Design of battery capacity estimation in UPS system using internal resistance method

Abstract: This paper develops a battery capacity estimation system for uninterruptible power supply systems based on the internal resistance method. This system estimates the residual capacity of the battery by measuring the change in the internal resistance of the battery during the charging and discharging process. This method can not only quickly detect the residual capacity of the battery, but also accurately predict its value. This system only needs to capture the changes in battery voltage and current to calculate the battery internal resistance, and then know the change in battery capacitance, without knowing the current battery capacity status in advance.

This detection system is also equipped with LabVIEW graphical programming language to complete a lead-acid battery energy management monitoring system, which can not only monitor the battery voltage and current in real time, but also has a data storage function, allowing users to grasp the battery usage status in the uninterruptible power supply system at any time through the computer screen.

Keywords: uninterruptible power supply system, internal resistance detection method, monitoring system

Preface

In recent years, due to the rapid development of science and technology, the pollution generated has also increased. As environmental awareness has risen, people have realized the importance of environmental protection and have begun to use environmentally friendly and pollution-free products, especially in transportation, such as electric scooters, electric bicycles, hybrid vehicles, electric motorcycles, etc., which have become indispensable and important means of transportation. However, there are still many shortcomings in many places, such as high prices, insufficient battery life, short endurance, poor charging efficiency, etc., all of which are the reasons for the low popularity. Among the above reasons, the most important key is the energy supply of the battery.

The popularization of various electrical equipment, the large-scale use of emergency lighting equipment and battery energy storage devices, has led to a leap in the demand for batteries and their peripheral equipment. In terms of the strict requirements of various precision instruments such as computer equipment, monitoring equipment, fire equipment, medical equipment, etc. on power quality, uninterruptible power supply (UPS) has become a necessary device that can truly solve power problems. When UPS encounters power quality problems such as voltage sags, spikes, voltage surges, noise, high (low) voltage transients, which are enough to affect the normal operation of the equipment, UPS will automatically stabilize the voltage and filter out the noise, providing the equipment with a stable and clean power environment. The main equipment for UPS is still the battery.

As can be seen from the above, the power supply requirements for batteries will inevitably increase in the future. In terms of electric locomotives, how to improve the charging efficiency, prevent the battery from being overcharged or overheated, and extend the battery life is one of the key points in charger design. In an uninterruptible power system, when the city power is cut off, the user must know how much power is available in the UPS, so how to quickly and accurately know the remaining battery capacity is a very important link in the residual capacity display ^[1,2] .

The purpose of this study is to establish a system that can monitor the residual capacity of the battery with a fast charging converter. The fast charging effect is achieved by controlling the converter, and the battery voltage and current are monitored by a computer to display the residual capacity, as shown in Figure 1. The system architecture can be divided into a Flyback converter and a Buck converter; the input end is supplied with an AC voltage of 110V _rms /60Hz by the mains, which is isolated by the Flyback and then stepped down by the Buck to charge the battery end. This paper uses DSP (TMS320LF2407A) and LabVIEWDAQ (DAQPad-6016) as the control core to achieve the realization of the converter and control law.

System control planning and process

The internal resistance estimation method used in this paper is the DC measurement method. The method is that when a large amount of current flows out of the battery, the terminal voltage of the battery will drop suddenly. When the battery discharges, the terminal voltage will rise suddenly. The slope of this sudden rise and fall is closely related to the internal resistance of the battery, as shown in Figure 2 ^[7] .

The above method can be used to measure the change of the internal resistance of the battery at the beginning and end of charging and at the beginning and end of discharging.

Using the above experimental method, Figure 3 is a diagram showing the relationship between battery discharge and internal resistance. It can be observed from the figure that when the charging capacity is fuller, the internal resistance of the battery will be smaller, and when the battery capacity is less, the internal resistance of the battery will be larger.

By using the above-mentioned properties of internal resistance, it can be shown that when the internal resistance is different, the battery capacity is also different. However, because its curve is a nonlinear curve, the curve data is input into the digital processor through the equation fitting method to estimate the residual capacity, and the curve fitting method in Matlab is used to solve the corresponding curve of the battery internal resistance and discharge time.

Then, using the curve fitting function of Matlab software, the values ​​of each point in Figure 3 are substituted into the equation (1). The circle in the figure represents the actual discharge time of the battery. Substituting the curve into equation (1), it can be seen that the results of Figure 3 and Figure 4 are very similar. In the formula, R is calculated in mΩ.

