[Paper Review] Implementation of an Improved Coulomb-Counting Algorithm Based on a Piecewise SOC-OCV Relationship for SOC Estimation of Li-IonBattery

Ines Baccouche, Sabeur Jemmali+3 more|arXiv (Cornell University)|Mar 27, 2018

Advanced Battery Technologies Research30 references18 citations

TL;DR

This paper presents an improved Coulomb-counting algorithm for Li-ion battery State-of-Charge (SOC) estimation using a piecewise linear SOC-OCV relationship to reduce cumulative errors. By periodically recalibrating the battery capacity using open-circuit voltage (OCV), the method achieves SOC estimation accuracy below 2% on a PIC18F MCU platform, significantly enhancing reliability for embedded multimedia systems.

ABSTRACT

Considering the expanding useofembedded devices equipped with rechargeable batteries, especially Li-ionbatteries that have higher power and energy density, the battery management systemis becomingincreasingly important. Infact, theestimationaccuracy of the amount of the remaining charges is critical as it affects the device operational autonomy.Therefore, the battery State-Of-Charge (SOC) is defined to indicate its estimated available charge. In this paper, a solution isproposed for Li-ion battery SOC estimation based on an enhanced Coulomb-counting algorithm to be implemented formultimedia applications.However,the Coulomb-counting algorithm suffers from cumulative errors due to the initial SOC andtheerrors ofmeasurements uncertainties,thereforeto overcome these limitations,we use the Open-CircuitVoltage (OCV),thushavinga piecewise linear SOC-OCV relationship andperformingperiodic re-calibration of the battery capacity. Thissolutionis implementedand validated on a hardware platform based onthePIC18F MCU family. The measured resultsarecorrelated withthetheoretical ones; they have shown a reliable estimation since accuracy is less than 2%.

Motivation & Objective

• To address the cumulative error in conventional Coulomb-counting methods due to initial SOC uncertainty and measurement noise.
• To improve SOC estimation accuracy in Li-ion batteries for embedded multimedia devices with high energy and power density.
• To implement a real-time, hardware-compatible solution using a piecewise linear SOC-OCV relationship for periodic recalibration.
• To validate the algorithm on a low-cost, resource-constrained microcontroller platform (PIC18F MCU).
• To achieve high estimation accuracy suitable for reliable battery management in portable electronic systems.

Proposed method

• The proposed method uses a piecewise linear approximation of the nonlinear SOC-OCV relationship to improve accuracy in SOC estimation.
• It integrates real-time Coulomb counting with periodic recalibration using measured open-circuit voltage (OCV) to correct cumulative errors.
• The algorithm estimates the current state of charge by integrating current over time and adjusting based on OCV-derived reference SOC values.
• The system performs recalibration at regular intervals or when OCV stabilizes, using lookup tables derived from experimental OCV-SOC data.
• The implementation is deployed on a PIC18F microcontroller, ensuring compatibility with embedded and low-power applications.
• The method leverages hardware-level ADC sampling and real-time processing to maintain accuracy under dynamic load conditions.

Experimental results

Research questions

• RQ1 How can cumulative errors in Coulomb-counting-based SOC estimation be effectively reduced in Li-ion batteries?
• RQ2 To what extent does using a piecewise linear SOC-OCV relationship improve SOC estimation accuracy compared to standard linear approximations?
• RQ3 Can periodic OCV-based recalibration significantly reduce drift in SOC estimation on a low-cost microcontroller platform?
• RQ4 What is the achievable estimation accuracy of the proposed algorithm in real-world embedded hardware?
• RQ5 How does the algorithm perform under varying temperature and load conditions in practical applications?

Key findings

• The proposed algorithm achieves an SOC estimation accuracy of less than 2% error, significantly improving upon conventional Coulomb counting.
• The use of a piecewise linear SOC-OCV relationship enables more precise mapping between OCV and SOC, reducing nonlinear error.
• Periodic recalibration using OCV effectively corrects cumulative integration drift in the Coulomb-counting process.
• The algorithm was successfully implemented and validated on a PIC18F MCU, demonstrating real-time feasibility in embedded systems.
• Measured results closely matched theoretical predictions, confirming the reliability and robustness of the method.
• The hardware implementation shows strong potential for deployment in portable multimedia and battery-powered devices.