.

Articles reporting recent advances in battery management systems, parameters and state estimation for batteries, learning-based battery control, data-driven battery fault diagnosis, optimal displacement of sensors for batteries and closely related topics are also welcome.

Dr. Enrico Barbierato
Dr. Daniele Toti
Dr. Andrea Pozzi
Guest Editors

Manuscript Submission Information

Manuscripts should be submitted online at www.mdpi.com by registering and logging in to this website. Once you are registered, click here to go to the submission form. Manuscripts can be submitted until the deadline. All submissions that pass pre-check are peer-reviewed. Accepted papers will be published continuously in the journal (as soon as accepted) and will be listed together on the special issue website. Research articles, review articles as well as short communications are invited. For planned papers, a title and short abstract (about 100 words) can be sent to the Editorial Office for announcement on this website.

Submitted manuscripts should not have been published previously, nor be under consideration for publication elsewhere (except conference proceedings papers). All manuscripts are thoroughly refereed through a single-blind peer-review process. A guide for authors and other relevant information for submission of manuscripts is available on the Instructions for Authors page. Sensors is an international peer-reviewed open access semimonthly journal published by MDPI.

Please visit the Instructions for Authors page before submitting a manuscript. The Article Processing Charge (APC) for publication in this open access journal is 2600 CHF (Swiss Francs). Submitted papers should be well formatted and use good English. Authors may use MDPI's English editing service prior to publication or during author revisions.