Battery Charging MPC with Predictive Waveform Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional battery charging methods are inefficient, taking hours and causing performance degradation due to the lack of optimized charging signals that account for various battery parameters like temperature, state of charge, and impedance.
Innovation Solution
A model predictive controller (MPC) architecture that uses processing units with executable instructions to generate optimized charge signals based on predicted battery parameters, such as state of charge, temperature, and impedance, by iterating through constraints and cost functions to alter charge signal parameters and shape the charge waveform.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional charging methods are used, then charging process is simple, but charging time is long and battery performance degrades
Solution Approach 1:
The controller performs preliminary predictions of battery state (temperature, SOC, impedance) before applying charge signals. By predicting future battery states based on current conditions and proposed charge signals, the system proactively adjusts charging parameters to prevent harmful effects before they occur, enabling faster charging without degradation.
Solution Approach 2:
The system implements closed-loop feedback by continuously measuring battery parameters (voltage, current, temperature), comparing actual states with predicted states, and adjusting charge signals accordingly. This feedback mechanism allows the controller to optimize charging in real-time, resolving the contradiction between fast charging and battery protection.
2Productivity
If high charge current is applied to reduce charging time, then charging speed increases, but battery temperature rises and performance degrades
Solution Approach 1:
The controller dynamically adjusts charge signal parameters based on real-time battery conditions. By continuously monitoring temperature and other state parameters, the system adapts charging current levels to maintain optimal temperature ranges, enabling fast charging while preventing thermal runaway and performance degradation.
Solution Approach 2:
The system changes charging parameters (current, voltage, frequency) based on predicted battery states. By adjusting these parameters dynamically according to temperature predictions and actual measurements, the controller achieves fast charging without exceeding safe temperature thresholds.
3Productivity
If optimized charge signals are generated using MPC architecture, then charging efficiency improves and battery health is optimized, but computational complexity and processing requirements increase
Solution Approach 1:
The controller segments the complex MPC problem into separate prediction models for different battery parameters (temperature prediction, SOC prediction, impedance prediction). Each model handles a specific aspect of battery behavior, making the overall system more manageable and computationally efficient while maintaining high charging efficiency.
Solution Approach 2:
The patent introduces intermediate prediction models that act as mediators between raw sensor data and final control decisions. These models (temperature prediction, SOC prediction, impedance prediction) process information in stages, reducing computational burden while achieving optimized charging efficiency and battery health management.
Data Source
AI summary
A model predictive controller and related charging components producing a charge signal for a battery wherein predicted battery parameters such as state of charge, battery temperature, state of health (e.g., anode overpotential), are used to generate constraints that are subsequently used, such as through an optimizer running a cost function, to produce a charge signal that may include one or more optimized charge attributes including a charge current magnitude or a mean current, a shaped leading edge, an edge time, a body time, and a rest time.