Battery pack active equalization system and method based on differential equation model

Abstract

The invention discloses a battery pack active equalization system based on a differential equation model, and aims to solve the problem of inconsistency of single batteries in a battery pack, improve the performance of the batteries and prolong the service life of the batteries. The system comprises a battery pack, an energy conversion and distribution module, a multipath gating module, a data acquisition module and a control circuit module. The control circuit module realizes intelligent balance management based on acquired data and an innovative differential equation model. According to the equalization method, an equivalent circuit model including equivalent inductance, equivalent dynamic internal resistance, equivalent capacitance and equivalent static internal resistance is constructed for a charging loop of a battery unit. By establishing and solving a second-order differential equation describing current changes, key electrical parameters of a loop can be accurately obtained, and multi-dimensional analysis is carried out by using the parameters. And through data driving and modeling control, high-precision, high-efficiency and high-intelligence active equalization of the battery pack is realized, and the service life of the battery is remarkably prolonged.

Description

Technical Field

[0001] This invention relates to the field of electric vehicle charging technology, and more specifically to a battery pack active balancing system and method based on a differential equation model. Background Technology

[0002] Currently, electric vehicles, especially those used for delivery, can travel up to 300 kilometers per day. However, existing electric vehicles typically have a range of only 50 to 60 kilometers, can only travel for about two hours on a single charge, and take up to 8 hours to charge. This makes them unable to effectively meet the demands of high-intensity daily use, such as food delivery.

[0003] Prolonged use of electric vehicle batteries can lead to battery performance aging, and improper charging methods are a major cause of many safety accidents. Battery malfunction, prolonged charging, incompatible chargers and batteries, and charging in unsafe areas such as indoor stairwells all account for a significant proportion of accidents.

[0004] Existing shared battery swapping stations typically have basic functions, such as storing batteries in each compartment, automatic charging devices, and allowing users to swap batteries themselves by scanning a code. Smart battery swapping station management platforms usually offer cloud-based monitoring capabilities, supporting features like card / code scanning for battery swapping, automatic shutdown when fully charged, GPS / BeiDou dual positioning, real-time battery status monitoring, fault warnings, and intelligent power-off. They also provide convenient online search for nearby battery swapping stations and deposit refunds via apps or mini-programs.

[0005] While existing systems possess a certain level of battery management and monitoring capabilities, the development of high-performance shared battery swapping cabinets, especially in the context of big data management and intelligent platforms sharing battery energy, still faces challenges in terms of the core safe charging system's ability to balance and control the internal battery cells during each charge, as well as improving battery safety and lifespan. Existing balancing methods may be inefficient or unable to deeply analyze changes in internal battery parameters, thus failing to fundamentally solve battery aging and safety issues.

[0006] Therefore, it is necessary to propose a battery pack active balancing system and method based on differential equation models to solve the above problems. Summary of the Invention

[0007] The purpose of this invention is to address the problem that existing battery pack management systems lack in-depth analysis and dynamic adjustment capabilities for changes in internal battery parameters during battery equalization charging, resulting in low equalization efficiency, shortened battery life, and potential abnormal fluctuations and safety hazards. Specifically, this invention aims to provide a solution that can accurately monitor the internal state of the battery in real time, optimize equalization control through differential equation models, and predict battery life, thereby improving battery pack safety, extending its service life, and increasing equalization efficiency.

[0008] To achieve the above objectives, the present invention specifically adopts the following technical solution:

[0009] A battery pack active balancing system based on a differential equation model includes:

[0010] Battery pack: Consists of multiple battery cells connected in series;

[0011] Energy conversion and distribution module: configured to obtain energy from the battery pack and convert it into controlled electrical energy output;

[0012] Multiplexing module: Composed of multiple controllable switching transistors, configured to selectively direct the output power of the energy conversion and distribution module to any one or more battery cells to be balanced in the battery pack according to control commands;

[0013] Data acquisition module: configured to collect key operating parameters of each battery cell in the battery pack in real time, including but not limited to voltage, current and temperature;

[0014] Control circuit module: electrically connected to the energy conversion and distribution module, the multiplexing module, and the data acquisition module, configured as follows:

[0015] a. Based on the parameters collected by the data acquisition module, identify the battery cells in the battery pack that need to be balanced;

[0016] b. Based on the identification result, drive the multiplexing module to direct the controlled electrical energy to the battery cells that need to be balanced;

[0017] c. Based on the parameters acquired by the data acquisition module and the preset differential equation model, the electrical characteristics, health status and lifespan of the battery cell are analyzed, predicted and managed.

[0018] Furthermore, the energy conversion and distribution module includes a DC isolation module and a constant voltage and constant current balancing module. The DC isolation module electrically isolates the battery pack from the constant voltage and constant current balancing module through an isolation transformer.

[0019] Furthermore, the multiplexing module includes a matrix switch selection module group a and a matrix switch selection module group b, wherein the switching transistors can be MOSFETs, IGBTs, or transistors.

[0020] Furthermore, the data acquisition module includes a battery gating and voltage / current acquisition module and an analog acquisition module. The analog acquisition module is used to acquire the electrode potential and temperature of each battery cell. The system also includes a feedback module, which is used to adjust the output of the energy conversion and distribution module according to the signal from the data acquisition module.

[0021] Furthermore, the control circuit module is also configured to communicate with external devices via a wireless transceiver module to achieve remote monitoring and data transmission.

[0022] A battery pack active balancing method based on a differential equation model, comprising the aforementioned battery pack active balancing system based on a differential equation model, including the following steps:

[0023] a. Construct an equivalent circuit model: For the charging circuit of any battery cell to be balanced in the battery pack, construct an equivalent circuit model, which includes an equivalent power source, equivalent inductance, equivalent dynamic internal resistance, equivalent capacitance, equivalent static internal resistance, and battery equivalent voltage source;

[0024] b. Establish differential equations: Based on the equivalent circuit model, establish second-order differential equations describing the current change in the charging circuit, with the equivalent inductance, equivalent dynamic internal resistance, equivalent capacitance, and equivalent static internal resistance as parameters;

[0025] c. Parameter acquisition: Key parameters in the equivalent circuit model are acquired in real time or non-real time through a measurement method that combines power-on and power-off conditions. These parameters include the equivalent dynamic internal resistance, equivalent static internal resistance, equivalent capacitance, equivalent inductance, and the maximum initial current rise rate of the charging circuit switch.

