A large-capacity commercial vehicle battery pack soh remote estimation and cloud health early warning system

Abstract

The application discloses a large-capacity commercial vehicle battery pack SOH remote estimation and cloud health early warning system, and relates to the technical field of new energy commercial vehicle battery management. The system is composed of three layers of architecture, namely, a vehicle-mounted layer, a communication layer and a cloud platform layer. The vehicle-mounted layer is provided with a BMS for uploading battery operation data. The communication layer is integrated with a wireless communication module and a charging station interface module. The cloud platform layer is provided with an SOH estimation module and an SOC calibration module. The SOH estimation module can acquire the battery operation data uploaded by the BMS through the wireless communication module, and estimate the health state SOH of the battery pack, so as to accurately identify the cell aging condition caused by temperature difference in the early stage. The SOC calibration module can fully utilize the charging standard working condition to acquire the charging pile metering data, and calibrate the state of charge SOC of the battery pack, so as to realize dynamic correction of the SOH estimation value. The application realizes online calibration and continuous optimization of the BMS capacity model parameters, and makes the SOH estimation truly reflect the aging rate of the battery pack.

Description

Technical Field

[0001] This invention relates to the technical field of battery management for new energy commercial vehicles, and more specifically, to a remote SOH estimation and cloud-based health early warning system for large-capacity commercial vehicle battery packs. Background Technology

[0002] With the rapid development of new energy commercial vehicles, heavy-duty trucks and buses commonly use large-capacity power battery packs as their driving energy source. Commercial vehicle battery packs typically consist of hundreds of individual cells connected in series and parallel, with capacities ranging from 200kWh to 500kWh, far exceeding the scale of passenger vehicle battery packs. Due to the complex operating conditions, drastic load variations, and long driving mileage of commercial vehicles, the health management of their battery packs faces more severe challenges than that of passenger vehicles.

[0003] The temperature gradient problem is particularly prominent in commercial vehicle battery packs. Due to the large size of the battery pack and the long internal heat dissipation path, the temperature difference between individual cells within the battery pack can reach 10°C to 20°C during high-current charging and discharging, far exceeding the typical 3°C to 5°C temperature difference range for passenger vehicle battery packs. Temperature has a significant impact on battery capacity, internal resistance, and aging rate. Excessive temperature differences will lead to inconsistent aging rates among individual cells, thereby accelerating the capacity degradation of the entire battery pack and increasing the risk of thermal runaway.

[0004] Regarding State of Health (SOH) estimation, existing technologies are primarily developed for small-capacity battery packs in passenger vehicles, failing to fully consider the temperature gradient characteristics of large-capacity battery packs in commercial vehicles. Existing SOH estimation algorithms based on Kalman filtering and neural networks show a significant decrease in accuracy when dealing with uneven temperature distribution within commercial vehicle battery packs. Regarding State of Charge (SOC) calibration, commercial vehicles typically undergo high-current fast charging at fixed charging stations, a process characterized by stable operating conditions and reliable data. However, existing technologies fail to fully utilize this window of opportunity for SOC benchmark calibration in conjunction with charging pile data. Several invention patents exist to address the battery health management issues in commercial vehicles, such as:

[0005] CN120352780A discloses a battery anomaly early warning system based on a smart battery gateway. It employs an improved capacitive Kalman filter algorithm and a gated recurrent unit algorithm with an attention mechanism to calculate State of Charge (SOC) and State of Harmony (SOH), constructing a multi-dimensional feature space and a Bayesian network inference engine for early warning. However, this technical solution primarily targets general-purpose batteries and is not specifically designed for the characteristics of large-capacity battery packs in commercial vehicles, such as large temperature gradients and complex series-parallel topologies, nor does it address collaborative calibration with charging stations.

[0006] CN119471405A discloses a method for online estimation of battery pack capacity and health status based on random charging information. It employs an optimization algorithm based on fuzzy Kalman filtering for temperature correction and establishes an Arrhenius model for lifespan prediction, making it applicable to new energy heavy-duty trucks. However, the system's temperature correction method is based on a single temperature parameter, lacking a temperature difference matrix model reflecting the internal temperature distribution of the battery pack, and it does not involve collaborative calibration with charging stations.

[0007] CN113960476B discloses a method and system for monitoring power battery faults based on cyber-physical fusion technology. Through the collaborative work of an onboard terminal and a cloud management system, it utilizes LSTM and BP neural networks for battery health state estimation and combines an extended Kalman filter for real-time voltage prediction. However, this solution also fails to specifically address the temperature gradient characteristics of large-capacity battery packs in commercial vehicles, nor does it involve collaborative SOC calibration with charging stations.

[0008] The existing technology has the following drawbacks:

[0009] 1) Existing SOH estimation algorithms are not specifically optimized for the large temperature gradients (temperature differences can reach 10℃ to 20℃) inside commercial vehicle battery packs. They use a single temperature parameter or average temperature for correction, which cannot accurately reflect the different impacts of uneven temperature distribution on the aging rate of each battery cell, resulting in large SOH estimation errors. This not only reduces the accuracy of the system but may also lead to distorted battery health status assessments.

