The most comprehensive explanation of electric vehicle BMS

Battery Management System (BMS) Overview

Table of contents

Cell characteristics

How lithium-ion batteries work

Voltage hysteresis characteristics

Capacity characteristics

state of charge

open circuit voltage

Resistance characteristics

Aging characteristics

The purpose of battery management system

BMS hardware topology

BMS electrical architecture

Function module of BMS

Signal Acquisition

SOC estimate

SOP estimate

SOH estimate

Insulation testing

Charging management

balanced management

National standard data upload

Troubleshooting

Cloud battery management

Recommended information

Battery system composition

A battery pack is generally composed of a battery module, a thermal management system, a battery management system (BMS), an electrical system and structural components. The battery module is composed of multiple cells.

The battery pack grouping methods include: first in series and then in parallel, and first in parallel and then in series.

• Batteries connected in series: The voltages are added, the capacity remains unchanged, and the internal resistance increases.
• Batteries connected in parallel: the voltage remains unchanged, the capacity is added, the internal resistance is reduced, and the power supply time is extended.

Cell characteristics

How lithium-ion batteries work

Lithium-ion batteries are composed of positive electrode materials, negative electrode materials, separators, electrolytes, and tabs.

When the battery is charged, lithium ions are released from the lithium-containing compound on the positive electrode, and the lithium ions move to the negative electrode through the electrolyte. The carbon material of the negative electrode has a layered structure with many micropores. The lithium ions that reach the negative electrode are embedded in the micropores of the carbon layer. The more lithium ions are embedded, the higher the charging capacity.

When the battery is discharged (that is, the process when we use the battery), the lithium ions embedded in the carbon layer of the negative electrode are released and move back to the positive electrode. The more lithium ions returned to the positive electrode, the higher the discharge capacity.

Voltage hysteresis characteristics

The figure below shows the battery cell charging and discharging current and voltage curves collected by the charging and discharging equipment during pulse discharge. When the current changes from -50A to 0A, the cell voltage slowly rises. Similarly, when the current jumps from 50A to 0A, the cell voltage slowly decreases. This characteristic is called the cell voltage hysteresis characteristic.

Capacity characteristics

• When the discharge rate is the same, the battery capacity decreases as the temperature decreases.
• When the temperature is the same, the battery capacity decreases as the rate increases.

state of charge

SOC refers to the battery's state of charge (state of charge), also called the remaining power, which is generally expressed as a percentage.

SOC=Q_remain/Q_max ×100%

Q_remain represents the remaining power of the battery, in Ah

Q_max represents the maximum available capacity of the battery, in Ah

• SOC is also the basic indicator for battery state estimation such as maximum charging and discharging power, internal resistance and aging.
• As can be seen from the previous introduction, the temperature and discharge rate of the battery are different, and the maximum available capacity and remaining power of the battery are also different. According to the above definition, the calculated SOC value is also different.
• In order to have consistent basic indicators for battery state estimation reference, in actual engineering applications, SOC is usually defined at 25℃@1/3C, and the SOC-OCV curve, power map and internal resistance map of the battery are calibrated based on this. etc. parameter table.

open circuit voltage

1. Open circuit voltage: The battery terminal voltage after the battery is not loaded and the hysteresis characteristic is eliminated. Generally, the battery needs to be stationary for more than 1 hour after disconnecting the load. The collected terminal voltage is called the open circuit voltage.
2. There is a one-to-one correspondence between the open circuit voltage and the battery's SOC.

Resistance characteristics

The figure below shows the relationship between battery internal resistance and battery SOC changes with temperature. The lower the temperature, the greater the internal resistance of the battery; the higher the SOC, the smaller the internal resistance of the battery

.Aging characteristics

As the number of cycles increases, the capacity of the battery gradually decreases.

The purpose of battery management system

Factors such as the difficulty of observing the status of the battery, continuous aging, and differences in consistency. After the battery is grouped, in order to ensure the safety of the system, a specialized controller BMS is required to solve these problems.

