Battery Management Systems - Accurate State-of-Charge Indication for Battery-Powered Applications

"Chapter 3

A State-of-Charge indication algorithm (p. 47-48)

As discussed in chapter 2, many advances have been made in State-of- Charge (SoC) indication in recent years, both through continued improvement of the SoC algorithms and through the development of more accurate hardware systems. Nevertheless, there is still no ""ideal"" SoC system that gives accurate indications under all realistic user conditions. The ""ideal"" SoC system is obviously one that is not expensive, can handle all battery chemistries, can operate over a wide range of load currents and can deal with the aging effect. Leading semiconductor companies (e.g. Philips [1]–[3], NXP Research, Texas Instruments [4]–[6], Microchip [7], [8] Maxim [9], [10], etc.) are paying more and more attention to accurate State-of-Charge indication in attempts to find that ideal system.

A SoC algorithm that combines some form of adaptivity with direct measurement and book-keeping systems was developed and implemented by Bergveld et al. in 2000 [1]–[3]. By implementing the mathematical models described in [1], this algorithm was found to be the most sophisticated and accurate [11], [12]. This chapter will give a complete description of this algorithm, which serves as the starting point of this book. This chapter is organised as follows.

An introduction to the algorithm is given in section 3.1. Section 3.2 describes the models and states of the SoC indication system. The main aspects of the algorithm are given in section 3.3. The focus in section 3.4 is on accuracy problems. Section 3.5 presents concluding remarks.

3.1 An introduction to the algorithm

The SoC indication algorithm presented by Bergveld et al. in [1]–[3] aims to eliminate the main drawbacks and combine the advantages of the direct measurement and book-keeping methods described in Chapter 2. The basis of the SoC algorithm is Electro-Motive Force (EMF) measurement during equilibrium and current measurement and integration during charge and discharge. During discharge, in addition to simple Coulomb counting, the effect of the overpotential is also considered [1]. A method has also been developed for updating the value of the maximum capacity for coping with capacity loss due to the aging effect. The algorithm will be described below for a Panasonic CGR17500 Li-ion battery, but the basis of the algorithm holds for other types of Li batteries, too. The rated capacity of this battery is 720 mAh.

3.2 Battery measurements and modelling for the State-of-Charge indication algorithm

The battery model applied in the developed SoC indication algorithm describes the battery EMF and overpotential behaviour, neither of which can be measured directly. The EMF and overpotential curves have been measured with an accurate battery tester and implemented in the Battery Management System (BMS) using mathematical-function approximations [1], [13]. Both the measurement and the implementation method contribute to the final accuracy of the SoC indication.
"