Battery capacity = -0.00013678R ³ +0.019894R ² -1.2632R+30.384 (1)

Figure 5 is a calculation flow chart of the internal resistance estimation method, and its calculation core is a digital single chip. When the program of the power estimation method starts to execute, it first obtains whether the terminal voltage of the battery is greater than 10.5V, and then decides whether to start internal resistance detection. After confirming that it is correct, it detects whether the battery starts to charge or discharge. When the battery has a sudden rise or drop, the current detection circuit and the voltage detection circuit are used to feed back the current and voltage signals to the DSP, and the terminal voltage of the battery is divided by the terminal current to obtain its internal resistance value. Then, the relationship between the battery internal resistance and capacity is obtained by using formula (1), and the battery capacity is known. Then, the remaining battery capacity is determined, and then the next sudden rise or drop of the battery is waited for to determine whether the battery current has changed and then recalculate the battery residual power. This paper uses the digital-to-analog function of the DSP to represent the estimated power with LED, and then uses the data collector to record the output changes of the estimated power. After the power calculation is completed, it is determined whether the battery voltage is lower than the discharge cut-off condition of 10.5V. If it is not lower than 10.5V, the discharge will continue, and the program will continue to calculate the residual power of the battery until the battery voltage is lower than 10.5V.

Charging rules and power estimation

Different charging methods will affect the life of the battery. If you follow the instructions provided by the battery manufacturer, the charging method specified is the safest and most efficient charging method, but the charging time provided is too long. Therefore, charging methods that can charge quickly without affecting its characteristics have been proposed one after another. This article will use the multi-stage charging method to charge the battery.

The multi-stage pulse charging method uses different constant currents to charge the battery, as shown in Figure 6. When the battery is charged to the set voltage, the current is reduced and the battery is continuously charged until it reaches the set voltage again and the current is reduced again. The advantages of this method are that it has a fast charging function and avoids overcharging the battery to ensure the battery life, and it can charge the battery to saturation state ^[3] .

The purpose of battery power estimation is to detect the remaining power in the battery. You can clearly know the remaining power of the battery at any time, and understand the remaining working time of the battery. Power estimation can not only understand whether the battery is fully charged or the capacity is approaching the end, but also prevent the battery from being overcharged or over-discharged.

The method used in this paper is the internal resistance detection method. The internal resistance of a battery includes two meanings. One refers to pure ohmic resistance, and the other is generated by electrode polarization in an electrochemical reaction, as shown in Figure 7. Battery internal resistance can be divided into metal resistance (R _Metallic ) and chemical material resistance (R _{Electrochemical} ); and in the metal category, it can be divided into terminal post (R _{Terminal Post} ), metal strap (R _Strap ), terminal grid (R _Grid ), and terminal grid to paste material (R _{Grid to Paste} ); and the chemical material category includes paste material (R _Paste ), electrolyte (R _Electrolyte ), and separator (R _Separator ). Changes in the concentration of the battery's electrolyte will affect the change in the battery's internal resistance. When charging, the internal resistance will decrease with the amount of electricity, but in the discharge state, the internal resistance will increase with the amount of electricity ^[4-6] .

The internal resistance of the battery will have different data depending on the output current, battery usage, temperature and aging conditions, so the internal resistance of the battery can be used as a representative of the battery's output capacity, because the internal resistance value of the battery itself will self-correct the parameters due to the above factors. Therefore, the internal resistance measurement method is to estimate the residual power of the battery by measuring the change of the internal resistance of the battery during the charging and discharging process, as shown in the relationship curve between the internal resistance and power of the lead-acid battery in Figure 8.

However, the change in the internal resistance of the battery is very small, usually in milliohms, so the accuracy and precision of the instruments and equipment used to measure the internal resistance must be very high, and a non-interference platform is required to measure more accurate data.

Graphical interface design

LabVIEW (Laboratory Virtual Instrumentation Engineering Workbench) is commonly used for data acquisition, instrument control, and industrial automation. It uses graphical object functions to edit programs, replacing the traditional text editing method, making it easier for users to understand the meaning of program structure. In addition, the LabVIEW system also provides analog-to-digital conversion functions, but it must use a data acquisition interface card to obtain analog signals and convert them into digital signals so that ordinary computers can interpret the captured digital signals. Similarly, the signal conversion function of the interface can also be used to convert computer commands from digital signals to analog signals to drive the controlled objects, thereby achieving the purpose of signal acquisition and automatic control ^[8] .