[0010] 2) Existing technologies fail to fully utilize the window of opportunity for commercial vehicles to undergo high-current charging under standard operating conditions at charging stations, and fail to coordinate with charging pile data for SOC benchmark correction, resulting in the inability to effectively correct accumulated SOC errors. This approach increases the long-term accumulation of errors in SOH estimation and reduces the reliability of battery health management. Summary of the Invention

[0011] The purpose of this invention is to provide a method for remote estimation of state of health (SOH) and cloud-based health warning for large-capacity commercial vehicle battery packs. By establishing an SOH correction model based on a temperature difference matrix and a charging station-coordinated state of charge (SOC) calibration mechanism, this method solves the problems of insufficient SOH estimation accuracy and difficulty in correcting accumulated SOC errors in existing technologies for large-capacity commercial vehicle battery packs. This enables high-precision remote estimation and effective warning of the health status of large-capacity commercial vehicle battery packs, improves the level of full life cycle management of commercial vehicle batteries, reduces SOH estimation errors, and eliminates accumulated SOC errors.

[0012] The technical solution of this invention is as follows: a remote SOH estimation and cloud-based health warning system for large-capacity commercial vehicle battery packs is provided. The system architecture includes an on-board layer, a communication layer, and a cloud platform layer. The on-board layer deploys a Battery Management System (BMS), and the communication layer integrates a wireless communication module and a charging station interface module. The cloud platform layer obtains battery operation data uploaded by the BMS through the wireless communication module and estimates the SOH of the battery pack. It also obtains standard metering data of the charging pile through the charging station interface module and calibrates the SOC of the battery pack to dynamically correct the SOH estimation results.

[0013] For a single battery pack, the specific steps for SOH estimation by the cloud platform layer are as follows:

[0014] S101: Real-time acquisition of battery operation data uploaded by BMS and raw SOH data of each individual cell. The battery operation data includes the cell temperature distribution matrix of the target battery pack. Based on the cell temperature distribution matrix, a temperature difference matrix of the target battery pack is constructed.

[0015] S102, Based on the temperature difference matrix, determine whether the temperature difference of the target battery pack reaches the preset temperature difference threshold. If it does not reach the threshold, directly calculate the comprehensive SOH value of the battery pack using the original SOH data. If it does reach the threshold, generate the corresponding weight coefficient matrix based on the temperature difference matrix, perform weighted correction on the original SOH data using the weight coefficient matrix, and calculate the comprehensive SOH value of the battery pack using the corrected SOH data.

[0016] S103, determine whether the comprehensive SOH value meets the early warning conditions; if it does, generate an early warning signal.

[0017] For a single battery pack, the specific steps for SOC calibration at the cloud platform layer are as follows:

[0018] S201: After the target vehicle is connected to the charging station, the charging pile metering data uploaded by the BMS is obtained in real time. The charging pile metering data includes the charging start voltage, charging end voltage and total charging capacity of the battery pack, and the initial SOC of the battery pack is obtained based on the charging start voltage.

[0019] S202 calculates the actual SOC of the battery pack at the end of charging based on the initial SOC and total charging capacity, and calculates the deviation between the actual SOC and the real-time SOC estimate at the BMS. The deviation is used to dynamically update the maximum available capacity parameter of the battery in the battery pack SOH estimation model. At the same time, a global calibration factor is generated based on the maximum available capacity parameter of the battery before and after the update. The global calibration factor and the current weight coefficient matrix are used to perform weighted mapping update of the maximum available capacity parameter of the battery in the SOH estimation model of each individual cell, so as to realize the online calibration and optimization of the BMS capacity model parameters.

[0020] Furthermore, S101 specifically includes:

[0021] During vehicle operation, the BMS in the onboard layer acquires the temperature data of each individual battery cell in real time and summarizes it to form a cell temperature distribution matrix. This cell temperature distribution matrix, along with the currently calculated raw SOH data, is then sent to the cloud platform layer via a wireless communication module integrated in the communication layer. The SOH estimation module in the cloud platform layer receives the cell temperature distribution matrix and the raw SOH data, and constructs a temperature difference matrix for the target battery pack based on the cell temperature distribution matrix, represented as follows: , where matrix element T i -T j This represents the temperature difference between the i-th and j-th individual cells. The temperature difference matrix is... The matrix, n represents the total number of individual battery cells in the battery pack that participate in temperature monitoring.

[0022] Furthermore, S102 specifically includes:

[0023] The SOH estimation module determines whether the temperature difference of the target battery pack reaches a preset temperature difference threshold based on the temperature difference matrix. If the maximum temperature difference value in the temperature difference matrix is ​​less than or equal to the preset temperature difference threshold T... th The overall SOH value of the battery pack is then calculated directly using the original SOH data, and is expressed as:

[0024] ;

[0025] In the formula, The overall SOH value of the target battery pack. To preset weights, This represents the minimum initial SOH value among all individual cells in the battery pack. This represents the arithmetic mean of the original SOH values ​​of all individual cells in the battery pack.

