A high-efficiency battery equalization architecture with EIS measurement function

Abstract

The utility model discloses a kind of high-efficiency battery equalization architecture with EIS measurement function, the architecture includes multiple parallel battery branch, partial power processing type isolated DC / DC converter and battery management system connected in each branch. Energy equalizer input connects branch total voltage, output is connected in series with branch, for compensating the voltage deviation of 0-40V between branch, and voltage difference can be made in safety circulation range to realize energy self-balancing;Battery management system is adjusted equalizer output according to branch SOC and current state by PI closed-loop control, realizes current or SOC equalization, and supports fault branch cutout. At the same time, using the sudden change current condition generated by equalizer, combined with Morlet complex wavelet transform can measure battery electrochemical impedance spectroscopy online. The utility model energy equalization efficiency is high, branch current equalization can be realized by processing only partial power, reduce system loss and cost;Control is flexible and reliable, realize current or SOC equalization by hierarchical closed-loop control, support fault branch fast cutout.

Description

Technical Field

[0001] This utility model relates to a high-efficiency battery balancing architecture with EIS measurement function, belonging to the field of power battery system technology. Background Technology

[0002] To address the capacity and range anxiety of new energy commercial vehicles and construction machinery vehicles, multi-cluster (branch) parallel power battery systems have become a development trend. However, due to differences in manufacturing processes and operating conditions of individual batteries, the voltage of different battery branches is prone to inconsistency, which in turn generates circulating current problems between parallel branches, seriously affecting battery life and system capacity efficiency.

[0003] In existing technologies, there are two main types of solutions for suppressing circulating current: one is to connect a relay in series in the battery branch. This solution is simple to implement and has a low cost, but it requires complex switching strategies and circulating current suppression strategies, and its reliability and stability are low. The suppression effect is heavily dependent on the performance of the strategies. The other is to connect a full-power processing type isolated bidirectional DC / DC converter in parallel at the output of the battery branch. Although it can flexibly control energy balance, it is costly and bulky, and energy loss is significant, especially in 800V high-voltage systems. Utility Model Content

[0004] To overcome the shortcomings of existing multi-branch parallel battery systems, such as poor circulating current suppression, high cost, or low reliability, this invention provides a high-efficiency battery balancing architecture with EIS measurement function. By optimizing the topology design and control logic, it achieves efficient energy balancing and circulating current suppression, and also has online impedance measurement function.

[0005] This invention is achieved through the following technical solution: a high-efficiency battery balancing architecture with EIS measurement function, comprising multiple parallel battery branches, an energy equalizer connected in series with each branch, and a battery management system (BMS). The energy equalizer adopts a partial power processing type unidirectional isolated DC / DC converter, with the input being the total voltage of the series-connected batteries in the branch, and the output being connected in series with the branch, used to compensate for the voltage difference between branches to achieve current balancing; the battery management system controls the output voltage of the energy equalizer by collecting branch status signals, and can actively generate voltage differences within a safe range to achieve energy self-balancing.

[0006] The energy equalizer topology can be either two-stage or single-stage: a two-stage topology consists of a front-stage LLC resonant converter (isolation buck) and a rear-stage Buck buck converter (precise voltage regulation); a single-stage topology uses a DAB or CLLC converter, providing isolation and wide-range voltage conversion capabilities. Through optimized modulation strategies and soft-switching technology, conversion efficiency exceeding 95% can be achieved, handling less than 5% of the power from the battery system (when compensating 40V for an 800V system), with a maximum loss as low as 0.25%.

[0007] The battery management system's control logic employs a PI regulator, with the state of charge (SOC) and current of each branch serving as the inner loop feedback, and the energy equalizer's output voltage as the outer loop control quantity. Current or SOC equalization is achieved based on the system state. When a branch is detected to be fully charged, fully discharged, or experiencing a fault, the energy equalizer can disconnect the branch from the system. Furthermore, by utilizing the sudden current conditions generated by the energy equalizer, combined with the Morlet complex-valued wavelet transform algorithm, the battery's electrochemical impedance spectroscopy can be measured online, improving the accuracy of SOX estimation and fault warning in the BMS.