The data acquisition interface card used in this system is NI-Pad 6016, which is a multifunctional DAQ (Data Acquisition) card with a sampling rate of 200 kS/s, 16 analog inputs, 32 digital I/Os, 2 analog outputs, 2 counters/timers, etc. The acceptable analog signal range is 10V~-10V.

Experimental Results

This article mainly constructs a fast charging system that uses the mains power supply through a flyback converter and then a step-down circuit to charge the battery using a pulse charging method. It uses the LabVIEW graphical programming language to create a real-time monitoring system. Through the computer screen, users can instantly grasp the use of lead-acid batteries. This chapter will prove the fast charging performance of the pulse charger and the battery power estimation experiment based on the aforementioned circuit architecture and control system.

Experiment 1 Multi-stage charger

FIG9 is a graph showing the change in charging current and battery terminal voltage during the entire charging process. It can be seen from the figure that the front section of the curve shows a continuous rise, and when the voltage reaches 14V, the charger control enters a multi-stage charging mode and changes the battery terminal charging current to charge the battery, as shown in the latter section of the curve in FIG9. The overall charging time is 4,283 seconds.

Experiment 2: Battery internal resistance estimation experiment

FIG10 is a graph showing the relationship between the internal resistance of the battery and the discharge current. It can be observed that no matter the battery capacity is 100% or 25%, the relationship between the internal resistance of the battery and the discharge current does not change much, and still maintains a fixed slope. FIG11 is a graph showing the relationship between the battery capacity and the internal resistance of the battery. It can be seen from the figure that when the discharge current is a large current discharge, the measured internal resistance of the battery is relatively small. When the discharge current is a small current discharge, the internal resistance of the battery will increase accordingly. From the above, it can be seen that when the battery capacity changes, the change in the internal resistance of the battery will change significantly, but the change in the discharge current will change slightly. The above-mentioned characteristics of resistance corresponding to capacity can be brought into the power estimation equation as a correction factor for the battery discharge current.

Experiment 3: Battery power estimation experiment

In the battery capacity estimation experiment in this chapter, the charger architecture described in the previous chapter is used to charge the battery in the first stage until the charging current drops to zero when the battery is fully charged, indicating that the battery is fully charged. The battery is then discharged at 1C as shown in Figure 12. The discharge time is 1,969 seconds, which is not much different from the data provided by the CSB battery manufacturer. Therefore, this capacity is defined as 100% of the maximum capacity that the battery can discharge at 1C. The internal resistance estimation method is then used to perform battery capacity estimation experiments for 75%, 50%, and 25%, respectively.

FIG13 is a graph showing the battery voltage, discharge current and power estimation curve changes when the initial capacity of the battery is 75%. The load is fixed to discharge the battery at 1C. When the terminal voltage of the battery reaches 10.5V, the discharge is judged to be completed. The total discharge time is 1,378 seconds, and the power estimation value after the discharge is completed is 2.14%.

FIG14 is a graph showing the battery voltage, discharge current and power estimation curve changes when the initial battery capacity is 50%. The total discharge time is 937 seconds, and the power estimation value is 3.4%.

FIG15 is a graph showing the battery voltage, discharge current and power estimation curve changes when the initial capacity of the battery is 25%. The total discharge time is 442 seconds, and the power estimation value after the discharge is completed is 1.92%.

The second stage uses the estimated changes in the residual capacity of the battery under various discharge currents. Before the experiment begins, to ensure that the battery is fully charged, the battery is first discharged to 10.5V at 0.1C, and then charged using the aforementioned multi-stage charger until the battery voltage reaches the set condition, indicating that it is fully charged. The battery is then left to stand for 30 minutes to wait for the electrolyte concentration in the battery to diffuse evenly, and then the discharge load is started with DC Load. During the discharge process, the changes in the battery voltage, discharge current, and estimated capacity are recorded.

In the experiment, the battery is discharged with five different loads, namely light load, heavy load, light load switching to heavy load, heavy load switching to light load, and mixed load. The internal resistance measurement method is used to estimate the residual power of the battery. The discharge is considered to be complete when the battery voltage drops to 10.5V.