[0026] Furthermore, S102 also includes:

[0027] The SOH estimation module determines whether the temperature difference of the target battery pack reaches a preset temperature difference threshold based on the temperature difference matrix. If the maximum temperature difference value in the temperature difference matrix is ​​greater than the preset temperature difference threshold T, then... th Then, a corresponding weighting coefficient matrix is ​​generated based on the temperature difference matrix, specifically including:

[0028] Based on temperature difference matrix Calculate the sum of the absolute values ​​of the temperature differences between each individual cell i and all other individual cells, expressed as:

[0029] ;

[0030] In the formula, It is the sum of the absolute values ​​of the temperature difference between the i-th individual cell and all other individual cells, which reflects the degree of alienation of the i-th individual cell in the overall heat distribution;

[0031] Will Substituting into the exponential decay function, the weighting coefficient of each individual cell is calculated and expressed as:

[0032] ;

[0033] In the formula, The weighting coefficients for the i-th individual cell are used to obtain the final weighting coefficient matrix for the target battery pack. , ,..., }

[0034] Furthermore, S102 also includes:

[0035] In obtaining the weighting coefficient matrix of each individual battery cell { , ,..., Afterwards, the original SOH data of each individual cell is weighted and normalized, and the vector corresponding to the original SOH data is defined as SOH. 0 =[ , ,..., Using the calculated weight coefficient matrix to adjust the vector SOH 0 The weighted adjustment is expressed as:

[0036] ;

[0037] In the formula, The corrected SOH value for the i-th individual cell. This is the estimated initial SOH value for the i-th individual cell.

[0038] Furthermore, in S201, the initial SOC of the battery pack is obtained based on the charging start voltage, specifically including:

[0039] First, calculate the open-circuit voltage based on the charging start voltage, charging start current, and the internal resistance of the battery pack. , is represented as:

[0040] ;

[0041] In the formula, I is the charging start voltage, and I is the charging start current. The internal resistance of the battery pack;

[0042] The initial SOC value of the battery pack is obtained by looking up the table using the open-circuit voltage through a pre-established SOC-OCV curve table. This initial SOC value is then corrected for temperature to obtain the final initial SOC, expressed as:

[0043] ;

[0044] In the formula, The initial SOC of the battery pack, Represents the reference temperature The reference SOC value is obtained by mapping the open-circuit voltage. The preset reference temperature, The actual temperature of the battery pack is measured at the current moment, and k is the temperature correction factor for the battery pack at the current voltage point.

[0045] Furthermore, S202 specifically includes:

[0046] The SOC calibration module in the cloud platform layer calculates the battery pack's true SOC at the end of charging based on the initial SOC and the total charging capacity, expressed as:

[0047] ;

[0048] In the formula, This represents the actual SOC value of the battery pack at the end of charging. The initial SOC value, Total charging capacity This is the charging efficiency coefficient. , The rated total energy storage capacity of the battery pack is specified at the factory; then the actual SOC value of the battery pack is... The data is sent to the vehicle-mounted BMS as a reference benchmark for deviation correction of the SOC calculation model, thereby triggering the BMS to smoothly correct and synchronize the current SOC estimate.

[0049] Furthermore, S202 also includes:

[0050] Calculate the deviation between the actual SOC and the real-time SOC estimate from the BMS. , = - ,in This is the real-time SOC estimate from the BMS side; using this deviation... The maximum usable capacity parameter of the battery in the dynamically updated SOH estimation model of the battery pack is expressed as:

[0051] ;

[0052] ;

[0053] ;

[0054] In the formula, This refers to the updated maximum usable battery capacity parameter in the battery pack SOH estimation model. To provide the maximum usable capacity parameter of the battery before the update in the battery pack SOH estimation model, It is the reciprocal of the capacity calculated based on the actual increase in electricity consumption. The incremental capacity is calculated based on BMS, where K is a preset gain coefficient. For symbolic functions, The nonlinear mapping representing the intensity of the deviation. As an exponential factor, By adjusting Perform strong correction for large deviations and weak correction for small deviations.

[0055] Furthermore, S202 also includes:

[0056] A global calibration factor is generated based on the battery's maximum available capacity parameters before and after the update, expressed as: This global calibration factor A correction ratio used to reflect the overall capacity of the battery pack; utilizing a global calibration factor. Using the currently obtained weight coefficient matrix, the maximum usable capacity parameter of the battery in the SOH estimation model of each individual cell is updated by weighted mapping to achieve parameter refinement, as shown below:

[0057] ;

[0058] In the formula, This represents the maximum usable capacity of the i-th individual cell after the update. This represents the maximum usable capacity of the i-th individual cell before the update. is the weighting coefficient for the i-th individual cell.