[0008] The energy equalizer employs a high-step-out isolated DC / DC converter topology, including a two-stage converter topology or a single-stage converter topology. In the two-stage converter topology, the front stage is an LLC resonant converter, and the rear stage is a Buck converter. The single-stage converter topology is a dual active bridge converter or a CLLC bidirectional resonant converter. The energy equalizer also works in conjunction with the battery management system to achieve online measurement of the battery's electrochemical impedance spectroscopy. Specifically, it generates a sudden current change by superimposing the output current onto the branch current. The battery management system then collects the current and voltage signals before and after the sudden change and performs wavelet transform to calculate the impedance.

[0009] The advantages of this invention are: high energy balancing efficiency, achieving branch current sharing by handling only a portion of the power, reducing system losses and costs; flexible and reliable control, achieving current or SOC balancing through hierarchical closed-loop control, supporting rapid switching out of faulty branches; rich extended functions, enabling simultaneous online measurement of electrochemical impedance spectroscopy during energy balancing, improving battery state management accuracy; and strong topology compatibility, supporting various converter topologies such as LLC+Buck, DAB, and CLLC, adapting to different application scenarios. Attached Figure Description

[0010] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0011] Figure 1 This is a schematic diagram of the multi-branch parallel battery system architecture of the series power processing equalizer of this utility model; Figure 2 This is a schematic diagram of the two-stage converter (LLC+Buck) topology of this utility model; Figure 3 This is a schematic diagram of the topology of the single-stage converter DAB converter of this utility model; Figure 4 This is a schematic diagram of the topology of the single-stage converter CLLC converter of this utility model; Figure 5 This is a control block diagram of the energy equalizer of this utility model; Figure 6 This is a schematic diagram of the voltage response to a sudden change in current according to this utility model; Figure 7 This is a flowchart of the online measurement of electrochemical impedance spectroscopy according to this utility model. Detailed Implementation

[0012] The technical solutions of the present utility model will be clearly and completely described below with reference to the accompanying drawings of the embodiments. Obviously, the described embodiments are only some embodiments of the present utility model, and not all embodiments. Based on the embodiments of the present utility model, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of the present utility model.

[0013] In the description of this utility model, it should be noted that the connection relationships indicated by terms such as "series", "parallel", and "communication connection" are only for the convenience of describing this utility model and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific connection method, and therefore should not be construed as a limitation of this utility model.

[0014] like Figure 1 The novel multi-branch parallel battery system architecture shown comprises n parallel battery branches, each with an energy equalizer connected in series. All energy equalizers are communicatively connected to the battery management system (BMS). The input to each energy equalizer is taken from the total voltage of the series-connected batteries in the branch, and its output is connected in series to its respective branch. By adjusting the output voltage, it compensates for the voltage difference between branches, thereby ensuring current balance within the battery branches. Within a safe circulating current range, the BMS can control the energy equalizer to create voltage differences between branches, achieving a self-balancing effect for the parallel branches.

[0015] The topology of the energy equalizer can be selected according to requirements: if a high step-down ratio is required and cost is a priority, then the topology of the energy equalizer can be adopted. Figure 2 The LLC+Buck two-stage topology shown uses a front-stage LLC resonant converter for isolation and initial voltage reduction, and a rear-stage Buck buck converter for precise output voltage regulation to meet the requirements of a wide voltage range. For higher integration and a simplified structure, a different approach can be used. Figure 3 and Figure 4 The DAB or CLLC single-stage topology shown achieves isolation and a wide voltage conversion range by optimizing the phase-shift modulation strategy or resonant parameters. The converter topology described above, through optimized modulation strategies and soft-switching technology, can achieve a conversion efficiency of over 95%. Since the energy equalizer only needs to compensate for voltage deviations of approximately 0-40V between battery clusters, for an 800V high-voltage system, the power it handles is only 5% (40 / 800) of the total power of the battery clusters, and the maximum loss can be as low as 0.25% (5%*(1-95%)).

[0016] The control logic of the battery management system is as follows Figure 5As shown, the SOC values ​​of each branch are collected in real time. SOC i ), current value ( i i ), calculate the average SOC value of the system ( SOC avg ) and average current value ( I avg After processing by the PI regulator, a voltage compensation command is generated. v Li ), and then converted into PWM control signals ( P ref_i The system controls the power electronic switching devices such as IGBTs and SiC in the energy equalizer. When the SOC of a branch deviates from the average value, the corresponding energy equalizer outputs a compensation voltage to form an energy transfer gradient within the safe circulating current range, thus achieving SOC equalization. When a current deviation is detected, current sharing is quickly achieved through voltage compensation. When the battery management system detects that any branch is in a fully charged, fully discharged, or faulty state, the energy equalizer can bypass that branch, cutting it out of the system.

