The BMS was originally designed for VRLR batteries in uninterruptible power supply systems (UPS), but was also tested in electric vehicles (VW CityStromer, BMW). In a second step, a universal battery management system (BMS) was realized as a LabView application. The use of a personal computer instead of a microcontroller leads to much higher flexibility of the BMS and allows easy adaptation to various kinds of battery applications and battery technologies. The LabView-based BMS controls data acquisition, performs data processing, visualization and storage and provides a graphic user interface.
Apart from monitoring features, the BMS evaluates the measured data and interacts with external components, such as the charger, the temperature regulation system and the inverter controller. A modem battery management system, in contrast to simple battery monitors, is capable of actively affecting battery operation. Before the presentation of the new measuring hardware and the LabView-based battery management system, the main principles and general structure of a BMS are discussed. I INTRODUCTION I1
Strong requirements concerning battery life-time, reliability and energy-efficiency are imposed on modem battery applications, e. g. , on batteries in unintermptible power supply systems (UPS) or batteries in electric vehicles (EV). These high demands can only be met by employing sophisticated battery monitoring and management systems. At present, battery users (e. g. in telecommunication energy supply or in power stations) know too little about the state of their batteries to draw economically optimized decisions concerning maintenance and replacement.
The employment of a battery monitoring and management system helps finding the right time for battery maintenance and replacement and, in addition, will lengthen the service intervals due to taking active influence on battery operation. GENERAL STRUCTURE OF A BMS Fig. 1 shows the general structure of a battery management system and divides the tasks to be performed into logical blocks [2, 8, 91. Battery management systems require battery data (battery block-voltages, current and temperature) and environmental data (temperature). Furthermore, application depending system data is needed.
To obtain these values, a specialized data acquisition system is used. An example of a suitable data acquisition system for BMS is pre1. sented in chapter 1 1 The following logical block of a BMS (see Fig. 1) performs the data processing. Apart from decoding the transmitted measured values, this block calculates battery quantities (e. g. entire battery voltage, average battery temperature, etc. ), integrates quantities like the battery current and performs statistical analysis (e. g. distribution of block voltages, deviation of block voltages from the average block voltage, etc. . Furthermore, the data procession unit needs information about the present maximum and minimum limit of block voltages, the maximum battery current etc. This information is either deduced from the system’s battery model or can be provided directly by the user (“parameter correction”). The parameters of the battery model have to be continuously adapted to the measured and computed quantities (e. g. , temperature, average load current) to ensure that the model accurately represents the battery’s present state. 630
To control battery operation, battery management systems require the accurate measurement of battery block voltages, battery current and battery temperatures. Therefore, in a first step, a modular data acquisition system, specifically designed for battery applications, was built. The measuring hardware consists of two different types of monitoring units, a battery block-voltage monitoring unit (VMU) and a battery current- and temperature monitoring unit (CMU), which have been developed in cooperation with SIEMENS AG, Erlangen.
VMUs have already been employed in the SIEMENS Masterguard UPS in order to provide the system’s microcontrollerbased BMS with the required measured data. 0-7803-5069-3 /98/$10. 00 01998 IEEE Fig. 1 Structure of a battery monitoring and management system (BMS). In Fig. 1, the logical block “parameter adaptation” is responsible for battery parameter determination and update. The “monitoring” block in Fig. 1 performs fault detection and user information activities. Fault detection checks if a battery quantity has exceeded or is likely to exceed its allowed limits.
Measured data, alarm messages and information related to service or maintenance needs are given to the user interface by the monitoring block. In contrast to the monitoring block, the “management” block is responsible for fault avoidance, which aims to keep the battery within certain limits of operation. If a quantity is going to exceed a limit, the management system reacts, for example, by activating the cooling system. In extreme situations, e. g. , in the case of exhaustive discharge, the fault avoidance may limit or even interrupt the current to protect the cells against reversion.
Furthermore, the management block is able to control intelmethod, preligent charging algorithms like the IU,,, sented in , which limits the maximum of each battery block-voltage instead of just limiting the entire battery voltage during the time of constant-voltage charge. In Fig. 1, the external components, which are influenced by the management block (charger, coolingheating system, inverter controller) are summed up in the “control” block. The hnctional block “in-/output” represents the user interface which gives selected data and messages to the user and which allows user intervention, e. . the correction of battery or system parameters. Data selection as well as data storage are represented through the logical block “data management”. 111 DATA ACQUISITION SYSTEM Fig. 2 shows the structure of the measuring and data acquisition system, consisting of battery block-voltage monitoring units (VMU) and current- and temperature monitoring units (CMU) [ 1,2]. Data transfer between the measuring units (any number and sequence is possible) and the data acquisition unit (DAU), which represents the interface to the data processing block of Fig. , is performed via a fiber-optic transmission system (FOTS). Optic data transmission provides electric insulation of the modular measuring system and leads to high immunity against electromagnetic noise. Power supply for the VMU and CMU is drawn directly from the measured batteries which minimizes the number of electrical connections and therefore leads to an easy and fast installation. The VMU measures up to eight voltages each ranging from 0. 3 V to 16 V with an accuracy of 0. 15 YO, which enables the measurement of single-cell and block voltages for any type of battery.
