The University of Auckland

Project #12: Machine learning models for estimating SoC of a Li-ion battery

Back

Description:

This project intends to design and develop a system that will track the charge of all cells inside a battery. It will then use this input to estimate where in the charge cycle the cell is and use this to adjust the belief of the cell’s capacity. The system will then be able to determine how to discharge the entire battery effectively by taking into account each cell’s capacity.

Objectives of the project are

Type:

Undergraduate

Outcome:

The end result of the project will ideally be a fully automated Smart Battery System consisting of the battery pack, an electrical load, a battery monitor and a controller. This system will be used to monitor the state of charge and capacity of each battery cell and adjust the discharge cycle accordingly.

The primary application for the Smart Battery System in this project is electric vehicles. However, the same technology can be applied to a vast class of applications such as mobile phones, laptops and any other device with a Lithium Ion battery pack.

Prerequisites

·   Power electronics for the circuitry of the battery system

·   Learning and decision-making algorithms

·   The specific application of lithium ion batteries in EVs

·   Programming of ARM core ST microcontroller

Specialisations

Categories

Supervisor

Co-supervisor

Team

Lab

Lab allocations have not been finalised