DYNAMIC CONTROL OF BATTERY CHARGING

Organization Name

Inventor(s)

Bharatkumar Hegde of Bloomfield Hills MI (US)

Chen-fang Chang of Bloomfield Hills MI (US)

DYNAMIC CONTROL OF BATTERY CHARGING - A simplified explanation of the abstract

This abstract first appeared for US patent application 18181092 titled 'DYNAMIC CONTROL OF BATTERY CHARGING

The system described in the patent application is designed to control a battery system during the charging process by utilizing real-time data and machine learning algorithms.

Acquires a set of charging parameter measurements during the charging process.
Estimates a dynamic performance variable related to an electrochemical phenomenon occurring within the battery system.
Selects a stored charging profile based on the charging parameter measurements.
Applies a charging current to the battery system based on the selected stored charging profile.
Inputs the charging parameter measurements, stored charging profile, and stored relation to a learning agent.
Evaluates the stored charging profile through a reward-based learning process.
Periodically updates the stored charging profile based on the learning process.

Potential Applications: - Electric vehicles - Renewable energy storage systems - Portable electronic devices

Problems Solved: - Optimizing battery charging processes for improved performance and longevity - Enhancing energy efficiency in various applications

Benefits: - Increased battery lifespan - Improved charging efficiency - Enhanced overall system performance

Commercial Applications: Title: "Smart Battery Management System for Enhanced Performance" This technology can be utilized in electric vehicle charging stations, solar energy storage systems, and consumer electronics to optimize battery performance and efficiency, leading to cost savings and improved user experience.

Questions about the technology: 1. How does the system adapt to different types of batteries and charging scenarios? The system uses machine learning algorithms to analyze data and adjust charging profiles accordingly.

2. What kind of data is considered in the learning process to update the charging profiles? The system takes into account charging parameter measurements, electrochemical performance variables, and historical data to optimize the charging process.

Original Abstract Submitted

A system for control of a battery system includes a processor connected to the battery system and configured to perform, in real time during a charging process, acquiring a set of charging parameter measurements, estimating a dynamic performance variable related to an electrochemical phenomenon occurring within the battery system during the charging process, and selecting a stored charging profile from a stored relation based on the charging parameter measurements. The processor is further configured to perform applying a charging current to the battery system based on the selected stored charging profile, inputting the charging parameter measurements, the stored charging profile and the stored relation to a learning agent, and evaluating the stored charging profile according to a reward-based learning process, the learning process including estimating a performance value associated with the stored charging profile. The processor is configured to periodically update the stored charging profile based on the learning process.