[0017] During online electrochemical impedance spectroscopy measurements, the energy equalizer actively outputs a pulsed current superimposed on the branch, causing a sudden current surge in the branch, such as... Figure 6 As shown. The battery management system collects data at the moments of sudden current changes. t 0 Within a certain period of time before and after [ t 0 - Δt , t 0 +Δt The current i i ( t ) and voltage v BCi ( t ) changes, and according to Figure 7 The process shown is as follows: Battery mutation points are processed. t 0 The battery impedance measurement analysis time was selected, and continuous wavelet transform (CWT) was performed on the voltage and current time-domain signals based on the Morlet complex mother wavelet (CMMW) to obtain the voltage wavelet coefficients. V BCi ( a,b ) and current wavelet coefficients i i ( a,b The generalized formula for wavelet transform is shown in formula (1): (1) In the formula, f ( t ) The signal to be transformed; a The wavelet transform scaling factor; b The wavelet transform shift factor; f c The wavelet center frequency; f b These are the wavelet bandwidth coefficients; f ( a,b ) indicates that when the scale factor is a Displacement factor is b At that time, the wavelet coefficients corresponding to the signal to be transformed.

[0018] Wavelet transform scaling factor a and wavelet center frequency f c Depends on the frequency of the actual signal being analyzed f It can be calculated using the following formula: (2) Wavelet transform shift factor b Set as the moment of sudden current change t 0 Since a sudden change in current approximates a step signal, it contains rich frequency components, and the minimum frequency that wavelet transform can effectively resolve is not less than 1 / (2). Δt In this embodiment, the minimum effective frequency is set to 10 / Δt Impedance identification results below this frequency range are considered invalid.

[0019] Finally, based on voltage wavelet coefficients V BCi ( a, t 0 ) and current wavelet coefficients i i ( a, t 0 Calculate battery impedance Z ( a, t 0 The formula is as follows: (3) By adjusting the scale factor a It can calculate the impedance at different frequencies, thereby obtaining the electrochemical impedance spectrum in the required frequency range, providing a basis for state of battery (SOX) calibration and fault early warning. Example

[0020] Taking an 800V high-voltage power battery system for a new energy commercial vehicle as an example, the system consists of three parallel battery branches (n=3), each branch composed of 250 strings of lithium iron phosphate batteries. The nominal voltage of a single battery cell is 3.2V, and the total voltage of the branch is approximately 800V. The total system voltage matches the 800V high-voltage platform requirement after the three branches are connected in parallel. Each branch is connected in series with an energy equalizer. Figure 2 The two-stage topology shown is "LLC resonant converter + Buck buck converter": LLC resonant converter parameters: resonant frequency 100kHz, turns ratio 20:1, input side connected to the total branch voltage (800V), output side rectified to obtain 0~40V voltage, realizing isolation and initial voltage reduction.

[0021] Buck buck converter parameters: switching frequency 200kHz, inductance 10μH, capacitance 100μF, output voltage regulation range 0-40V, efficiency is increased to 96% through soft switching technology.

[0022] Battery Management System (BMS): It uses an STM32H743 microcontroller as the main control MCU, and integrates a 16-bit ADC module (sampling rate 1MHz) to collect branch current (accuracy ±1%) and voltage (accuracy ±0.5%). It communicates with the energy equalizer through a CAN bus (baud rate 500kbps).

[0023] II. Energy Balance Control Process Static flow sharing control After the system is powered on, the BMS collects the current of each branch in real time. i i ), calculate the average current ( I avg When the current in a certain branch is... I avg When the deviation exceeds 5A, the flow sharing regulation is activated: If the current in branch 1 is i1 = 105A I avg =100A, the BMS calculates the required compensation voltage V through a PI regulator. L1 =5V (proportional coefficient Kp=0.1, integral coefficient Ki=0.01), generate a PWM signal to control the Buck converter to output -5V, and increase the branch impedance to reduce the current.

[0024] If the current i2 in branch 2 is 95A, the BMS controls its equalizer to output 5V, reducing the branch impedance to increase the current, ultimately keeping the current deviation of the three branches within ±1A.

