Electrical Energy Storage Technology

Battery Management - Hardware and Algorithms

Battery management includes the monitoring, control, and protection of batteries, making it an essential part of any battery system. Battery management must meet different complex requirements based on the storage application and cell chemistry. Thus, the number of tasks to be monitored and parameters to be regulated is many times greater for a traction battery in an electric vehicle than for a mobile phone battery.

Management scope

Exceeding the maximum permissible current or cell voltage limits can lead to cell damage or even failure. As such, single cell monitoring of the electrical variables as well as limiting the battery current are among the core tasks of battery management.

Temperatures which are either too high or too low have a negative impact on the life of the cell and can in extreme cases result in internal short circuits or a thermal runaway. Temperature monitoring of the battery system, the modules or even the individual cells by the battery management system is therefore necessary. For passive systems, the battery current is limited to a greater extent if necessary. The temperature can be controlled directly using active air conditioning.

The battery management system is also responsible for balancing cells with different states of charge within a series connection

as well as determining the state of the battery system. In particular, accurate state of charge (SOC) estimation, both for the individual cells of a pack and for the entire battery, is significant for the superordinate energy management and estimation of the remaining range or service life in mobile applications.

Important cell parameters, such as capacity and internal resistance, change over the life of a cell. Therefore, the value or state of health (SOH) of the parameters is continuously estimated based on the electrical measurements and stored models.

In order to avoid violating the operating limits, the charge acceptance (CA), the maximum possible performance (CC, cold cranking) as well as the energy reserves are calculated on the basis of measurements and the estimated states and cell parameters. This information (SOF, state of function) is communicated to the superordinate energy management.

Address

Technische Universität Berlin
Electrical Energy Storage Technology
Institute of Energy and Automation Technology
Faculty IV
Office code EMH 2
Einsteinufer 11
D-10587 Berlin

Contact

Office EMH 2
Building EMH
Room EMH 255

Battery management systems

To perform the necessary tasks, the battery management system (BMS) must have circuitry for detecting cell voltages, battery current, and temperatures. Balancing circuits are used to balance the state of charge of cells within a row. There are various systems that are able to discharge individual cells (passive balancing systems) and transfer energy between the individual cells (active balancing systems) as well as systems that are able to provide the energy of single cells for the application or charge cells individually (redistribution). Because of their simpler structure, passive balancing systems are most widely used, even though the capacity of a series connection is limited by the cell with the smallest capacity. Active systems must have inductors or capacities and a transformer for the energy transfer between single cells in addition to the pure resistors and switches. This increases the available capacity but also the cost. Redistribution requires a DC-to-DC converter for each cell. As a result, these systems are very costly. However, the proportion of actual available energy of the series connected cells is greatest with this system.

Algorithms/methods

Various methods and algorithms are used to determine the battery states and parameters. To determine the state of charge, the ampere-hour count with open-circuit voltage correction or non-linear Kalman filters can be used. For parameter tracking of the capacity, these methods are expanded to include, e.g. least-square fitting methods or the use of dual or joint Kalman filters. Cell impedance can be determined online during load jumps or via pulse measurements. Vehicle-suitable EIS measurements are also conceivable to detect the impedance over a wide frequency range. Moreover, self-learning algorithms and back-end based methods can be used to make a prognosis in addition to the parameter determination.


In addition to the methods for state and parameter determination, balancing algorithms must be implemented. These algorithms are divided into voltage and ampere hour-based methods. There are also algorithms in development and research that allow for regenerative balancing and thus actively contribute to extending the life of the battery.

Hardware-in-the-loop simulation

Due to the high energy densities and potential dangers in some cell chemistries, BMS are subject to particularly stringent requirements. They are essential for safety. The functionality of newly developed BMS must therefore always be tested according to established standards. This entails a long development and evaluation period for BMS.

As an alternative to real cells, hardware platforms that emulate the behavior of the cells can be used in the test phase. Using these battery simulators, hardware-in-the-loop tests are performed to obtain reliable test results with minimal resources. The use of HiL simulators capable of delivering and receiving power allows validation of both the BMS circuits and algorithms. In addition, the behavior of BMS outside the system boundaries can be examined without causing a fire or explosion risk by exceeding limit values ​​of real cells.