Depending on the shunt resistor, the CMU is able to measure battery current, e. g. , over a range fiom OA up to f 300 A. The CMU also measures battery temperature (-5OC to 6OOC). When the battery management system does not need measured data, the monitoring units rest in a stand-by mode during which the modules’ power consumption is negligible compared to the self-discharge of the batteries. As soon as measured values are needed, the units are switched on by the central data acquisition unit using the fiber-optic transmission ring.
After some setup-routines are completed, the VMUs and CMUs start the measuring process. As soon as data is available, the modules store the measured values in their transmission buffers. Receiving a so called “data-locomotive”, the first module 631 214 –. . – I t Data Acquisition Unit I 1 I —-I Fibre Optic Transmission System (FOTS) T T t 8 block voltages Fig. 2 8 block voltages 8 block voltages 5 temperatures and 3 currents t 5 temperatures and 3 currents t Structure of the measuring and data acquisition system [ 11.
The electric data input and data output signals of the serial port (TxD, RxD) are transformed into optic signals and are transmitted via the fiber optic transmission system. The electric handshake signals RTS and CTS as well as DSR, DTR and DCD are not transmitted so that, on the one hand, only one transmission ring is needed, but, on the other hand, hardware handshake signals cannot be used. Therefore, during the development of the LabView BMS much effort had to be taken to ensure the synchronization of data communication between the measuring units (CMU and VMU) and the personal computer.
Having started the Labview application for the first time, the user is asked to provide obligatory system information, e. g. , the number and sequence of the measuring units, the value of the shunt resistor for current measurement, the time between new measurement requests and the paths for data storage. At the moment, the presented system is pre-configured to deal with up to eight measuring modules. The default time interval between measurement requests is set to four seconds, but can be augmented as well as shortened (down to one second) by the user. After the receipt of a valid data chain, LabView decodes the transmitted data.
If the decoded data is recognized as measured values (and not as error message) these values are compared with the corresponding maximum and minimum limits. If a measured value exceeds the tolerated interval an alarm message is generated and shown on the graphic user interface. The minimum and maximum limits may be set by the user, or be deduced from an appropriate battery model. User interface: A screen shot of the user interface called “front panel”, can be seen in Fig. 3. On the front panel, the user can provide input information. Besides, the average battery interrupts the measuring process and starts sending the data-locomotive ollowed by its own measured data to the next module in the ring. This module recognizes the locomotive and the transmitted data and attaches to the data chain its own measured values. The central acquisition unit is able to decode the received data chain as it knows the order of the monitoring units [l]. IV LABVIEW-BASED UNIVERSAL AND BATTERY MONITORING MANAGEMENT SYSTEM (BMS) LabView (Laboratory Virtual Instrument Engineering Workbench) from National Instruments Corporation is a software development application which uses a graphical programming language, G, to create programs in block diagram form.
Since LabView includes libraries of functions for data acquisition, serial instrument control, data analysis, data presentation and data storage it commends itself for the BMS application. LabView programs are called “virtual instruments (VI)”. These VIS consist of an interactive user interface (“front panel”, see Fig. 3), a dataflow diagram that serves as the source code, and icon connections that allow the VI to be called from higher level VIS [ 5 ] . As the BMS is executed on a personal computer, it offers higher flexibility and much more graphic tools for data visualization than microcontroller-based systems.
Therefore, the central unit of the diagram in Fig. 1 as well as the in-/output interfaces have been realized as a LabView application. Data acquisition: For communication between the measuring system and LabView (request for and receipt of measured values) one of the serial ports of the personal computer is used. 632 27-4 Fig. 3 Front panel of the LabView-based BMS eight measuring channels and can localize the battery block, which is exceeding its operation limits. Fig. 4 shows the display for module 1. The measured voltage of block 1 has exceeded ts maximum limit and, therefore, has caused the alarm message on the front panel. If the is in doubt, whether the battery,s usable ca- temperature, up to six currents of parallel battery strings and the entire battery voltage are constantly monitored. Fault detection: In case of an alarm message, the corresponding monitoring unit is marked. Clicking on the “module #,’button, the user gets detailed information about the module’s block 1 1 block2 1 block3 1 block4 block5 block6 block7 block8 i Fig. 4 Detailed information about measuring module 1 633 acity might have decreased, it is possible to run a battery-check discharge, which helps recognizing and localizing defective cells. This battery-check, which could be defined as a part of the fault detection functions (monitoring) of the BMS, but also as a part of active intervention (management) is described below (see “fault avoidance and intervention”). Data processing: ment, the battery charger, which performs a blockvoltage limiting and therefore battery protecting charging algorithm (1Uma), is connected to the BMS via a second RS 232 serial port. It is anticipated that future BMS applications, e. . , BMS in electric vehicles, will use modem fieldbus configurations, such as CAN bus, which will lead to more flexible, open systems. Data management: Apart from calculating the entire battery voltage and the average temperature, LabView calculates the battery’s state of charge using a modified Peukert equation [2,6,7]. Switching the button “SOC” the state of charge of the battery strings is shown on the display. Furthermore, the distribution of block voltages is displayed. The histogram of the measured voltages can be seen by clicking on the “voltage distribution” button.
