button battery cr1620

Research and implementation of a new button battery cr1620 management system for electric vehicles - system debugging and operation results

System debugging and trial operation

The debugging of the management system is divided into debugging of various functional templates, software function debugging and overall debugging of the system. After the system runs normally, the accuracy of current, voltage, temperature, etc. is calibrated. Then a period of bench testing is carried out. Finally, the system is installed on the vehicle for vehicle performance testing and trial operation. During the debugging and operation process, the following problems were encountered:

1. The system has poor anti-interference ability. When the battery is discharged with a large current or the motor on the vehicle is running, the CAN bus communication will lose data or errors.

2. The field effect tube and operational amplifier in the single battery voltage measurement circuit will be damaged due to static electricity and the influence of plugging and unplugging.

3. In the 5,000-kilometer operation experiment in July 2003, due to continuous rainy weather for several days, the environmental humidity was very high, and the system's CAN communication was completely interrupted. After in-depth inspection and testing, it was found that most of the CAN interface circuits were damaged, mainly due to the leakage of 450 volts of high voltage in the battery pack.

In response to the above problems, we mainly adopted the following methods and measures.

1. Improve the circuit board and rewire. The digital and analog parts on the circuit board should be wired separately, and finally a single point grounding should be achieved. The power supply and CAN interface parts should establish sufficient isolation areas with other circuits to prevent mutual leakage interference of the circuits. Add protection devices for field effect tubes and op amps to the single battery measurement circuit. The CAN interface chip is changed from 82C250 to the high-voltage 82C251.

2. Filtering. In the BMS-Ⅲ system, a filter must be added to the system power supply inlet. The current of the filter should not be too large, generally 2 to 3 times the load, so as to prevent external interference from impacting the system. It was also found in the experiment that a filter must be added to the output end of the charger to effectively suppress high-frequency interference, otherwise, the single battery voltage measurement will be inaccurate. A magnetic ring is placed on the outside of the battery measurement line, which also plays a certain role in suppressing interference.

3. Shielding. It was found in the experiment that if shielded wires are not used, the CAN bus can still work normally when the current is small or the motor DC/DC has no studio. Once the current increases or the motor starts to work, the CAN bus may fail. Finally, we switched to shielded wires, and the shielding of the entire CAN bus was connected together, so that the CAN bus could work properly. Electric vehicles are a strong source of interference. It is very important to use shielded wires wherever shielded wires can be used to prevent problems before they happen.

4. Improve the power supply system. In the past, the CAN bus was powered by one point, so when the air was humid, the 450-volt high-voltage leakage of the battery pack could easily break down the CAN interface circuit. On the one hand, the interface chip was changed to a high-voltage resistant device, and on the other hand, we changed the power supply system to a two-point power supply at one end and one end, which greatly improved the reliability.

5. Modify the software. In the software, we strengthened the monitoring of the error status of the CAN bus. Once the bus fails, the program will automatically reset the CAN node. Enhance the anti-interference ability from the software aspect.

8.1 Running results

8.2.1 Accuracy experiment

In the laboratory, we conducted a comprehensive test on the accuracy of the system. The conclusion is that effective anti-interference measures must be taken to achieve a single battery voltage accuracy of 15 millivolts. Figure 8.1 is a 14-channel single battery voltage accuracy test diagram.

8.2.2 Balancing experiment

The bypass shunt balancing module we developed was tested in the laboratory for 42-way balancing. The control algorithm is full-process voltage balancing control. The balancing was performed with two single cells as one unit, and the balancing accuracy was about plus or minus 0.02 volts. Figure 8.2 is the balancing experiment diagram.

8.2.3 Diagnostic experiment

In the laboratory, we used 8 lithium-ion batteries to perform expert system diagnostic experiments. Figure 8.3 is the single cell terminal voltage variation curve sampled during the discharge process of the diagnostic experiment, and Figure 8.4 is the diagnostic result given by the battery diagnostic fuzzy expert system.

8.2.4 Bench test

The system prototype was debugged and operated intermittently in the battery laboratory for nearly 3-4 months, and the on-site bench test was carried out on the whole vehicle group at the end of September 2002. The improved prototype was installed on the fuel cell platform vehicle for debugging and trial operation.