[0026] d. Model-based analysis and control: Using the established differential equations and acquired parameters, the charging process of the battery cell is analyzed, diagnosed, predicted, and controlled. The analysis, diagnosis, prediction, and control include at least one of the following:

[0027] i. Electrical characteristics analysis: Evaluate the current waveform, peak value, and transient characteristics of the charging circuit;

[0028] ii. Component status assessment: Determine whether the battery cell connection points are loose, aged, or faulty;

[0029] iii. Component selection and verification: Verify whether the maximum allowable current of the switching transistor meets the requirements, and optimize the circuit inductance design to reduce impact;

[0030] iv. Resonance detection: Determine whether resonance exists in the circuit based on abnormal fluctuations in current and voltage and temperature changes;

[0031] v. Lifetime Prediction and Management: Establish a full lifetime control group database and use fitting curves and interpolation algorithms to predict the lifetime and quality grade of battery cells.

[0032] Furthermore, the parameter acquisition includes:

[0033] The current flowing through the battery cell to be balanced is calculated by measuring the current of the battery gating and voltage and current acquisition module and the current of the DC isolation module in real time.

[0034] The equivalent static internal resistance and equivalent dynamic internal resistance are calculated by measuring the instantaneous no-load voltage and current changes after the power is cut off.

[0035] The equivalent capacitance is calculated according to the formula for the discharge circuit of a resistor and capacitor; the equivalent inductance is measured using an inductance meter.

[0036] Furthermore, the equivalent static internal resistance is used to characterize the discreteness of the battery body, which can guide the classification and grouping of batteries during the manufacturing process.

[0037] The equivalent dynamic internal resistance and equivalent capacitance are used to characterize the battery manufacturing characteristics and connection point impedance. When the equivalent capacitance value deviates significantly from the normal range, it is determined that the battery connection point is loose or unreliable.

[0038] Furthermore, the verification of the switching transistor includes: using the maximum initial current rise rate of the switching transistor as an initial condition, substituting it into the second-order differential equation to calculate the maximum instantaneous current at the moment the charging current is powered on; verifying whether the maximum allowable current of all switching transistors in the circuit is greater than the calculated maximum instantaneous current, and leaving a margin; if not satisfied, adjusting the circuit design.

[0039] Furthermore, the lifespan prediction and management includes:

[0040] a. Establish a full-lifetime control group database: Based on the measurement data of battery packs and switching transistors at different charging cycles, establish full-lifetime data including the equivalent inductance, equivalent dynamic internal resistance, equivalent static internal resistance, equivalent capacitance, and the maximum initial current rise rate of the switching transistor.

[0041] b. Generate boundary data table: Use the least squares method to perform curve fitting on the control group data, and on this basis generate a full life cycle boundary data table for the high-quality boundary, the qualified boundary, and the warning boundary;

[0042] c. Pre-shipment control: Measure various parameters of the product at the time of shipment, compare them with the data of the control group, predict the product's service life and quality grade, and adjust the design parameters if the design requirements are not met;

[0043] d. Post-delivery control: During actual product operation, battery cell parameters are collected in real time and compared with the boundary data table to determine whether the battery cell status is in the high-quality zone, qualified zone, warning zone, or out-of-boundary zone; an alarm signal is issued for data in the out-of-boundary zone.

[0044] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0045] 1. This invention, by constructing an equivalent circuit for battery charging and establishing a differential equation model, can deeply analyze the electrical characteristics of the charging circuit, providing a theoretical basis for charging circuit design, thereby achieving more accurate and efficient battery equalization charging.

[0046] 2. The system of this invention can measure key parameters of the battery pack such as voltage, current, and temperature in real time, and calculate the equivalent internal resistance, equivalent capacitance, and equivalent inductance using a differential equation model. By monitoring abnormal fluctuations in these parameters, it is possible to promptly determine whether there are problems such as resonance in the circuit system, loose battery connection points, or aging, effectively preventing potential safety hazards.

[0047] 3. This invention, by calculating the maximum initial current rise rate and the maximum instantaneous current at the moment the switch is turned on, can verify whether the maximum allowable current of the selected switch meets the requirements during the circuit selection stage, and leave a margin to avoid the switch breakdown and damage, thereby improving the reliability of the system.

[0048] 4. This invention, through the classification and management of equivalent static internal resistance, groups batteries with similar resistance values ​​together, thus preventing voltage imbalances caused by differences in internal resistance from the outset. Monitoring the equivalent capacitance helps determine the quality of battery connection points, facilitating the timely detection and handling of defective products. These measures collectively extend the overall lifespan of the battery pack.

[0049] 5. This invention establishes a full-lifecycle control group database and utilizes fitting curves and interpolation algorithms to predict battery pack lifespan, determine product quality levels, and make feedback adjustments based on actual operating data. This makes battery management more intelligent and predictive. Attached Figure Description

[0050] Figure 1 This is a schematic diagram of the active balancing system principle of the present invention.

[0051] Figure 2 The equivalent circuit diagram for charging battery dc2 by the active balancing system of the present invention is shown.

[0052] Figure 3 The series inductor L0-L between the battery and the switching transistor in this invention m The schematic diagram.

[0053] Figure 4 The current i of the present invention when the capacitor C2 changes DC3 The waveform.

[0054] Figure 5 For the present invention, when the inductor L DX Changing current i DC3 The waveform.

[0055] Figure 6 For the present invention, when the inductor L DX Changing current i DC3 The waveform.

[0056] Figure 7 This is a control procedure diagram for the control group data table of the present invention.

[0057] Figure 8 This is a diagram illustrating the pre-shipment testing steps of the present invention.

[0058] Figure 9 This is a diagram showing the post-shipment operation steps of the present invention.

[0059] Figure 10 This is a schematic diagram of the dynamic resistance boundary of the present invention.