[0059] Furthermore, S103 specifically includes:

[0060] The health early warning module in the cloud platform layer receives the comprehensive SOH value output by the SOH estimation module and compares the comprehensive SOH value with the first preset warning threshold and the second preset warning threshold respectively. If the comprehensive SOH value is greater than or equal to the first preset warning threshold, the warning condition is not met. If the comprehensive SOH value is less than the first preset warning threshold but greater than or equal to the second preset warning threshold, a first-level warning signal is generated. If the comprehensive SOH value is less than the second preset warning threshold, a second-level warning signal is generated.

[0061] The beneficial effects of this invention are:

[0062] First, the technical solution of this invention constructs a distributed architecture consisting of an on-board layer, a communication layer, and a cloud platform layer. The cloud platform layer is equipped with a State of Health (SOH) estimation module and an early health warning module, which can solve the industry technical problem of insufficient SOH estimation accuracy for large-capacity battery packs in commercial vehicles, and generate graded warning signals based on the SOH estimation results. During vehicle operation, the system collects temperature data of each individual cell in the battery pack in real time, constructs a temperature difference matrix characterizing the non-uniformity of internal heat distribution, and calculates the dispersion of each individual cell in the overall heat distribution of the pack by introducing an adaptive weighting mechanism. An adaptive weighting coefficient is dynamically generated for each cell using an exponential decay function. Subsequently, the original SOH estimation value of each individual cell is corrected by weighted normalization using the weighting coefficient matrix. The technical solution of this invention can accurately identify cell aging caused by temperature differences at an early stage, thereby calculating the comprehensive SOH value. This effectively eliminates the estimation errors caused by traditional methods using a single temperature or average temperature correction, enabling the SOH estimation to truly reflect the differentiated aging rate of each cell within the battery pack due to uneven temperature distribution, significantly improving the accuracy and robustness of judging the true upper limit of the battery pack's lifespan.

[0063] Secondly, based on the cell-level SOH differentiation correction, the technical solution in this invention sets up a SOC calibration module. The SOC calibration module can synchronously acquire high-precision metering data of the charging pile under standard charging conditions, calculate the true SOC during the charging process, and calculate the deviation between the true SOC and the predicted SOC at the BMS end. Then, a nonlinear damping algorithm based on deviation feedback is introduced, and the maximum available capacity parameter in the battery pack SOH model is smoothly and dynamically updated using this deviation and nonlinear feedback compensation terms (including gain coefficient and exponential factor). This effectively eliminates the cumulative error caused by long-term operation of the ampere-hour integration method, avoids model parameter oscillation caused by charging sample fluctuations, and realizes online calibration and continuous optimization of BMS capacity model parameters. Furthermore, this invention also uses the maximum available capacity parameters of the battery pack before and after the update to generate a global calibration factor. Combined with the weight coefficient matrix constructed in the early stage, the maximum available capacity parameters in each individual cell model are updated by weighted mapping, forming a closed-loop calibration mechanism from global observation to local refinement. This not only accurately corrects the current SOC value, but also effectively refines the global capacity correction to the individual cell level, improving the accuracy of range prediction for commercial vehicles throughout their entire life cycle. Attached Figure Description

[0064] The advantages of the above and additional aspects of the present invention will become apparent and readily understood in the description of the embodiments in conjunction with the following drawings, wherein:

[0065] Figure 1This is a schematic diagram of the overall structure of a large-capacity commercial vehicle battery pack SOH remote estimation and cloud-based health early warning system according to an embodiment of the present invention;

[0066] Figure 2 This is a schematic flowchart illustrating SOH estimation at the cloud platform layer according to an embodiment of the present invention;

[0067] Figure 3 This is a schematic flowchart illustrating SOC calibration at the cloud platform layer according to an embodiment of the present invention;

[0068] Figure 4 This is a schematic diagram of the complete data flow from vehicle data collection and cloud processing to health warning output and cross-carrier data sharing, according to an embodiment of the present invention. Detailed Implementation

[0069] To better understand the above-mentioned objectives, features, and advantages of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments. It should be noted that, unless otherwise specified, the embodiments of the present invention and the features thereof can be combined with each other.

[0070] In the following description, many specific details are set forth in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and therefore the scope of protection of the invention is not limited to the specific embodiments disclosed below.

[0071] like Figure 1 As shown, this embodiment provides a remote SOH estimation and cloud-based health warning system for large-capacity commercial vehicle battery packs. The system includes: an on-board layer, a communication layer, and a cloud platform layer. The on-board layer deploys a Battery Management System (BMS) to collect and upload battery operating data. The communication layer integrates a wireless communication module and a charging station interface module for remote data transmission. The cloud platform layer obtains the battery operating data uploaded by the BMS through the wireless communication module and estimates the SOH of the battery pack. It also obtains charging pile metering data through the charging station interface module and calibrates the SOC of the battery pack to achieve dynamic correction of the estimated SOH value.