BMS hardware topology

The hardware topology architecture of BMS is divided into two types: centralized and distributed.

centralized

Concentrate all electrical components on one board. The advantage of this hardware architecture is that the circuit design is simple and the cost is low; the disadvantage is that the single sampling wire harness is relatively long and the sampling voltage drop is different, the sampling wire harness design is complex, and the number of sampling channels is limited, so it is suitable for smaller battery packs.

distributed

The distributed hardware architecture includes the main board and the slave board. The advantage of this hardware architecture is that the sampling harness distance is uniform; the disadvantage is that the cost is high and additional chips are required to send the entire information of each module to the BMS motherboard.

BMS electrical architecture

The electrical architecture of the BMS consists of a high-voltage part and a low-voltage part.

1. By collecting the temperature and voltage of individual cells from the control board, and achieving balance between batteries, the bus current is collected interactively with the current sensor.
2. Collect high-voltage values and diagnose the battery's relay closure and disconnection as well as insulation conditions.
3. Control relay closing and opening
4. Interact with other components of the vehicle through low-voltage wiring harness

Function module of BMS

BMS is an important link between vehicle power batteries and electric vehicles. The BMS collects, processes, and stores important information during the operation of the battery module in real time, and exchanges information with external devices such as vehicle controllers to ensure the safe and reliable operation of the lithium battery system.

Signal Acquisition

According to the requirements of GBT 38661-2020, the sampling accuracy of BMS needs to meet the following requirements

• Total pressure sampling

Usually, BMS uses a voltage divider circuit to measure the voltage of V1~V4 relative to V0. Used to diagnose the connectivity status of fuses and relays.

• Current sampling

There are two types of current sampling, one is the Hall sensor and the other is the shunt.

Hall sensor

When the primary current flows through the conductor, a magnetic field with a magnetic field intensity proportional to the current is generated around the conductor. The Hall element outputs a voltage signal proportional to the magnetic induction intensity at the air gap. The amplifier circuit amplifies the signal and outputs it.

shunt

The principle of the shunt is to connect a micro-ohm level resistor in series in the bus circuit, and calculate the current according to Ohm's law by measuring the voltage drop.

• Cell voltage and temperature sampling

At present, BMS generally uses mature AFE sampling chips to sample cell voltage and temperature. Commonly used sampling chip manufacturing companies include: ADI, TI and NXP, etc.

Take the ADBMS6815 produced by Analog Devices as an example. A 6815 chip has 12 cell voltage acquisition channels and 2 temperature acquisition channels. A general battery pack needs to collect 96 cell voltages, so 8 6815 chips are needed. Each 6815 chip communicates with each other through SPI, and finally transmits the collected voltage and temperature sampling values to the BMS motherboard.

As shown in the figure below, the chip contains two analog-to-digital converters (ADCs). When collecting cell voltage, the chip controls the operation of two multiplex switches respectively, and measures C1 to C0, C2 to C1,..., C12 to The voltage value of C11 is used to obtain the voltage of each cell.

The temperature sensor in the battery pack is a negative temperature coefficient thermistor (NTC). The higher the temperature, the smaller the resistance of the thermistor. Usually, BMS uses a voltage dividing circuit to collect the divided voltage of the thermistor to determine the resistance of the thermistor, thereby obtaining the temperature value.

SOC estimate

SOC estimation methods include:

• Traditional methods: ampere-hour integration method, open circuit voltage method
• Methods based on battery model: Kalman filter method, particle filter algorithm
• Neural Network Algorithm: Neural Network Algorithm

ampere-hour integration method

Algorithm principle: Starting from the definition of SOC, calculate the changing amount of electricity to estimate SOC.

Advantage

• Simple and reliable
• Batteries suitable for various chemistries

Disadvantages

• Depends on initial SOC
• Depends on current sensor accuracy

open circuit voltage method

Algorithm principle

According to the characteristic that there is a one-to-one correspondence between the open circuit voltage of the battery and the SOC of the battery, after the battery pack is stationary for 1 hour to obtain the OCV of the battery, the battery SOC is obtained by looking up the table according to the SOC-OCV curve.