[0060] Reference numerals: 1. DC isolation module; 2. Feedback module; 3. Constant voltage and constant current equalization module; 4. Battery gating and voltage and current acquisition module; 5. Analog acquisition module; 6. Control circuit module; 7. Wireless transceiver module; 21 includes voltage feedback module; 22. Current feedback module; 23. Given adjustment feedback module. Detailed Implementation

[0061] The technical solutions of the present invention will be clearly and completely described below with reference to the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention.

[0062] Please see Figures 1 to 10An active battery equalization system based on a differential equation model includes a battery pack dc1~dcm, a DC isolation module 1, a feedback module 2, a constant voltage and constant current equalization module 3, a matrix switch selection module A group, a matrix switch selection module B group, a battery selection and voltage and current acquisition module 4, an analog acquisition module 5, a control circuit module 6, and a wireless transceiver module 7; the feedback module 2 includes a voltage feedback module 21, a current feedback module 22, and a given adjustment feedback module 23;

[0063] The battery pack DC1~DCM is connected to DC isolation module 1, analog acquisition module 5, and matrix switch selection module A.

[0064] The control circuit module 6 is connected to the DC isolation module 1, the analog acquisition module 5, the matrix switch selection module group a, the matrix switch selection module group b, the battery selection and voltage and current acquisition module 4, the feedback module 2, and the wireless transceiver module 7.

[0065] The constant voltage and constant current equalization module 3, DC isolation module 1, feedback module 2, battery selection and voltage and current acquisition module 4 are connected; the feedback module 2 is used to control the switching state of the constant voltage and constant current equalization module and the magnitude of output voltage, current and power.

[0066] The battery selection and voltage / current acquisition module 4 is also connected to the matrix switch selection module group b and the feedback module 2;

[0067] The matrix switch selection module group A is also connected to the matrix switch selection module group B.

[0068] The battery pack dc1~dcm consists of m batteries connected in series. The analog acquisition module 5 acquires the potential and current signals at each electrode connection point 0~m in the battery pack dc1~dcm, and obtains the voltage of each battery dcl~dcm based on the potential at each electrode connection point 0~m. The DC isolation module 1 obtains the voltage U from the battery pack dc1~dcm. BAT Then, the voltage U is obtained by using a transformer isolation-step-down method. DC1 The constant voltage and constant current balancing module 3 is powered by the DC isolation module 1, which has a built-in isolation transformer to isolate the electrical connection between the battery pack DC1 to DCM and the constant voltage and constant current balancing module 3. Under the control of the feedback module 2, the constant voltage and constant current balancing module 3 outputs electrical energy through the battery selection and voltage and current acquisition module 4, the matrix switch selection module b group, and the matrix switch selection module a group to charge the individual batteries with lower voltages in the battery pack DC1 to DCM until the voltages of all batteries in the battery pack reach the same level.

[0069] The analog acquisition module 5 is connected to m temperature acquisition units cw1 to cwm, which collect the temperature data of each battery in the battery pack dc1 to dcm and send it to the control circuit module 6 in real time.

[0070] The specific control circuit module 6, based on the battery voltages U1 to Um acquired by the analog acquisition module 5, identifies the battery with the lower voltage among the battery pack dc1 to dcm. The control circuit module 6 selects the corresponding switch by driving the matrix switch selection module a group, and simultaneously selects the corresponding switch by driving the matrix switch selection module b group, so that the constant voltage and constant current equalization module 3 supplies power to the battery selection and voltage and current acquisition module 4. The battery selection and voltage and current acquisition module 4 then conducts the selected switch and charges the battery with the lower voltage among the battery pack dc1 to dcm. The switches in the matrix switch selection modules a group and b group are electronic switches, usually composed of electronic components such as MOSFETs, IGBTs, or transistors.

[0071] Example 1: The analog acquisition module 5 acquires the voltage U2 of battery DC2, which is defined as the potential at electrode connection point 2 being significantly lower than the potential at connection point 1 than other batteries. Then, the control circuit module 6 drives the selection switches SK1 and SK2 via group a, and drives the selection switches SKb3 and SKb0 via group b, causing the battery selection module 4 to output voltage U. DC3 After being selected, the switching transistors SK1, SK2, SKb3 and SKb0 charge the battery dc2;

[0072] When it is necessary to charge a certain battery dc1~dcm, the corresponding switch in the matrix switch selection module a group connected to the positive and negative terminals of this battery is closed; for example, when it is necessary to charge battery dc2, the switch SK1 connected to the negative terminal connection point 1 of this battery is closed, and the switch SK2 connected to the positive terminal connection point 2 of this battery is closed.

[0073] When it is necessary to charge an odd-numbered battery string dc1, dc3, dc5, dc7..., the switch SKb1 connected to the positive terminal of the odd-numbered battery string is closed, and the switch SKb2 connected to the negative terminal of the odd-numbered battery string is closed; for example, when it is necessary to charge battery dc1, the switches SKb2 and SKb1 connected to the positive and negative electrode connection points 1 and 2 of this battery are closed.

[0074] When it is necessary to charge an even-numbered battery string dc2, dc4, dc6..., the switch SKb3 connected to the positive terminal of the even-numbered battery string is closed, and the switch SKb0 connected to the negative terminal of the even-numbered battery string is closed; for example, when it is necessary to charge battery dc2, the switches SKb3 and SKb0 connected to the positive and negative electrode connection points 1 and 2 of this battery are closed.

[0075] It also includes a wireless transceiver module 7; and a battery gating and voltage / current acquisition module 4 for acquiring the voltage U at points c1 and c0. DC3 Signal and current IDC3 The signal is received and sent to the control circuit module 6; the control circuit module 6 controls the feedback module 2 based on the voltage and current signals collected by the battery selection and voltage / current acquisition module 4; the voltage feedback module 21, current feedback module 22, and setpoint adjustment feedback module 23 included in the feedback module 2 output an analog feedback signal U after receiving the control signal. FB Adjust the output current I of the constant voltage and constant current equalization module 3 DC2 and output voltage U DC2 ;

[0076] DC isolation module 1 measures voltage U BAT Signal and current I BAT The signal is sent to the control circuit module 6;

[0077] The control circuit module 6 communicates with the wireless transceiver module 7, and the wireless transceiver module 7 communicates with external devices.