[0072] Specifically, the vehicle-mounted layer (deployed on the target vehicle) includes a heavy-duty commercial vehicle BMS, a multi-cell temperature acquisition module, a current acquisition module, and a voltage acquisition module. The heavy-duty commercial vehicle BMS connects to these modules via a CAN bus for data transmission. The heavy-duty commercial vehicle BMS manages the charging and discharging process of each battery pack within the battery pack and collects operational data for each battery pack (the battery pack as a whole corresponds to the battery's operational data). For a single battery pack, the operational data includes the temperature of each cell, the overall charging and discharging current of the battery pack, the overall charging and discharging voltage of the battery pack, and the charging and discharging voltage of each cell. The multi-cell temperature acquisition module can use a temperature sensor array to collect the temperature of each cell within each battery pack (each temperature sensor collects the temperature of one cell). The current acquisition module can use a current sensor array to collect the charging and discharging current of each battery pack. The voltage acquisition module can use a voltage sensor array to collect the charging and discharging voltage of each battery pack and the charging and discharging voltage of each cell within a single battery pack.

[0073] The communication layer (located on the target vehicle) integrates a wireless communication module (4G / 5G) and a charging station interface module. The wireless communication module is used to upload the target vehicle's battery operation data to the cloud platform layer; the charging station interface module is used to establish a communication connection with the charging pile and obtain the charging pile's metering data during the charging process.

[0074] In this embodiment, the wireless communication module can adopt an industrial-grade 5G / LTE-Cat.1 full-network compatible wireless communication module, which can realize high-speed, low-latency bidirectional data transmission between the vehicle and the cloud platform, ensuring extremely high connection stability and anti-interference capability even in complex electromagnetic environments (such as high-power charging areas). The vehicle-side charging station interface module can adopt a commercial vehicle-mounted DC charging socket conforming to the GB / T 20234.3 standard (physically integrated in the charging compartment at the rear or side of the vehicle, serving as the physical connection medium between the vehicle and the external charging station). After the target vehicle and the charging pile establish a physical connection through the vehicle-mounted DC charging socket, the vehicle-mounted BMS communicates with the charging pile. The BMS receives the charging message (including charging pile metering data) sent by the charging pile through the CAN bus, and aggregates the charging pile metering data with the local battery operation data (battery voltage, current, and temperature data) and sends them together to the cloud platform layer through the wireless communication module, thereby realizing accurate tracking and remote monitoring of the entire charging process.

[0075] The cloud platform layer (deployed on the host computer) includes a SOH estimation module, a SOC calibration module, an early health warning module, a cross-carrier data sharing module, and a database storage module. The SOH estimation module performs SOH estimation based on battery operating data. An embedded temperature difference compensation module calculates the temperature difference matrix based on the battery operating data and corrects the estimated SOH value. The SOC calibration module performs SOC correction in conjunction with battery operating data and charging pile metering data under charging conditions. The early health warning module generates tiered warning signals based on the SOH estimation results to provide early warnings of battery health status. The cross-carrier data sharing module enables battery health data exchange between different carrier platforms. The database storage module stores battery operating data, charging pile metering data, and SOH estimation results.

[0076] For a single battery pack, the specific implementation steps for SOH estimation at the cloud platform layer are as follows:

[0077] S101: Real-time acquisition of battery operation data uploaded by BMS and raw SOH data of each individual cell. The battery operation data includes the cell temperature distribution matrix of the target battery pack. Based on the cell temperature distribution matrix, a temperature difference matrix of the target battery pack is constructed.

[0078] Specifically, during vehicle operation, the BMS in the vehicle layer acquires the temperature data of each individual cell in real time through a temperature sensor array pre-placed on the target battery pack, and summarizes it to form a cell temperature distribution matrix. This cell temperature distribution matrix, along with the currently calculated raw SOH data, is then sent to the cloud platform layer via a wireless communication module integrated in the communication layer. The SOH estimation module in the cloud platform layer receives the cell temperature distribution matrix and the raw SOH data, and constructs a temperature difference matrix for the target battery pack based on the cell temperature distribution matrix, represented as follows: , where matrix element T i -T j This represents the temperature difference between the i-th and j-th individual cells. The temperature difference matrix is... The matrix, n represents the total number of individual battery cells in the battery pack that participate in temperature monitoring.

[0079] The raw SOH data is calculated in real time by the BMS itself. In this embodiment, the SOH estimation model in the BMS is defined as the ratio of the current maximum usable capacity of the battery to the rated capacity. For a single cell, the SOH calculation formula is expressed as:

[0080] ;

[0081] In the formula, This refers to the SOH value of a single battery cell. This represents the current actual maximum usable capacity of a single battery cell (i.e., maximum discharge capacity, unit: Ah; this parameter is one that needs to be corrected in the future). This refers to the rated capacity (i.e., nominal capacity) of a single battery cell.

[0082] In this embodiment, the vehicle operation process can be a charging condition, a discharging condition, and a stationary condition. The charging condition is the physical process in which the vehicle is parked and the battery pack is replenished with energy through external power supply equipment (such as a DC fast charging pile or an AC slow charging pile). The discharging condition is the physical process in which the vehicle is in a driving or auxiliary power consumption state and the battery pack outputs electrical energy through the power system to drive the motor or load. The stationary condition is the physical process in which the vehicle is not in a charging or discharging state and the battery pack is in the open circuit voltage (OCV) stable period, which is used for system charge balancing or stationary calibration testing.