Algorithm advantages

• The calculation method is simple and accurate

Algorithm disadvantages

• Depends on voltage sampling accuracy
• It is not applicable when the slope of the OCV-SOC curve is small.
• The OCV of the battery cannot be obtained in real time and cannot meet the demand for real-time calculation of battery SOC.

Kalman filter method

• Kalman filter algorithm principle

Assume that the linear system can be expressed by the following state space equation:

The state vector in the system cannot be measured directly with measuring equipment, such as the SOC of the cell. The observation vector can be measured, such as the cell voltage and temperature. The system control vector is the independent variable that causes the system state to change, such as the current. The Kalman filter algorithm estimates the state vector of the system based on the observation vector.

The iterative process of the Kalman filter algorithm is divided into two parts: time update and measurement update. For the specific derivation process, please refer to: DanSimon. Optimal state estimation: Kalman, H∞ and nonlinear filtering [M]. National Defense Industry Press, 2013.

• Battery equivalent circuit model

According to the equivalent circuit model, the following equation can be derived:

After discretization we can get:

• Disadvantages of the Kalman filter algorithm

1. High-precision battery models are needed as support;
2. The algorithm is complex, the amount of calculation is large, and it is difficult to implement the algorithm;
3. The demand for hardware computing resources is large;

Current mass production plan

Used more often: ampere-hour integral + full charge correction + OCV correction + Kalman filter

1. Ampere hour integral: calculate SOC in real time during charging and discharging process
2. Full charge correction: When the battery is fully charged according to the prescribed method, the SOC is corrected to 100%
3. OCV correction: After discharging, let it sit for a period of time. If the slope of the OCV-SOC curve is large enough, the SOC can be corrected based on OCV.
4. Kalman filter: During the charging and discharging process, the accumulated error of SOC is corrected in real time. The correction timing needs to be determined based on the accuracy of the battery model.

SOP estimate

• The goal of the SOP algorithm

Based on the current and previous charging and discharging status of the battery, estimate the maximum charging and discharging capacity of the battery, including: the maximum allowable charging power and the maximum allowable discharge power.

• SOP algorithm principle

1. According to the SOC and temperature of the battery, look up the table to determine the maximum continuous charge and discharge power and the maximum instantaneous charge and discharge power.
2. The depolarization speed of the battery cell determines the frequency of current maximum power use. When the Li ion accumulation speed on the surface of the SEI film is greater than the absorption speed of the negative electrode, the voltage will drop and the maximum power cannot be maintained.
3. Therefore, the difficulty in calculating SOP is how to exceed peak power and continuous power?

• Implementation solutions for difficulties in SOP algorithm

1. When the power is greater than the corresponding sustained power, energy integration acc_E is performed on the part greater than the sustained power.
2. When acc_E is greater than a certain threshold, the peak power cannot be maintained, and the corresponding limit power needs to be reduced according to a certain slope.
3. When the current power is less than the corresponding constant power, integration continues. When the energy of the two gradually cancels out, the limit power begins to return to the peak power according to a certain slope.
4. In the figure below, when the area of C is greater than a certain threshold, the limit power needs to be limited in the direction of continuous power. At the same time, when the area of D is roughly equal to the area of C, the limit power needs to be restored to the peak power again.

Reference factors for energy threshold selection

1. Battery polarization characteristics
2. Battery heat accumulation

SOH estimate

Principle of SOH calculation

Two-point method to calculate SOH

Determine two accurate SOC values based on the OCV-SOC curve, and calculate the accumulated charging or discharging power between the two SOCs in ampere-hours, and then calculate the battery capacity to obtain the SOH.

calendar life

According to the storage time of the battery, linear interpolation is used to check the calendar life table to obtain the calendar life SOH of the battery.

cycle life

According to the cumulative charge and discharge capacity of the battery, calculate the equivalent number of cycles, check the cycle life table with linear interpolation, and obtain the cycle life SOH of the battery.

Advantages and Disadvantages of Algorithms

1. A large amount of cell aging test data is required. The test cycle is long and the cost is high.
2. The two-point method to calculate SOC has high requirements on the accuracy of the current sensor and the accuracy of OCV measurement. The following table deduce the deviation of the two-point method to calculate the battery capacity by assuming the SOC error and the deviation of the current sensor. The calculation results show that the ΔSOC deviation is 2%, the current sensor deviation is 0.125A, and the accumulation is 20H. When ΔSOC>60%, the battery capacity deviation calculated according to this method will be less than 5%.

Insulation testing

The purpose of insulation testing is to detect the insulation resistance of the positive electrode to the casing and the negative electrode to the casing of the battery pack to prevent safety accidents caused by battery pack leakage.

The most commonly used method at present is the balanced bridge method, because it is the method recommended by GB 18384-2015 and is also called the national standard method.

principle

A standard resistor is connected in parallel between the positive and negative high-voltage busbars and the vehicle body ground. By switching the switch, the voltage dividing ratio of the positive and negative busbars to the vehicle body ground is changed to calculate the insulation resistance of the battery pack.

Balanced bridge method schematic diagram

Operating mode

Problems with the algorithm

Since there are Y capacitors (C1 and C2 in the figure below) in the entire vehicle, every time the switch is closed, the Y capacitor needs to be charged, and the voltage cannot be measured quickly, resulting in a deviation in the calculated insulation resistance, or even the inability to calculate the insulation resistance. .

Solution: Through actual vehicle calibration, after confirming that the switch is closed, wait for a certain period of time before detecting the voltage, so that the calculated insulation resistance will be relatively accurate.

Charging management

Charging methods are divided into: fast charging and slow charging

AC charging (slow charging): Use an AC car charger to charge the battery system

DC charging (fast charging): Use external DC charging piles to charge the battery system

slow charge

The slow charging process is divided into three stages:

Confirm connection

energy transfer

end outage

fast charge

The fast charging process is divided into three stages:

Initialization and data interaction phase. Charging connection confirmation, electronic lock locking, handshake, fast charging equipment self-test and parameter matching.

energy transfer stage. Information exchange between electric energy transfer and charging (charging status, changes in charging demand).

closing phase

balanced management

There are two methods of balancing within the battery pack: passive balancing and active balancing

.At the end of charging, the parallel resistor of the high-power battery is closed to maintain the constant voltage state of the high-power battery, and the remaining batteries continue to be charged with constant current until all single cells in the module reach the same voltage.

National standard data upload

Data upload path

Upload content

Troubleshooting

BMS diagnostic services need to be developed in accordance with ISO 14229-1:2013 Unified diagnostic services (UDS). UDS diagnosis includes 6 categories and 26 types of services. Each service has its own independent ID, namely SID (Service Identifier).

According to the requirements of GB/T 38661, the BMS needs to diagnose 6 basic items and 11 scalable items, for a total of 17 fault items. In addition to these basic requirements, the BMS also needs to set more faults based on the functional design of the entire vehicle and the specific needs of the battery system.

Cloud battery management

Fault detection

Use feature extraction technology to extract features from the original data.

Artificial intelligence technology is used to identify potential risk battery packs in market operations through machine learning methods such as SVM, neural network, and Bayesian classifiers.

The neural network algorithm identifies the mapping relationship between battery fault characteristics and battery failure modes to achieve battery fault identification and early warning.

Battery life assessment and prediction

• Data cleaning and feature data extraction.
• Use deep learning algorithms to automatically extract battery aging features.
• The machine learning method fits the relationship between aging characteristics and aging degree.
• Identify the battery's decay mode based on aging characteristics, and predict the battery's aging trend based on different aging modes and aging levels.

Recommended information

1. Website of the Advanced Energy Storage Science and Application Research Group of Beijing Institute of Technology : Website of the Advanced Energy Storage Science and Application Research Group

2. Core algorithm of power battery management system + Machinery Industry Press

3.GB/T 38661-2020+Technical conditions for battery management system for electric vehicles

4.GB/T 27930-2015+ Communication protocol between off-board conductive charger and battery management system for electric vehicles

5.GB/T 18487.1-2015+Electric vehicle conductive charging system+Part 1: General requirements

6.GB/T 32960.3-2016+Technical Specifications for Electric Vehicle Remote Service and Management System+Part 3: Communication Protocol and Data Format