[0025] SOC equalization control When a branch SOC deviation exceeds 5% (e.g., branch 3 SOC=85%, SOCavg=80%), the BMS actively controls the equalizer of that branch to output a +2V compensation voltage, creating a small voltage difference (approximately 2V) between branch 3 and other branches, achieving energy transfer within a safe circulating current range (≤20A). After 30 minutes of adjustment, the SOC deviation of each branch can be reduced to less than 1%.

[0026] Faulty branch cut-out When the BMS detects that the voltage of a single cell in branch 1 exceeds 3.65V (full charge threshold), it immediately sends a cut-off command to the equalizer of that branch: controlling the LLC converter to stop working, and simultaneously driving the internal relay of the Buck converter to disconnect, isolating branch 1 from the system to avoid the risk of overcharging.

[0027] III. Practical Steps for EIS Online Measurement During the system charging process, at t=300s, the BMS controls the energy equalizer to create voltage differences between the branches. Δ t 0.25s, during which the BMS waveform recording function is triggered to record the data of the branch cell voltage and branch current from 299.75s to 300.25s. A suitable fitting algorithm is selected to fit the voltage and current data into a curve. The wavelet coefficients of the cell voltage and the branch current at the signal frequency of 40 to 1000Hz are calculated according to the wavelet transform formula. The cell impedance at different signal frequencies is calculated according to formula (3) to form an impedance spectrum.

[0028] To verify the accuracy of online EIS measurement, the second-order RC equivalent circuit model of a single cell can be used as a reference. Combined with laboratory tests, the impedance waveform of the single cell in the frequency domain is calculated based on the ohmic resistance, polarization resistance, polarization capacitance, diffusion resistance, and diffusion capacitance. The error of the online EIS measurement function proposed in this invention is then calculated at the selected frequency points.

[0029] IV. Verification of Key Parameters Efficiency test: Under the conditions of 800V system and 40V compensation voltage, the input power of the energy equalizer = 40V × 200A = 8kW (only 5% of the total system power of 160kW), the measured loss = 8kW × (1-95%) = 400W, the total system loss accounts for 0.25%, which meets the design expectations.

[0030] Reliability verification: After 1000 charge-discharge cycle tests, the equalizer switching device (IGBT model FF450R12ME4) was undamaged, the current sharing accuracy remained within ±2%, and the EIS measurement repeatability error was ≤3%.

[0031] As can be seen from the above specific implementation methods, this architecture can stably complete EIS online measurement while achieving efficient energy balance, and is suitable for high-power power battery systems such as new energy commercial vehicles and construction machinery vehicles.

Claims

1. A high-efficiency battery equalization architecture with EIS measurement function, characterized in that: It includes multiple parallel battery branches, energy equalizers connected in series in each battery branch, and a battery management system; the energy equalizer is a partial power processing type unidirectional isolated DC / DC converter, whose input is connected to the total voltage of the series-connected batteries in the battery branch, and whose output is connected in series with the battery branch; the battery management system is communicatively connected to each energy equalizer and is used to control the output voltage of the energy equalizer to compensate for the voltage difference between the battery branches.

2. The high-efficient battery equalization architecture with EIS measurement function according to claim 1, characterized in that: The energy equalizer adopts a high-step-out isolated DC / DC converter topology, including a two-stage converter topology or a single-stage converter topology; the two-stage converter topology has an LLC resonant converter in the front stage and a Buck step-down converter in the back stage; the single-stage converter topology is a dual active bridge converter or a CLLC bidirectional resonant converter.

3. The high-efficient battery equalization architecture with EIS measurement function according to claim 1, characterized in that: The battery management system has a built-in PI regulator. By collecting the SOC value and current value of each battery branch, it calculates the average SOC value and average current value of the system. Using SOC and current as the inner loop control quantity and the output voltage of the energy equalizer as the outer loop control quantity, it generates a PWM control signal and sends it to the corresponding energy equalizer.

4. The high-efficient battery equalization architecture with EIS measurement function according to claim 1, characterized in that: The battery management system is also used to disconnect a battery branch from the system via an energy equalizer when the battery branch is detected to be in a fully charged, fully discharged, or faulty state.

5. The high-efficient battery equalization architecture with EIS measurement function according to claim 1, characterized in that: The energy equalizer is also used in conjunction with the battery management system to realize online measurement of the battery electrochemical impedance spectrum. Specifically, it generates a sudden current condition by superimposing the output current onto the branch current. The battery management system collects the current and voltage signals before and after the sudden current change and performs wavelet transform to calculate the impedance.

Images