EV power battery pack power measurement system based on single chip microcomputer

introduction

With the rapid development of the automobile industry, the best solution to the air pollution problem caused by automobile exhaust emissions is to develop air pollution-free transportation, and electric vehicles have emerged. Electric vehicles using batteries as power sources are the most feasible both technically and economically.

The power battery used in electric vehicles is usually composed of multiple single cells connected in series or in parallel. Generally, the number of single cells in series can reach ten to dozens. The voltage of the single cell is generally 12V, the total voltage is above 100V, and the total capacity is above 100Ah. The battery performance detection system described in this article can increase or decrease the number of batteries to be detected according to the needs in terms of structural design. It has high flexibility, and the single-chip microcomputer control can achieve an accuracy of 1%. The improved open circuit voltage method is used for data calculation, which further improves the accuracy.

Lead-acid batteries have limited capacity and are corrosive, and cadmium in nickel-cadmium batteries is a polluting heavy metal. However, the capacity and charge-discharge characteristics of nickel-metal hydride batteries meet the requirements and are environmentally friendly. Therefore, nickel-metal hydride batteries are the future development direction of batteries for electric vehicles. The electric vehicle battery pack management and power metering system in this article is designed for nickel-

metal hydride batteries .

Nickel-metal hydride batteries are composed of a nickel-metal hydride positive electrode, a hydrogen storage alloy negative electrode, and an alkaline electrolyte (such as a 30% potassium hydroxide solution). The electrochemical reaction during charging and discharging is as follows:

Battery positive terminal:

Ni(OH)2+OH-→NiOOH+H2O+e

Battery negative electrode: M+H2O+e→MH+OH-

Overall battery reaction:

M+Ni(OH)2→NiOOH+MH

In the above equations, the (positive direction) forward chemical reaction direction is the chemical reaction direction when the battery is charged, and the (negative direction) reverse chemical reaction direction is the chemical reaction direction when the battery is discharged. M is a hydrogen storage alloy; MH is a hydrogen storage alloy that has absorbed hydrogen atoms.

At the beginning of constant current charging of NiMH batteries, the battery terminal voltage rises rapidly, and then drops slightly when the battery is nearly fully charged. NiMH batteries have a small internal charging resistance, so they have a higher charging efficiency. The terminal voltage of a fully charged NiMH battery drops slowly at the beginning of constant current discharge, and only when the battery is nearly fully discharged does the battery terminal voltage begin to drop sharply. During the discharge process, the internal resistance of the NiMH battery remains at a constant value with little change. Only when the discharge is nearly complete does the internal resistance of the battery increase sharply, and the time is very short, indicating that the NiMH battery has a high discharge efficiency.

System Overview

System Introduction

This system has a battery patrol detection function, which can detect the battery terminal voltage, charge and discharge current and battery temperature online during the battery charging and discharging process, and can calculate the remaining ampere-hours based on the detected current and display it as required. In addition, this system also has a fault prediction function. Its structural block diagram is shown in Figure 1.

Main technical parameters

The technical indicators that this system is intended to achieve are as follows:

(1) It is required to collect and process one data in 1s.
(2) It can detect up to 45 12V battery cell voltages, 1 current and 2 temperatures. The voltage measurement accuracy is 1%, the temperature measurement accuracy is 5%, and the current measurement accuracy is 1%.
(3) The system's operating environment temperature is 0℃~40℃.
4) The system display uses serial port communication, and the single-chip microcomputer circuit has functions such as data power-off protection and power supply detection.

Hardware structure of the system

The system consists of three major modules: main circuit module, voltage acquisition expansion module and display module.

The main circuit module is the core part of the system, which includes the signal processing and storage circuit composed of a single-chip microcomputer system, A/D converter, signal conditioning circuit, logic control circuit, power supply voltage monitoring circuit and EEPROM circuit, 20-channel voltage acquisition sub-circuit integrated in the main circuit, the main circuit module and the interface between the main circuit module and the other two modules. The specific block diagram is shown in Figure 2.

The voltage acquisition expansion module consists of 25 voltage acquisition sub-circuits and is integrated with a battery voltage input socket. The circuit board is made into a plug-in board, which can be plugged into the slot on the main circuit when expansion is needed.

The display module consists of 7 digital tube displays, three buttons and two alarm circuits.

