Estimation of State of Charge in Lithium-Ion and Solid-State Batteries Using Machine Learning Models

Innovation in electric mobility and energy storage from renewable energy resources are two key drivers in the fast growth of a battery industry that is striving to enhance performance of battery systems with great urgency. HQ Center of Excellence is actively working on the development of an advanced battery management system (BMS) and intelligence platform. Machine learning helps extract value from existing data to accelerate the optimization in the design of more effective BMSs. The primary objective is to build BMS technologies that improve the life and performance of lithium-ion and solid-state batteries in power electric vehicles and energy storage systems. Implementations of the BMS asks for the integration of both software and hardware, which includes battery state-of-charge (SOC) estimation, state-of-health (SOH) estimation, fault detection, control and monitoring tasks. This project will help Hydro-Québec to assess methods for predicting electric vehicle battery states. The project develops a data-driven machine learning model offering the most accurate predictions for SOC and SOH. It provides a case study for machine learning techniques accurately predicting the health and life of a battery.

Faculty Supervisor:

Vladimir Makarenkov

Student:

Partner:

Hydro-Quebec

Discipline:

Computer science

Sector:

Artificial Intelligence

University:

Université du Québec à Montréal

Program: