Researchers from Chalmers University of Technology in Sweden have developed a new rapid charging method utilising artificial intelligence (AI) that reduces battery wear without affecting charging speed.
Fast or rapid charging an electric vehicle (EV) will always be a necessary part of a vehicle’s life – whether that’s for passenger EVs on long journeys or commercial vehicles like taxis or buses or construction equipment. However, fast charging takes its toll on the EV battery, shortening its lifespan due to unwanted side reactions when high currents are pushed into the battery cell.
Similarly, EV batteries currently have a lifespan anywhere from 8 to 15 years, although these numbers are still in dispute, with some companies and real-world examples touting longer lifespans. Consumers are understandably wary of these numbers, with some studies of the European EV market demonstrating such concern.
In a new study published in the journal IEEE Transactions on Transportation Electrification, the researchers from Chalmers University of Technology and Victoria University of Wellington in New Zealand show that it is possible to not only increase the lifespan of an EV battery but to do so without avoiding or reducing charging speed.
Specifically, the researchers present an AI-based charging strategy that adapts the current during east fast charge to the battery’s chemistry and ‘state of health’ (SoH).
Measuring battery lifetime in so-called equivalent full cycles (EFCs) – the number of charge and discharge cycles a battery undergoes before its capacity drops to 80 per cent of the original value, the threshold commonly considered the end of life for an EV battery – the researchers’ new AI-based charging strategy achieved a 22.9 per cent extension in EFCs compared with conventional charging.
Significantly, charging time remained virtually unchanged, coming in at 24.12 minutes on average, compared to 24.15 minutes for conventional charging.
The research was carried out by Changfu Zou, professor at the Department of Electrical Engineering at Chalmers, and Meng Yuan, Assistant Professor at Victoria University and a former researcher at Chalmers.
The new AI-based charging strategy relies on reinforcement learning – a variation of machine learning in which an algorithm learns by interacting with an environment (the battery), gradually improving its decisions based on received feedback. Or, put another way, right actions by the algorithm are rewarded and thus reinforced.
In this case, the AI model developed by Changfu Zou and Meng Yuan was trained to adapt the charging based on how charged or discharged the battery was at the time of charging, as well as the overall health of the battery.
This is in contrast to typical charging strategies which use fixed voltage and current limits regardless of whether the battery is new, partially charged, or older.
“This work shows that the true bottleneck of fast charging is not simply current limits, but the evolving electrochemical state inside the battery,” said Changfu Zou.
“By integrating AI with physics-based understanding, we move closer to health-aware charging strategies that maximize both performance and lifetime.”
Most importantly, this new charging method would be both easy and cost-effective to implement, requiring only a software update in the vehicle’s battery management systems – though the researchers acknowledge “some adaption” is needed for the method to be used more widely.
Similarly, the researchers’ next step is to test the method directly on physical batteries, suggesting there is some way to go before real-world benefits can be realised.
“There are not so many different battery types today, but the method needs to be calibrated for it to be used by everyone,” said Changfu Zou.
“Using transfer learning, we can take advantage of what our AI model has already learned and thus adapt the AI model to new batteries more quickly.”
Sign up for The Driven’s free daily newsletter and get the latest EV news and analysis delivered straight to your inbox.