[0078] The circuit that charges the lower-voltage battery in the battery pack DC1~DCM using the selected switch can be simplified into an equivalent circuit, as follows: Figure 2 As shown;

[0079] Taking Example 1 as an example, if the voltage U2 of battery dc2 is lower than the voltage of other batteries, then the circuit can be equivalent to the equivalent circuit shown in Example 2: the circuit output by the isolation transformer in DC isolation module 1, the constant voltage and constant current equalization module 3, and the battery gating circuit of battery gating and voltage and current acquisition module 4 can be equivalent to a power supply U. DC3 The battery selection and voltage / current acquisition module outputs power to ports c1 and c0; the switching transistors SK1, SK2, SKb3, and SKb0 are equivalent to a single switching transistor K2, which is connected in series with connection point c1; the equivalent inductance L... DX This includes the equivalent switching transistor K2, the battery DC2, and the equivalent power supply U. DC3 There is stray inductance in the negative terminal and its associated circuit, as well as inductive components such as inductors and ferrite beads added in series in this circuit; an inductor L0-L is added in this circuit. m The schematic diagram is as follows Figure 3 As shown;

[0080] Equivalent inductance L DX One end is connected to the battery terminal, and the other end is connected in series with the equivalent switching transistor K2; the battery dc2 is equivalent to the equivalent dynamic internal resistance r. 21 The equivalent capacitor C2 is connected in parallel, and then the equivalent voltage source U is connected in series. 20 Then, the equivalent static internal resistance r is connected in series. 22 ;

[0081] When the voltages of dc1, dc3 to dcm are lower than the voltages of other batteries, the equivalent circuit for actively balancing and charging other batteries dc1, dc3 to dcm is similar to that in Example 2. The battery is equivalent to an equivalent dynamic internal resistance connected in parallel with an equivalent capacitor, then connected in series with an equivalent voltage source, and finally connected in series with an equivalent static internal resistance. The numbering of the selection switches in matrix switch selection modules A and B is as described above.

[0082] Based on the equivalent circuit given in Example 2, the differential equation of the equivalent circuit is established to provide a theoretical basis for the design of the charging circuit. The derivation process and control method of the differential equation are as follows:

[0083] Step 1: Measure the current I in real time using DC isolation module 1. BAT The signal, battery gating and voltage and current acquisition module 4 measures the current I in real time. DC3 The two signals are converged to the control circuit module 6, which calculates the current i2 = I of battery dc2. DC3 -I BAT ;

[0084] Step 2: Obtain the equivalent internal resistance r by measurement method 22 r 21 and voltage U 20 When the equivalent switch K2 is closed and the system's transient process has passed, the equivalent capacitor C2 in battery dc2 reaches a charging-discharging equilibrium, and the terminal voltage stabilizes. The current i2 flowing through battery dc2... (1) The voltage U2 in battery DC2 (1) It's U 20 Add capacitor voltage u C2 In addition to the equivalent internal resistance r 22 voltage u r22 Since the capacitor voltage cannot change abruptly, the instantaneous open-circuit voltage U2 after the power supply is cut off... (2) It's U 20 With the voltage across capacitor C2 applied, and the time of this instant being t2, the internal resistance r can be calculated. 22 =(U2) (1) -U2 (2) ) / I2 (1) ;Measure the no-load voltage U2 when the power supply is disconnected and the system has not undergone a transient process. (3) It's U 20 The sum of the voltage across capacitor C2 and the voltage remaining at this instant is set to time t3; the no-load voltage U2 is measured after the power supply is disconnected and the system's transient process has passed. (4) It's U 20 The voltage of capacitor C2 discharges to 0V, and the time at which this instant occurs is set as time t4. From this, the equivalent dynamic internal resistance r can be calculated. 21 =(U2) (2)-U2 (4) ) / I2 (1) According to the formula for the resistor-capacitor discharge circuit:

[0085]

[0086] The internal resistance can be calculated from this:

[0087]

[0088] Step 3: Disconnect all switches in matrix switch selection module group a and matrix switch selection module group b, and measure the equivalent inductance L using an inductance meter. DX ;

[0089] Step 4: Establish the differential equation of the equivalent circuit given in Example (2) as shown in Equation (1):

[0090]

[0091] The differential equation (1) holds when t≥0; where ε(t) is a step function, ε(t)=0 when t<0 and ε(t)=1 when t>0. Equation (1) is simplified as shown in equation (2):

[0092]

[0093] Among them, (U) DC3 -U 20 )+I BAT ·r 22 The constant is used; substituting it into the equation yields the equivalent circuit based on current i. DC3 The second-order differential equation is shown in equation (3):

[0094]

[0095] Due to the equivalent internal resistance r 22 r 21 Voltage U 20 And capacitor C2, inductor L DX The voltage U2 and current I can only be measured after the power is off. BAT i DC3 and voltage U DC3 The current i2 can be measured in real time and controlled by the control circuit module 6 based on the real-time measured current i DC3 I BAT Subtracting them, we get the current i2 = i DC3 -I BAT This requires using mathematical calculations to determine sensitive information regarding the state of the equivalent switch K2 at the moment of closure and after closure.

[0096] Control method 1: Equivalent static effective internal resistance r22 Discrete data characterizing the battery body, taking lithium batteries as an example, the magnitude of the equivalent internal resistance r22 is determined by indicators such as the consistency, uniformity, and purity of the lithium metal thickness during the lithium battery manufacturing process. In the lithium battery production process, the equivalent internal resistance r22 should be considered. 22 The resistance value is classified, and lithium batteries with similar resistance values ​​are manufactured in the same battery group to avoid problems caused by equivalent internal resistance r from the source. 22 The difference leads to an imbalance in battery voltage within the battery pack;

[0097] Control Method 2: Equivalent Dynamic Internal Resistance r 21 The equivalent capacitance C2 characterizes discrete data such as battery manufacturing features and battery connection point impedance. Taking lithium batteries as an example, the equivalent dynamic internal resistance r 21 The equivalent capacitance C2 is mainly generated by the contact resistance of the battery connection point and the battery winding. The size of the equivalent capacitance C2 is the basis for judging whether the battery connection point is loose or the connection is unreliable. For batteries with an equivalent capacitance C2 that is significantly larger than that of qualified batteries, they can be regarded as unqualified products with unqualified connection points.