Charging and discharging tests at different rates were carried out on the bench, and the temperature, voltage, and current variation curves during the charging and discharging process were recorded. In particular, a large number of single-cell voltage change curves were collected, the consistency of the battery pack was analyzed, the static SOC was calibrated and tested, and the reliability of the system was evaluated. Figure 8.5 is the voltage change curve of the single-cell charging during the bench test, Figure 8.6 is the voltage change curve at the initial stage of 1/2C discharge, and Figure 8.7 is the voltage change curve at the later stage of 1/2C discharge.

8.2.5 Operation experiment This system was installed on a fuel cell bus, and the whole vehicle circuit debugging was completed, and the whole vehicle performance test and 1,600 kilometers of operation were carried out. Figure 8.8 is the discharge current and power curve during the vehicle driving process.

Figures 8.9 and 8.10 are the data curves of the vehicle experiment on October 25, 2003. When the car was initially driven, the road surface was relatively stable, with few slopes and small slopes, and the car ran smoothly. It can be seen from the following two figures that when the car is in a stable running state, the battery is in a small current charging state, the total voltage does not change much, and the single cell voltage is basically in a constant state.

This management system has undergone laboratory performance tests, nearly 4 months of bench tests and vehicle debugging, and 5,000 kilometers of actual operation. During the whole process, the system operated basically normally. The single-cell voltage measurement, total voltage, total current, temperature measurement, SOC estimation and other functions have met the requirements of the vehicle. The balancing method and fuzzy expert diagnosis have also been studied. The results show that this system has high reliability and practicality. The important achievements are as follows:

1. The system realizes an advanced system with distributed structure, modularization, multi-CAN communication and multi-function.

2. The measurement achieves high precision, with total current and total voltage accuracies of 0.5% and 0.2% respectively, making the power metering more accurate.

3. The unique lithium-ion battery single cell voltage measurement circuit has reached 108-126 channels, which can be expanded to more channels, with an accuracy of (0.1-0.2)%.

4. The balancing circuit and balancing algorithm of lithium-ion batteries have been studied and designed, and basic experiments have been conducted on the fuzzy diagnosis expert system of lithium-ion batteries.

5. The new SOC estimation method fully considers the influence of various factors including consistency on power estimation, adds various compensations, and improves the accuracy of power estimation.

6. The system is running on the vehicle, solving a series of problems such as automatic control of the system 24V power supply, anti-static interference, anti-motor DC/DC interference, and anti-high voltage leakage. It has passed the bench test and completed the actual operation test of the whole vehicle for 5,000 kilometers, solving a series of technical problems that have arisen, and the engineering level and reliability have been greatly improved.

At the same time, the system inevitably has some shortcomings. Regarding the next improvement of the system, there are several suggestions:

1. Considering the expansion of the diagnostic system in the future and the long-term tracking of SOC on hybrid vehicles, it is recommended to replace the CPU. Philips' 32-bit ARM series embedded microcontroller can be considered. On the basis of balancing performance and cost, it is recommended to use the 32-bit microcontroller LpC2129. LpC2129 has a very small 64-pin package, extremely low power consumption, multiple 32-bit timers, 4-way 10-bit ADC, 2-way CAN, pWM channels, 46 GpIOs and up to 9 external interrupts, making them particularly suitable for automotive, industrial control applications, medical systems and fault-tolerant maintenance buses. This can not only reduce costs, but also reduce the size of the measurement circuit board, which is of great significance for the battery management system to truly enter the market.

2. The balancing circuit has only been studied in a preliminary manner, using a simple bypass shunt method, and the control algorithm is full-process voltage balancing. What kind of balancing circuit and control algorithm can minimize energy loss, charging balancing or discharging balancing are all issues worth studying.

3. The fuzzy diagnosis expert system is still far from being truly practical. The rules used for fault diagnosis and the determination of each membership value need to be discussed in depth with battery experts, and adjusted continuously through a large number of experiments.

At present, the diagnosis of the system is mainly static or slow-changing. The dynamic diagnosis on the actual vehicle needs further research and improvement in data acquisition and SOR evaluation algorithm.

Read recommendations:

Electric vehicle lithium battery GN-24100-FAP

Lithium iron phosphate battery has these characteristics.solar energy storage battery 10kwh 48v 100a

Composition of marine power battery packs

14250 battery

button battery 2032