FIG16 is a graph showing the battery terminal voltage curve, discharge current and power estimation curve under a light-load discharge test. The load discharges the battery at a fixed current of 3A. The total discharge time is 6,790 seconds. The power estimation value after the discharge is 1.25%.

FIG17 is a graph showing the battery terminal voltage curve, discharge current and power estimation curve under a heavy-load discharge test. The load discharges the battery at a fixed current of 9A. The total discharge time is 1,512 seconds. The power estimation value after the discharge is -2.38%. A negative estimation value means that the estimated time is longer than the actual discharge time.

FIG18 is a graph showing the battery terminal voltage curve, discharge current and power estimation curve changes under a light-load to heavy-load discharge test. The load first discharges the battery at a fixed current of 3A and then at 9A. The total discharge time is 2,728 seconds, and the power estimation value after the discharge is -2.1%.

FIG19 is a graph showing the battery terminal voltage curve, discharge current and power estimation curve changes in a heavy-load to light-load discharge test. The load first discharges the battery at a fixed current of 9A and then at 3A. The total discharge time is 4,554 seconds, and the power estimation value after the discharge is -2.3%.

FIG20 is a graph showing the battery terminal voltage curve, discharge current and power estimation curve under a mixed load discharge test. The load first discharges the battery at a fixed current of 10A, then discharges the battery at 4A, and finally discharges the battery at 9A. The total discharge time is 2,410 seconds, and the power estimation value after the discharge is -2.6%.

Experiment 4 LabVIEW graphical interface and functions

Figure 21 is the front panel of the LabVIEW monitoring system, which has real-time digital value display, over-protection warning light, data acquisition storage and residual detection functions. Figure 22 is the actual parameter measurement diagram, which shows the change of the terminal voltage and terminal current of the battery, and provides users with a display light to know whether the battery is currently charging or discharging, and a light that shows whether the battery terminal voltage is abnormal. The real-time voltage and current average data and average power are displayed at the bottom of the screen.

If the battery terminal voltage is abnormal, such as the terminal voltage is less than 10.5V, a warning signal will be generated to the user to indicate the current problem. As shown in Figure 23, it indicates the current problem status. At the same time, a set of digital I/O signals will be generated to the controller. After receiving this set of signals, the controller will command all switches of the converter to be cut off to protect the lead-acid battery. In terms of residual power display, when the battery voltage changes, the residual capacity display on the right side of the screen will display the current residual power accordingly, informing the user of the current remaining capacity of the battery. In terms of data storage, data can be stored as files of Excel, Txt or Word types. When the program starts, a file will be generated internally to store the data. The type of file stored in this article is Excel file, and the storage format is shown in Figure 24. Finally, when the user wants to stop the program, just press the STOP button on the upper left to stop the operation of the LabVIEW monitoring system.

Conclusion

This paper develops a charger system with DSP as the core, which has a fast charging circuit that improves the performance of traditional chargers and has power estimation performance. The charger has the following features: 1. In the initial charging stage, the battery is charged with a larger current to shorten the battery charging time; 2. In the later stage of charging, the battery is charged with a multi-stage constant current to prevent the battery from being overcharged and damaged.

The characteristics of the internal resistance estimation method are: 1. This system can measure the internal resistance of the battery as long as there are changes in voltage and current, and then know the change in battery capacity, without having to know whether the battery is fully charged in advance; 2. The article explains the characteristics of the change slope under different discharge current conditions, and it will have a better performance to detect the battery internal resistance value at the initial moment of discharge or the moment of discharge end, which is the advantage of this method; 3. The accuracy required by the internal resistance estimation method is slightly higher, and it can only have better performance in an interference-free environment, and it requires extremely high technical requirements in product implementation. If other types of residual detection methods are combined with the internal resistance detection method, it is believed that a more accurate residual detection system can be obtained.

The graphical interface has the following features: 1. The voltage and current of the battery can be monitored at any time, and the changes in its internal resistance can be calculated, and the residual capacity of the battery can be displayed in real time, and recorded in an Excel file for users to view at any time to know the changes in battery power; 2. The system has an alarm function. When the battery is abnormal, the system will detect the error and display the error situation; and send a signal to the digital signal processor to stop charging or discharging, and inform the user of the abnormal situation.