[0098] When the voltage of batteries dc1, dc3 to dcm is lower than the voltage of other batteries, the calculation method of the equivalent circuit for actively balancing and charging batteries dc1, dc3 to dcm is similar to that in embodiment (2).

[0099] Circuit parameters are selected and verified using a method that calculates differential equation models;

[0100] Taking the equivalent circuit shown in Example (2) as an example, the equivalent capacitance, equivalent internal resistance and other data are calculated online for the differential equation model of the equivalent switch K2 at the moment of closing and after closing;

[0101] Method 1: Based on the calculation results of the differential equation, the effect of the change in the equivalent capacitance within the battery on the current i is determined. DC3 Conclusions regarding waveform peak values ​​during the power-on phase; given multiple capacitance values ​​within the range of the equivalent capacitance C2, with other parameters remaining constant, substitute these multiple capacitance values ​​into the differential equation to calculate the current i. DC3 The peak waveform during the power-on phase is recorded and stored in a data table; when other parameters are fixed, the current i can be measured. DC3 The peak value of the waveform during the power-on phase is estimated online by consulting a data table to determine the range of the equivalent capacitance C2 in the current circuit. The more detailed the data table, the closer the estimated value of the equivalent capacitance C2 will be to the actual value.

[0102] Construct a system where the capacitance C2 changes, and the current i DC3 The following is an example of the calculated data:

[0103] (1) The parameters of the differential equation are set as follows: voltage UDC3 =5V, voltage U 20 =4V, equivalent static internal resistance r 22 =0.06Ω, equivalent dynamic internal resistance r 21 =0.02Ω, current I BAT =10A, Inductance L DX =21μH and capacitor C2=2μF, i DC3 (0)=0A and i DC3 ′(0)=2×10 8 (A / s), substituting these data into the second-order differential equation (3), and with other parameters remaining unchanged, under the condition that the equivalent capacitance C2 = 2μF, the current i is calculated. DC3 During the power-on phase, the waveform slowly rises to a steady-state value of 22.5A; the current i DC3 Formula (4) for calculating time t:

[0104]

[0105] (2) Set the voltage U obtained from the test. DC3 =5V, voltage U 20 =4V, equivalent static internal resistance r 22 =0.06Ω, equivalent dynamic internal resistance r 21 =0.02Ω, current I BAT =10A, Inductance L DX =21μH and equivalent capacitance C2 =20μF (compared to the previous parameters, other parameters remain unchanged, only the capacitance C2 is increased by 10 times). Substituting these data into the second-order differential equation (3), under the condition that other parameters remain unchanged and the equivalent capacitance C2 =20μF, the current i is calculated. DC3 The peak waveform during the power-on phase reaches approximately 79.43 A, then gradually decreases to a steady-state value of approximately 22.5 A; the current i DC3 Formula (5) for calculating time t:

[0106]

[0107] According to the current i DC3 The current i can be calculated using formulas (4) and (5) for time t. DC3 The correspondence with time t (or waveform) is as follows: Figure 4 As shown.

[0108] (3) By comparing the calculation waveforms of formula (4) and formula (5), a data table consisting of two data points is obtained: ① When the capacitance C2 = 2μF, the current i DC3 The peak value is 22.5A; ② When the capacitor C2 = 20μF, the current i DC3 The peak value was 79.43A;

[0109] For example, when other parameters are fixed, if the measured current i DC3 The peak waveform during the power-on phase is 30A. By consulting the data table, the effective capacitance C2 in the current circuit is estimated to be between 2μF and 20μF, approximately 10μF.

[0110] By estimating the value of the equivalent capacitance C2 online in real time, the stability of the battery charging system can be monitored. If the dynamic increase of the equivalent capacitance C2 exceeds 10 times the measured value when the power is off, it can be determined that there is a connection fault in the circuit, such as loose connection, exceeding a certain temperature range, or aging of the connection point.

[0111] Method 2: Calculate the equivalent internal resistance r based on differential equations (1) and (3), as well as online real-time voltage and current measurement results and calculation data. 22 With r 21 In the steady-state charging of a battery in an active balancing system, if there is no series, parallel, or mixed resonance between capacitors and inductors in the circuit system, the equivalent internal resistance can be approximated in real time. Furthermore, the voltage and current within the circuit system are direct current, where the voltage U DC3 Current i DC3 Current i BAT Voltage U can be measured in real time. 20 It can be measured using the methods described above;

[0112] If the equivalent capacitance and equivalent inductance in the circuit system are connected in series, parallel or mixed resonance, some currents and voltages in the circuit system will fluctuate abnormally and the rate of battery temperature rise will also increase abnormally. Therefore, the abnormal fluctuation characteristics of the temperature, current and voltage of the battery being charged are used as the basis for judging that resonance has occurred in the circuit system.

[0113] Among them, when the voltage of dc1, dc3 to dcm is lower than the voltage of other batteries, the method of selecting and verifying circuit parameters by calculating the differential equation model for the circuit charging batteries dc1, dc3 to dcm is similar to that in embodiment (2).

[0114] Differential equations (1) to (3) can be used for the selection and verification of the switching transistors in group a and group b of the matrix switching transistor gating module.

[0115] Taking Example (2) as an example, the equivalent switch K2 is obtained by equivalent switching of the switches in the matrix switch selection module a group and the matrix switch selection module b group. Therefore, the equivalent switch K2 should meet the following conditions: all switches in the matrix switch selection module a group and the matrix switch selection module b group must meet the following conditions: the maximum allowable current of all switches in the matrix switch selection module a group and the matrix switch selection module b group must be greater than the maximum instantaneous operating current, and the maximum instantaneous operating current may mainly occur at the moment of power-on.

