Machine learning algorithm predicts how to get the most out of electric vehicle batteries — ScienceDaily
Researchers have developed a machine learning algorithm that could help reduce charging times and extend battery life in electric vehicles by predicting how different driving patterns will affect battery performance, improving safety and reliability.
The University of Cambridge researchers say their algorithm could help drivers, manufacturers and businesses get the most out of the batteries that power electric vehicles by suggesting routes and driving patterns that minimize battery drain and charging times.
The team developed a non-invasive way to examine batteries and get a holistic view of battery health. These results were then fed into a machine learning algorithm that can predict how different driving patterns will affect the future health of the battery.
In commercial development, the algorithm could be used to, for example, recommend routes that get the driver from point to point in the shortest amount of time without damaging the battery, or recommend the fastest way to charge the battery without damaging it. The results are published in the journal nature communication.
The health of a battery, whether in a smartphone or a car, is far more complex than a single number on a screen. “Battery health, like human health, is a multidimensional thing and can deteriorate in many different ways,” said first author Penelope Jones of the Cavendish Laboratory in Cambridge. “Most battery health monitoring methods assume that a battery is always used in the same way. But that’s not how we use batteries in real life. When I stream a TV show on my phone, the battery drains a lot faster than when I use it for news. The same goes for electric cars – how you drive affects how the battery degrades.”
“Most of us will replace our phones well before the battery degrades to the point where it’s unusable, but for cars, batteries need to last five, 10 years or more,” said Dr. Alpha Lee, who led the research. “Battery capacity can change drastically over that time, so we wanted to find a better way to check battery health.”
Researchers developed a non-invasive probe that sends high-dimensional electrical pulses into a battery and measures the response, providing a range of “biomarkers” of battery health. This method protects the battery and does not lead to any further degradation.
The battery’s electrical signals were converted into a description of the battery’s condition, which was fed into a machine learning algorithm. The algorithm was able to predict how the battery would behave in the next charge-discharge cycle, depending on how fast the battery was charged and how fast the car would go on the next trip. Tests on 88 off-the-shelf batteries showed that the algorithm did not need any information about the battery’s previous use to make an accurate prediction.
The experiment focused on lithium cobalt oxide (LCO) cells, which are widely used in rechargeable batteries, but the method is generalizable to the different types of battery chemistries used in electric vehicles today.
“This method could add value to so many parts of the supply chain, whether you are a manufacturer, end user or recycler, because it allows the condition of the battery to be understood beyond a single number and because it is predictive,”, said Lee. “It could reduce the time needed to develop new battery types because we can predict how they will degrade under different operating conditions.”
The researchers say that in addition to manufacturers and drivers, their method could also be useful for companies that operate large fleets of electric vehicles, such as logistics companies. “The framework we’ve developed could help companies optimize the use of their vehicles to improve overall fleet battery life,” Lee said. “There is so much potential in a framework like this.”
“It was such an exciting framework to build because it could solve so many challenges in today’s battery space,” Jones said. “It’s a great time to get involved in battery research, which is so important to addressing climate change by moving away from fossil fuels.”
Researchers are now working with battery manufacturers to accelerate the development of safer, longer-lasting next-generation batteries. They are also investigating how their framework could be used to develop optimal fast-charging protocols to reduce EV charging times without causing degradation.
The research was supported by the Winton Program for the Physics of Sustainability, the Ernest Oppenheimer Fund, the Alan Turing Institute and the Royal Society.