[0083] S102, based on the temperature difference matrix, determine whether the temperature difference of the target battery pack reaches the preset temperature difference threshold. If it does not reach the threshold, directly calculate the comprehensive SOH value of the battery pack using the original SOH data. If it does reach the threshold, generate the corresponding weight coefficient matrix based on the temperature difference matrix, use the weight coefficient matrix to perform weighted correction on the original SOH data, and use the corrected SOH data to calculate the comprehensive SOH value of the battery pack.

[0084] Specifically, the SOH estimation module determines whether the temperature difference of the target battery pack reaches a preset temperature difference threshold based on the temperature difference matrix. If the maximum temperature difference value in the temperature difference matrix is ​​less than or equal to the preset temperature difference threshold T, the module will determine the SOH. th The overall SOH value of the battery pack is then calculated directly using the original SOH data, and is expressed as:

[0085] ;

[0086] In the formula, The overall SOH value of the target battery pack. In this embodiment, the weights are preset. It can be set to 0.6. This represents the minimum initial SOH value among all individual cells in the battery pack. This represents the arithmetic mean of the original SOH values ​​of all individual cells in the battery pack.

[0087] Conversely, if the maximum temperature difference in the temperature difference matrix is ​​greater than the preset temperature difference threshold T... th Then, a corresponding weighting coefficient matrix is ​​generated based on the temperature difference matrix, specifically including:

[0088] Based on temperature difference matrix Calculate the sum of the absolute values ​​of the temperature differences between each individual cell i and all other individual cells, expressed as:

[0089] ;

[0090] In the formula, This is the sum of the absolute values ​​of the temperature differences between the i-th individual cell and all other individual cells, reflecting the degree of alienation of the i-th individual cell in the overall thermal distribution; Substituting into the exponential decay function, the weighting coefficient of each individual cell is calculated and expressed as:

[0091] ;

[0092] In the formula, The weighting coefficients for the i-th individual cell are used to obtain the final weighting coefficient matrix for the target battery pack. , ,..., }

[0093] In obtaining the weighting coefficient matrix of each individual battery cell { , ,..., Afterwards, the system performs weighted normalization on the original SOH data of each individual cell. This process aims to reflect the differentiated aging degree of each individual cell after being affected by temperature through weighting coefficients, transforming discrete temperature characteristics into corrected SOH values; the vector corresponding to the original SOH data is defined as SOH. 0 =[ , ,..., Using the calculated weight coefficient matrix to adjust the vector SOH 0 The weighted adjustment is expressed as:

[0094] ;

[0095] In the formula, The corrected SOH value for the i-th individual cell. The original SOH estimate for the i-th cell; .

[0096] After correction, the overall SOH value of the battery pack is calculated using the corrected SOH data, and is expressed as follows: The calculation method was modified so that the maximum temperature difference is less than or equal to the preset temperature difference threshold T. th The method for calculating the overall SOH value is the same under the same circumstances, and will not be repeated here.

[0097] This embodiment further improves the temperature compensation logic by constructing a temperature difference matrix and calculating the dispersion index D of each individual cell relative to the entire group of cells. iThis indicator reflects the degree of isolation of a specific cell in the thermal distribution field by accumulating the absolute temperature difference between various temperature measurement points. Then, it uses this dispersion index to generate adaptive weighting coefficients, which are then used to weight the cell SOH data, thus improving the accuracy of battery state of health (SOH) estimation. Compared to the traditional mean-deviation method, this method can more accurately identify extreme regions in the thermal distribution, effectively highlighting the impact of these easily aging cells on the system when calculating the final SOH. This more accurately reflects the true upper limit of the battery pack's lifespan, more effectively identifies and highlights the bottleneck effect, and improves the sensitivity and robustness of health warnings.

[0098] S103, determine whether the overall SOH value meets the warning conditions. If it does, generate a corresponding graded warning signal based on the overall SOH value and use the graded warning signal to provide a warning of battery pack aging.

[0099] Specifically, the health early warning module in the cloud platform layer receives the comprehensive SOH value output by the SOH estimation module and compares the comprehensive SOH value with the first preset warning threshold and the second preset warning threshold respectively. If the comprehensive SOH value is greater than or equal to the first preset warning threshold, the warning condition is not met. If the comprehensive SOH value is less than the first preset warning threshold but greater than or equal to the second preset warning threshold, the warning condition is met, and a first-level warning signal is generated to trigger the first-level warning. If the comprehensive SOH value is less than the second preset warning threshold, the warning condition is met, and a second-level warning signal is generated to trigger the second-level warning.

[0100] In this embodiment, preferably, the first preset warning threshold can be set to 78%-82%, and the second preset warning threshold can be set to 68%-72%.

[0101] For a single battery pack, the specific implementation steps for SOC calibration at the cloud platform layer are as follows:

[0102] S201 After the target vehicle connects to the charging station, the BMS communicates with the charging pile. The cloud platform layer obtains the charging pile metering data uploaded by the BMS in real time. The charging pile metering data includes the charging start voltage, charging end voltage and total charging capacity of the battery pack, and obtains the initial SOC of the battery pack based on the charging start voltage.