[0116] In the circuit of embodiment (2), the maximum initial current rise rate di at the instant the switch is turned on DC3 / dt(t=0+) is i DC3 ′(0), maximum initial rate of ascent i DC3 ′(0) can be found in the selection manual provided by the product manufacturer; the maximum initial current rise rate of the switching transistor. Because the rate of current rise during the conduction phase of the switching transistor is higher at the beginning and lower at the end, a coefficient k3 is used in the equation; the maximum initial rate of rise can also be obtained through testing; with i DC3 ′(0) is the initial condition. Substituting it into the differential equations (1)~(3), we can calculate i. DC3 The maximum instantaneous current at power-on; the maximum allowable current of all switching transistors in the circuit must be greater than the maximum instantaneous operating current, with a certain margin, otherwise the switching transistors will break down and be damaged; during the circuit selection stage, verify whether the maximum allowable current of the selected switching transistor is greater than the calculated current i. DC3 If the peak waveform during the power-on phase does not meet the requirements, the circuit design should be modified to ensure that the maximum allowable current of the switching transistor is greater than the calculated current i. DC3 The peak value of the waveform during the power-on phase.

[0117] For example, if the rise time of a certain MOSFET is 10ns, and k3 = 4 (which can be adjusted according to product characteristics), then... Let the voltage U obtained from the test be... DC3 =5V, voltage U 20 =4V, equivalent static internal resistance r 22 =0.06Ω, equivalent dynamic internal resistance r 21 =0.02Ω, current I BAT =10A, Inductance L DX =21μH and capacitor C2=2μF, i DC3 (0) = 0A. Substituting these data into the second-order differential equation (3), the current i is calculated. DC3 During the power-on phase, the waveform gradually approaches a steady-state value of approximately 22.5A; the calculated current i DC3 Formula (6) for calculating time t:

[0118]

[0119] If other parameters remain unchanged, and a high-speed MOSFET is selected, assuming its on-current rise time is 2ns, then Substituting this initial value into the second-order differential equation (3), the current i is calculated. DC3 The peak waveform during the power-on phase reaches approximately 79.92 A, then gradually decreases to a steady-state value of approximately 22.5 A; the calculated current i DC3 Formula (7) for calculating time t:

[0120]

[0121] According to the current i DC3 The current i can be calculated using formulas (6) and (7) for time t. DC3 The correspondence with time t (or waveform) is as follows: Figure 5 As shown in the example, this demonstrates that by replacing the original switching transistor with a high-speed switching device, the current rise time is shortened, and the calculated current i DC3 The peak value increases; therefore, it is concluded that if high-speed switching devices are selected to replace the original switching devices in the matrix switching transistor selection module a group and matrix switching transistor selection module b group circuits, the current i should first be calculated using the second-order differential equation (3). DC3 Peak value verification is performed to avoid power-on breakdown of switching devices during high-speed switching.

[0122] When the voltages of dc1, dc3 to dcm are lower than the voltages of other batteries, the method for selecting and verifying circuit parameters by calculating differential equation models for the circuit that charges batteries dc1, dc3 to dcm is similar to that in embodiment (2).

[0123] Based on the calculation results of the differential equation, the equivalent inductance L is obtained. DX Affecting current i DC3 The conclusion regarding the transient time, and the equivalent inductance L DX The smaller the value, the shorter the transient time; the equivalent inductance L DX This includes the equivalent switching transistor K2, the battery DC2, and the equivalent power supply U. DC3 There is stray inductance in the negative terminal and its associated circuit, as well as inductive components such as inductors and ferrite beads added in series in this circuit; series inductors L0 to L1 are added in this circuit. m The schematic diagram is as follows Figure 5 As shown;

[0124] Constructing an inductor L DX Under varying conditions, using the equivalent circuit shown in Example (2), the current i DC3 The following is an example of the calculated data:

[0125] The parameters of the differential equation are set as follows: voltage U DC3 =5V, voltage U 20 =4V, equivalent internal resistance r 22 =0.06Ω, equivalent dynamic internal resistance r 21 =0.02Ω, current I BAT =10A, Inductance L DX =2μH and capacitor C2=2μF, i DC3 (0)=0A and i DC3 ′(0)=2×10 8 (A / s), substituting these data into the second-order differential equation (3), the current i is calculated. DC3 During the power-on phase, the waveform gradually approaches a steady-state value of approximately 22.5A; the calculated current i DC3 Formula (8) for calculating time t:

[0126]

[0127] With other parameters remaining constant, if the value of the equivalent inductance is reduced, the equivalent inductance L... DX =0.2μH, and using this initial value, we can substitute it into the second-order differential equation (3) to calculate the current i. DC3 The peak waveform during the power-on phase reaches approximately 67.59 A, then gradually decreases to a steady-state value of approximately 22.5 A; the calculated current i DC3 Formula (9) for calculating time t:

[0128]

[0129] According to the current i DC3 The current i can be calculated using formulas (8) and (9) for time t. DC3 The correspondence with time t (or waveform) is as follows: Figure 6 As shown; this example illustrates: design the value of the equivalent inductance L DX =2μH and the value of the design equivalent inductance L DX Compared to 0.2μH, the former has a current i DC3 The current is more stable when powered on, resulting in less impact on the switching transistors in the circuit.

[0130] When the voltages of dc1, dc3 to dcm are lower than the voltages of other batteries, the method for selecting and verifying circuit parameters by calculating differential equation models for the circuit that charges batteries dc1, dc3 to dcm is similar to that in embodiment (2).