The comparison of several commonly used voltage data acquisition circuit solutions is shown in Table 1. Since the battery powers the motor through an inverter or a chopper, the electromagnetic interference is relatively serious, so a data acquisition circuit with strong anti-interference ability should be used. The voltage data acquisition circuit solution composed of optocouplers is low in cost, small in size, meets the system requirements in accuracy, and has strong anti-interference ability. Therefore, this system adopts this voltage data acquisition circuit solution, and the principle block diagram is shown in Figure 3.

The current detection circuit uses the Hall current sensor LT208-S7 of LEM Company; the temperature acquisition circuit uses the integrated temperature sensor LM35; the sampling and holding and A/D conversion circuits are composed of the fast successive comparison 12-bit A/D converter AD1674.

System software design

This system adopts a modular programming scheme. Each module subroutine is relatively independent, and the entire system software structure is clear and easy to expand. These subroutines include: system initialization subroutine, A/D conversion subroutine, filtering and calculation subroutine, LED display subroutine and external interrupt subroutine. The flow chart of the main program is shown in Figure 4.

In response to existing interference, software methods can be used to implement digital filtering to improve signal reliability and reduce the impact of false information.

There are two methods for electricity metering algorithms in the world: (1) combining several methods of measuring open circuit voltage, load voltage, internal resistance, and electricity, and then compensating for factors such as temperature and aging, such as the internal resistance-ampere-hour method and the Peukert-ampere-hour method; (2) using more complex models, such as the combination of the TNO model, the Shepard model, and the Martin model. These two methods have their own advantages and disadvantages: the former is simple, with less calculation and relatively low hardware requirements, but the accuracy is lower; the latter is more accurate, but the method is complex, with a large amount of calculation and high hardware requirements.

The algorithm used in this system is an algorithm that organically combines the open circuit voltage method, the ampere-hour method and the Peukert equation. The open circuit voltage method is used because there is a clear relationship between the open circuit voltage and the initial power, which can be determined by actual measurement; while the ampere-hour method is more accurate in calculating the power used, and is applicable to both charging and discharging. This method has much less computational complexity than complex models, and does not require high hardware, especially the CPU, and is easy to complete in real time.

When the power is turned on, the initial power CtI: is determined based on the open circuit voltage U0, where a and b are constants. Sampling is performed once per second to obtain the voltage, current, and temperature, and the consumed power Cu: is calculated using the integral method. The initial power CtI minus Cu is the remaining power: Cr(t)=CtI-Cu(t).

This algorithm has a prerequisite, that is, the battery must have been idle for a period of time before starting up, so that the measurement result is more accurate.
The remaining power is affected by many factors, mainly the impact of discharge current on battery capacity, and the impact of temperature and number of cycles on capacity, which all need to be quantitatively compensated or corrected.

Measures to improve system reliability

(1) Software anti-interference measures. Software anti-interference is a method that sacrifices a small amount of running speed and program space to achieve the purpose of anti-interference. This system adopts software anti-interference measures such as instruction redundancy, setting software traps, and data redundancy technology.

(2) Hardware anti-interference measures. The input channel uses a photocoupler to reduce the interference caused by changes in battery voltage. At the same time, a second-order low-pass active filter is constructed in the signal processing circuit to filter out some interference waves. The microcontroller is connected to a power monitoring chip WATCHDOG, which can monitor the power supply voltage and prevent the program from running away or entering an infinite loop.

(3) Proper handling of data pointers. The address where the current data is stored (data pointer) is very important information in the system. Loss or error of the pointer will lead to data loss and misreading. In order to ensure the correctness of the pointer, this design adopts a multi-pointer and strict verification strategy: that is, multiple pointers are stored, and strict verification is performed each time the pointer is stored. When multiple write errors occur, the RAM here is blocked; before each application, several pointers are randomly read out and the correct value is determined by voting. If the vote does not form a majority, all pointers are read and voted again. If it still fails, it is displayed and the system is reinitialized.

Conclusion

This paper completely analyzes and discusses the design and implementation of the battery pack management and power metering system. The system is based on the ATMEL microcontroller and fully develops and utilizes the I/O resources of the microcontroller. The overall design of the system adopts a building block structure, which is convenient for the expansion of the number of measurement paths. This system can calculate the remaining power more accurately and provides an experimental platform for further research work, which can adapt to various calculation methods that may be used in the future.