[0103] Specifically, after the target vehicle connects to the charging station, the charging station interface module and the charging pile are physically connected. The vehicle-mounted BMS communicates with the charging pile through a handshake. The charging pile sends periodic message data to the BMS in real time, including charging voltage, charging current, and cumulative charging power increment. The BMS, as the central node, uploads the received message data to the cloud platform layer in real time. The SOC calibration module in the cloud platform layer receives the message data and extracts the charging start voltage at the moment charging begins. Then, it obtains the initial SOC of the battery pack based on the charging start voltage. When the SOC calibration module receives the charging end message (CEM), it locks and extracts the charging end voltage and the cumulative charging power increment recorded during charging. Finally, it obtains the charging pile metering data, including the charging start voltage, charging end voltage, and total charging power, where the total charging power is the sum of the cumulative charging power increments.

[0104] It should be noted that after the physical connection between the charging station interface module and the charging pile, the BMS and the charging pile will establish a handshake communication. During the charging process, the BMS and the charging pile will exchange information in real time. The charging pile will continuously send information such as the current voltage, current, and cumulative charging amount (energy increment) to the vehicle BMS through messages (such as BHM, BCS, etc.). As a charging controller, the BMS needs this data to determine whether current reduction is needed, whether the cutoff voltage has been reached, or whether the stopping conditions are met. Although there is real-time data during the process, the final cumulative energy and termination voltage at the end of charging are confirmed in the charging end message (such as the CEM message). This process relies on existing onboard hardware, and the principle will not be elaborated here.

[0105] The initial state of charge (SOC) of the battery pack is obtained based on the charging start voltage, specifically including:

[0106] First, calculate the open-circuit voltage based on the charging start voltage, the charging start current (obtained at the instant charging begins), and the internal resistance of the battery pack. , is represented as:

[0107] ;

[0108] In the formula, I is the charging start voltage, and I is the charging start current. Given the internal resistance of the battery pack (known data); by using a pre-established SOC-OCV curve table and open-circuit voltage for lookup mapping, the original SOC value of the battery pack is obtained. Then, temperature correction is applied to this original SOC value to obtain the final initial SOC, expressed as:

[0109] ;

[0110] In the formula, The initial SOC of the battery pack, Represents the reference temperature The reference SOC value is obtained by mapping the open-circuit voltage. This is the preset reference temperature (e.g., 25℃). The actual temperature of the battery pack measured at the current moment (the average temperature of all individual cells) is given by k, which is the temperature correction factor of the battery pack at the current voltage point (open circuit voltage) (unit: % / ℃).

[0111] It should be noted that the SOC-OCV curve table in this embodiment can be obtained in advance through low-rate charge-discharge tests in the laboratory; the internal resistance of the battery pack is the data of a pre-constructed function model based on pulse power characteristic (HPPC) testing. This model uses battery temperature and state of charge as independent variables, and in actual calculations, the corresponding values ​​are obtained by combining table lookup with bilinear interpolation to achieve dynamic voltage drop compensation under different operating conditions; the temperature correction coefficient is obtained by fitting the charge-discharge cycle test data of the battery under different ambient temperatures; these testing or calculation methods are all commonly used methods in the field and will not be described in detail here.

[0112] In step S202, the cloud platform layer calculates the actual SOC of the battery pack at the end of charging based on the initial SOC and total charging capacity obtained in step S3, and calculates the deviation between the actual SOC and the real-time SOC estimate from the BMS. This deviation is used to dynamically update the maximum available capacity parameter of the battery in the battery pack SOH estimation model. At the same time, a global calibration factor is generated based on the maximum available capacity parameter before and after the update (this factor reflects the correction ratio of the overall battery pack capacity). The global calibration factor and weight coefficient matrix are used to perform weighted mapping and update of the maximum available capacity parameter of the battery in the SOH estimation model of each individual cell, so as to realize the online calibration and optimization of the BMS capacity model parameters and improve the accuracy of subsequent range prediction.

[0113] Specifically, the SOC calibration module in the cloud platform layer calculates the battery pack's true SOC at the end of charging based on the initial SOC and the total charging capacity, expressed as:

[0114] ;

[0115] In the formula, The true SOC value of the battery pack at the end of charging (dimensionless, usually expressed as a percentage). The initial SOC value, Total charging capacity (the cumulative increase in charging capacity recorded by the charging station within the charging range, in kWh). This is the charging efficiency coefficient. , The rated total energy storage capacity (in kWh) of the battery pack is specified at the factory; then the actual SOC value of the battery pack is... The data is sent to the vehicle-mounted BMS as a reference benchmark for deviation correction of the SOC calculation model, thereby triggering the BMS to smoothly correct and synchronize the current SOC estimate and eliminate the cumulative error of the ampere-hour integral.