[0131] A full-lifecycle control group database was established to estimate the lifespan of the battery pack's active balancing system before delivery and to monitor the aging characteristics of the battery pack after delivery. A method of dividing the recording parameters into intervals was used to achieve dimensionality reduction storage, which reduced the data storage space and achieved an organic combination of statistics and storage. The charging algorithm is as follows:

[0132] Methods for establishing a control group database include Figure 7 As shown: First, based on the method of dividing the entire lifespan into stages according to the number of charge cycles, battery packs and switching transistor groups with 0, 200, 400, 600, 800, 1000, and 1200 charge cycles were installed in active balancing circuits, resulting in 7 sets of active balancing circuits. The usage time of each set of battery packs and switching transistor groups was evenly distributed, and a full-lifespan control group database was established using the measurement results. The number of sets used to establish the control group database can be increased or decreased according to actual conditions. Then, the values ​​of m equivalent inductors contained in each set of active balancing circuits were measured, and the average value of all values ​​was stored in the control group data table. Using a combination of power-on and power-off methods, voltage, current, and temperature signals were collected, and the dynamic and static equivalent internal resistance, equivalent capacitance, and maximum initial current rise rate di of the switching transistor during charging of each battery were calculated. DC3 / dt(t=0+), and these values ​​are stored in a data table as the test results of these parameters, which is called the full life control group database; then, the least squares method is used to select a suitable curve to approximate the expected full life curve, and a three-part boundary data table is generated based on the superimposed inherent resonant frequency fundamental noise. The three boundaries are the high-quality boundary, the qualified boundary, and the warning boundary, and the three-part boundary data table is stored; since the test data is discrete, an interpolation restoration algorithm is constructed using the selected curve equation to restore the control group three-part boundary data of the required data points, and the algorithm is written into a callable function;

[0133] Pre-set control methods before leaving the factory, such as Figure 8 As shown: First, the equivalent inductance of each battery's active balancing charging circuit is measured, and the value of each equivalent inductance is stored as a constant in the factory data sheet. Then, using a combination of power-on and power-off methods, voltage, current, and temperature signals are collected to calculate the dynamic and static equivalent internal resistance and equivalent capacitance during active balancing charging of each battery, and the equivalent internal resistance and equivalent capacitance values ​​are stored as factory values ​​in the factory data sheet. The maximum initial current rise rate di of the switching transistor in each battery's active balancing charging circuit is measured. DC3 / dt(t=0+) is stored as the factory value in the factory data table; then it is compared with the data of the system aging control group to predict the service life, determine the product quality level, and report to the quality management system; before delivery, is the predicted product lifespan met by the design requirements? If the predicted product lifespan does not meet the design requirements, the design parameters need to be adjusted until the predicted product lifespan meets the design requirements before it can be delivered.

[0134] Control methods after delivery and use, such as Figure 9 As shown: First, disconnect all switches in groups a and b of the matrix switch selection module; read the data information table, disconnect the switches, collect voltage, current, and temperature information, and determine which battery needs to start equalization charging; then, using a combination of power-on and power-off methods, collect voltage, current, and temperature signals, and calculate the dynamic and static equivalent internal resistance, equivalent capacitance, and maximum initial current rise rate di of the switch for each battery during charging. DC3 / dt(t=0+), and store the equivalent internal resistance and equivalent capacitance values ​​as the current user values ​​in the temporary data table; then read the current system's cycle charging number, and use the fitting curve interpolation method to calculate the corresponding cycle charging number data in the full-life control group data table and the three-part boundary data table; qualitatively analyze the dynamic and static equivalent internal resistance, equivalent capacitance, and maximum initial current rise rate di of the switching transistor during charging of each battery in the temporary data table. DC3 / dt (t=0+) indicates whether the circuit is in the high-quality zone, qualified zone, warning zone, or out-of-range zone. Each time the circuit is actively balanced and charged, a set of data is recorded and entered into the historical data table. For out-of-range zone data, the maximum value of the out-of-range zone data is updated and an alarm signal is sent to the external network control device. Several maximum values ​​of each parameter are also stored. If the external network control device requires this circuit to stop operating, this circuit will stop operating; otherwise, this circuit will continue to operate.

[0135] A schematic method for calculating the quality boundary: Let τ represent the number of charge cycles of a certain battery in the battery pack dc0~dcm. Batteries and switching transistors close to the quality boundary are divided into battery packs and switching transistor groups based on the number of charge cycles: 0 cycles, 200 cycles, 400 cycles, 600 cycles, 800 cycles, 1000 cycles, and 1200 cycles. These are then installed in an active balancing circuit, resulting in 7 sets of active balancing circuits. Taking the active balancing charge of the second battery as an example, the dynamic resistance r is calculated. 21 The optimal boundary is illustrated by the calculation method for aging characteristics; the dynamic resistance r is given. 21 =f(τ) is approximately an exponential function: r 21 =k2e pτ Where k and p are undetermined coefficients; for r 21 =k2e pτ Taking the logarithm of both sides, we get:

[0136] log 10 r 21 =(p·log10e)τ+log10k (10)

[0137] Let a = p·log 10 e = 0.4343p, b = log 10 If k, then expression (1) can be written as:

[0138] log 10 r 21 =aτ+b (11)

[0139] According to the calculation formula of the least squares method:

[0140]

[0141] Figure 10 The data measured after constructing the active equalization circuit using high-quality batteries and switching transistors are represented by black dots. The values ​​and parameters calculated from these values ​​are shown in Table 1.

[0142] Table 1. Measured data after constructing the active balancing circuit using high-quality batteries and switching transistors.

[0143]

[0144]

[0145] Substitute into the system of equations: We obtain: a≈0.0006128660714; b≈-2.294005357.

[0146] Substituting the result into the expression, we get: log 10 r 21 =aτ+b, we can get:

[0147] r 21 ≈e 0.00141117628τ-5.282142538 (13)

[0148] When the number of charging cycles τ = 898, substituting into the expression, we get: r 21 ≈0.01804464636. This value is the dynamic resistance r when the number of charging cycles τ = 898. 21 The high-quality boundary is obtained by calculation; the high-quality boundary diagram obtained by calculation through expression (13) is shown in the figure. Figure 10 As shown, the calculation methods for the qualified boundary and the warning boundary are similar to those for the high-quality boundary. Figure 10 The boundary for medium-quality is represented by a solid line, the boundary for qualified is represented by a dashed line, and the boundary for warning is represented by a dotted line.

[0149] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. The scope of patent protection of the present invention shall be determined by the claims. Similarly, any equivalent structural changes made based on the content of the present invention's specification shall also be included within the scope of protection of the present invention.