Fault avoidance and intervention: Depending on the width of the voltage distribution, e. g. during floating operation, or depending on a decreasing usable battery capacity, the BMS is able to recognize if a so-called “battery conditioning” (in case of lead acid batteries) should be performed and gives a corresponding message to the user . A “battery conditioning” algorithm performs a special sequence of discharging and recharging periods which reverses to a certain extent aging of the batteries. A typical example of reversible damages of lead acid batteries is the so-called “premature capacity loss” .
To allow analysis of the data at a later moment, the measured values as well as the messages given to the user are continuously stored. This default storage mode is useful for battery examination in research and development. Using the BMS for common battery applications an event-controlled storage mode can be chosen, which stores a measured value, if it is significantly different form the value before. For future commercial battery applications, for example, battery leasing, another simple storage mode, which only records the alarm messages on the user display, could easily be employed.
V SUMMARY To determine the decreasing battery capacity in case of batteries during floating operation (UPS-batteries), it is possible to run a battery-check discharge (e. g. 10% of nominal battery capacity). During the battery check the battery block-voltages are recorded and compared with the corresponding values from previous tests. In case of significant deviations, the battery’s usable capacity Qo has probably decreased as well. Therefore, a “battery conditioning” has to be recommended.
If a measured temperature exceeds its limits or the differences between the measured temperatures exceed a predefined limit, a water circulation system is switched on. This system may either cool or heat the batteries or may just establish a uniform temperature distribution. Parameter adaptation: In this paper, a new developed, flexible LabView-based battery management system is presented. Through the LabView realization of the BMS (in contrast to older microprocessor based concepts ) it was possible to create an easily adaptable monitoring and management system for any kind of battery application.
Additional monitoring or management features for BMS can easily be added to the LabView software. Even unforeseen demands on BMS, that may turn up with the usage of new battery technologies, can be met without problems. VI ACKNOWLEDGEMENTS The authors would like to thank SIEMENS AG, Erlangen for supporting the development of the presented battery measuring system and, furthermore, acknowledge D. Linzen for his contributions to this project. VI1 REFERENCES A. Lohner, S. Buller, E. Karden, R. W.
De Doncker, “Development of a Highly Accurate, Universal and Inexpensive Measuring System for Battery Management Systems”, 15” Electric Vehicle Symposium (EVS), 1998, Brussels (Belgium) E. Karden, P. Mauracher, A. Lohner, “Battery Management Systems for Energy-Efficient Battery Operation: Strategy and Practical Experience”, 1 3 ~ Electric Vehicle Symposium, 1996, Osaka (Japan), Vol. 2 pp. 91-98 The amount of charge Qo which can be stored in and discharged from a battery at nominal values of current and temperature, is in general not equal to the nominal capacity QN, which is an effect of aging.
Therefore, which is a model parameter of the modified Peukert equation, has to be updated regularly, e. g. , during the conditioning cycles. Due to the modular structure of the LabView application, more sophisticated battery models may be implemented according to individual user’s requirements. Control interfaces: eo, Apart from communication between LabView and the measuring system for data acquisition, the BMS has to have other interfaces to control external components such as the cooling system or the battery charger. At the mo634 A. Lohner, E.
Karden, R. W. De Doncker, “Charge Equalizing and Lifetime Increasing with a New Charging Method for VRLA Batteries”, 19” International Telecommunication Energy Conference, 1997, Melbourne (Australia), pp. 407-41 1 27-4  A. Lohner, “Batteriemanagement fQr verschlossene Bleibatterien am Beispiel von USV-Anlagen”, Dissertation am Institut fiir Stromrichtertechnik und Elektrische Antriebe, RWTH-Aachen, 1998  National Instruments Corporation, “LabView User Manual”, Part Number 320999A-01, 1996 Edition  R. Giglioli, A. Buonarota, P. Menga, M.
Ceraol “Charge and Discharge Fourth Order Dynamic Model of the Lead-Acid Battery”, 10* Electric Vehicle Symposium, Hongkong, 1990, pp. 371-382  W. Peukert, “Uber die Abhangigkeit der Kapazitat von der Entladestromstiirke bei Bleiakkumulatoren”, ETZ, Band 18, 1897 [SI E. Dowgiallo jr. “Innovative On-Board Instrumentation for EV Battery Characterization”, DOEEPRI Beta Battery Workshop 8,1991, pp. 215-218  H. Kahlen, B. Hauck “Batteriemanagementsysteme fiir Traktionsbatterien”, Fachtagung der Deutschen Gesellschaft fiir elektrische StraBenfahrzeuge (dges), 1995, Berlin 635