Claims

1. A battery pack active balancing system based on a differential equation model, characterized in that, include: Battery pack: Consists of multiple battery cells connected in series; Energy conversion and distribution module: configured to obtain energy from the battery pack and convert it into controlled electrical energy output; Multiplexing module: Composed of multiple controllable switching transistors, configured to selectively direct the output power of the energy conversion and distribution module to any one or more battery cells to be balanced in the battery pack according to control commands; Data acquisition module: configured to collect key operating parameters of each battery cell in the battery pack in real time, including but not limited to voltage, current and temperature; Control circuit module: electrically connected to the energy conversion and distribution module, the multiplexing module, and the data acquisition module, configured as follows: a. Based on the parameters collected by the data acquisition module, identify the battery cells in the battery pack that need to be balanced; b. Based on the identification result, drive the multiplexing module to direct the controlled electrical energy to the battery cells that need to be balanced; c. Based on the parameters acquired by the data acquisition module and the preset differential equation model, the electrical characteristics, health status and lifespan of the battery cell are analyzed, predicted and managed.

2. The battery pack active balancing system based on a differential equation model according to claim 1, characterized in that: The energy conversion and distribution module includes a DC isolation module and a constant voltage and constant current balancing module. The DC isolation module electrically isolates the battery pack from the constant voltage and constant current balancing module through an isolation transformer.

3. The battery pack active balancing system based on a differential equation model according to claim 1, characterized in that: The multi-channel selection module includes a matrix switch selection module group a and a matrix switch selection module group b, wherein the switching transistors can be MOSFETs, IGBTs, or triodes.

4. The battery pack active balancing system based on a differential equation model according to claim 1, characterized in that: The data acquisition module includes a battery selection and voltage / current acquisition module and an analog acquisition module. The analog acquisition module is used to acquire the electrode potential and temperature of each battery cell. The system also includes a feedback module, which is used to adjust the output of the energy conversion and distribution module according to the signal from the data acquisition module.

5. The battery pack active balancing system based on a differential equation model according to claim 1, characterized in that: The control circuit module is also configured to communicate with external devices via a wireless transceiver module to achieve remote monitoring and data transmission.

6. A battery pack active balancing method based on a differential equation model, comprising the battery pack active balancing system based on a differential equation model as described in any one of claims 2-5, characterized in that: Includes the following steps: a. Construct an equivalent circuit model: For the charging circuit of any battery cell to be balanced in the battery pack, construct an equivalent circuit model, which includes an equivalent power source, equivalent inductance, equivalent dynamic internal resistance, equivalent capacitance, equivalent static internal resistance, and battery equivalent voltage source; b. Establish differential equations: Based on the equivalent circuit model, establish second-order differential equations describing the current change in the charging circuit, with the equivalent inductance, equivalent dynamic internal resistance, equivalent capacitance, and equivalent static internal resistance as parameters; c. Parameter acquisition: Key parameters in the equivalent circuit model are acquired in real time or non-real time through a measurement method that combines power-on and power-off conditions. These parameters include the equivalent dynamic internal resistance, equivalent static internal resistance, equivalent capacitance, equivalent inductance, and the maximum initial current rise rate of the charging circuit switch. d. Model-based analysis and control: Using the established differential equations and acquired parameters, the charging process of the battery cell is analyzed, diagnosed, predicted, and controlled. The analysis, diagnosis, prediction, and control include at least one of the following: i. Electrical characteristics analysis: Evaluate the current waveform, peak value, and transient characteristics of the charging circuit; ii. Component status assessment: Determine whether the battery cell connection points are loose, aged, or faulty; iii. Component selection and verification: Verify whether the maximum allowable current of the switching transistor meets the requirements, and optimize the circuit inductance design to reduce impact; iv. Resonance detection: Determine whether resonance exists in the circuit based on abnormal fluctuations in current and voltage and temperature changes; v. Lifetime Prediction and Management: Establish a full lifetime control group database and use fitting curves and interpolation algorithms to predict the lifetime and quality grade of battery cells.

7. The active balancing method for battery packs based on a differential equation model according to claim 6, characterized in that: The parameter acquisition includes: The current flowing through the battery cell to be balanced is calculated by measuring the current of the battery gating and voltage and current acquisition module and the current of the DC isolation module in real time. The equivalent static internal resistance and equivalent dynamic internal resistance are calculated by measuring the instantaneous no-load voltage and current changes after the power is cut off. The equivalent capacitance is calculated according to the formula for the discharge circuit of a resistor and capacitor; the equivalent inductance is measured using an inductance meter.

8. The active balancing method for battery packs based on a differential equation model according to claim 6, characterized in that: The equivalent static internal resistance is used to characterize the discreteness of the battery body and can guide the classification and grouping of batteries during the manufacturing process. The equivalent dynamic internal resistance and equivalent capacitance are used to characterize the battery manufacturing characteristics and connection point impedance. When the equivalent capacitance value deviates significantly from the normal range, it is determined that the battery connection point is loose or unreliable.

9. The active balancing method for battery packs based on a differential equation model according to claim 6, characterized in that: The verification of the switching transistor includes: using the maximum initial current rise rate of the switching transistor as an initial condition, substituting it into the second-order differential equation to calculate the maximum instantaneous current at the moment the charging current is powered on; verifying whether the maximum allowable current of all switching transistors in the circuit is greater than the calculated maximum instantaneous current, and leaving a margin; if not satisfied, adjusting the circuit design.

10. The active balancing method for battery packs based on a differential equation model according to claim 6, characterized in that: The lifespan prediction and management includes: a. Establish a full-lifetime control group database: Based on the measurement data of battery packs and switching transistors at different charging cycles, establish full-lifetime data including the equivalent inductance, equivalent dynamic internal resistance, equivalent static internal resistance, equivalent capacitance, and the maximum initial current rise rate of the switching transistor. b. Generate boundary data table: Use the least squares method to perform curve fitting on the control group data, and on this basis generate a full life cycle boundary data table for the high-quality boundary, the qualified boundary, and the warning boundary; c. Pre-shipment control: Measure various parameters of the product at the time of shipment, compare them with the data of the control group, predict the product's service life and quality grade, and adjust the design parameters if the design requirements are not met; d. Post-delivery control: During actual product operation, battery cell parameters are collected in real time and compared with the boundary data table to determine whether the battery cell status is in the high-quality zone, qualified zone, warning zone, or out-of-boundary zone; an alarm signal is issued for data in the out-of-boundary zone.

Images