[0116] Calculate the deviation between the actual SOC and the real-time SOC estimate from the BMS. , = - ,in This is the real-time SOC estimate from the BMS, calculated automatically by the BMS using the ampere-hour integration method at the end of charging; this deviation is used... The maximum usable capacity parameter of the battery in the dynamically updated SOH estimation model of the battery pack is expressed as:

[0117] ;

[0118] ;

[0119] ;

[0120] In the formula, This refers to the updated maximum usable battery capacity parameter in the battery pack SOH estimation model. To provide the maximum usable capacity parameter of the battery before the update in the battery pack SOH estimation model, The item represents the actual effective energy input during this charge (kWh); It is the reciprocal of the capacity calculated based on the actual increase in electricity consumption. The difference between the two is the reciprocal of the capacity calculated based on the BMS estimation increment, which is the capacity correction amount caused by the deviation; K is a preset gain coefficient, which is a calibration intensity adjustment parameter used to control the sensitivity of each charge to capacity update (for example, K=0.1 can avoid the system instability caused by excessive capacity update at one time). For a sign function, when If the value is greater than 0 (indicating that the BMS has underestimated the capacity), the function returns +1, increasing the capacity. If the value is less than 0 (indicating that the BMS has overestimated), the function returns -1, reducing the capacity. The nonlinear mapping representing the intensity of the deviation. As an exponential factor, By adjusting It can achieve strong correction for large deviations and weak correction for small deviations, effectively suppressing random noise interference within the system.

[0121] It should be noted that, to avoid system oscillations (i.e., drastic capacity jumps with each charge) caused by directly using deviation values ​​for correction, this application, when correcting the maximum available capacity parameter in the battery pack SOH estimation model, introduces deviation compensation (i.e., based on the increase in charge) in addition to using basic correction (i.e., correction based on the increase in charge). The damping feedback corresponds to the sign function part in the formula (the sign of the deviation directly drives the direction of capacity correction), by introducing the gain coefficient K and the exponential factor. It damped the capacity update amount, effectively avoiding model parameter oscillations caused by anomalies in a single charging sample, and had adaptive smoothing capabilities, thus improving the robustness of range prediction.

[0122] Simultaneously, a global calibration factor is generated based on the maximum usable battery capacity parameters before and after the update, expressed as: This global calibration factor It can reflect the correction ratio of the overall battery pack capacity; it utilizes a global calibration factor. Compared with the currently obtained weight coefficient matrix (i.e., the weight coefficient matrix calculated by S102 and recorded in real time in the system) , ,..., The maximum usable capacity parameter of the battery in the SOH estimation model of each individual cell is updated by weighted mapping to achieve parameter refinement, as shown below:

[0123] ;

[0124] In the formula, This represents the maximum usable capacity of the i-th individual cell after the update. This represents the maximum usable capacity of the i-th individual cell before the update. is the weighting coefficient for the i-th individual cell.

[0125] In this embodiment, the above steps complete a closed-loop process from global observation (charging pile) to model update (overall capacity) and then to local refinement (individual cell parameters). The weighted coefficient matrix system not only performs overall calibration of the battery pack but also uses historical temperature distribution data to differentiate the degradation characteristics of each cell, enabling online calibration and optimization of the BMS capacity model parameters and improving the accuracy of subsequent range predictions.

[0126] The cloud platform layer also enables cross-carrier data sharing. Specifically, the cloud platform layer includes a cross-carrier data sharing module and a database storage module. The database storage module is used for structured storage and indexing of the target vehicle's full lifecycle data. The cross-carrier data sharing module can obtain vehicle information and battery pack status information. Its data acquisition logic is configured as follows: it obtains basic vehicle information from the vehicle terminal through the vehicle gateway protocol (such as the GB / T 32960 protocol) and synchronizes the battery pack status information in real time through the BMS. The vehicle information includes the vehicle identification code encrypted with the SHA256 algorithm. The battery pack status information includes the battery pack model configuration (such as lithium iron phosphate 300S1P), SOH evolution trend data within the predetermined period, battery pack operating temperature distribution statistics (including key statistical indicators such as the average cell temperature, variance, and maximum temperature difference, such as the average 33.5℃, variance 52.3℃², and maximum temperature difference 17℃), and abnormal event records (such as historical temperature over-limit warnings). The cross-carrier data sharing module can also encapsulate and standardize the collected vehicle information and battery pack status information. Specifically, it performs unified formatting on the collected vehicle information and battery pack status information, encapsulates them into standardized data packets that conform to industry standard protocols to support cross-platform compatibility, and performs full data synchronization with the authorized third-party operator cloud platform through the industry unified data exchange platform at a fixed time every day (such as 02:00 am) using REST API calls.

[0127] After receiving data, other operators' cloud platforms perform format and integrity verification on the data packets. Once verified, the data is parsed and incorporated into the platform's battery health database. Other operators can then use this cross-operator shared data to conduct industry benchmark comparative analysis, such as comparing the average SOH levels and temperature management effectiveness of different operators' fleets, supporting improvements in industry-level battery health management.

[0128] The steps in this invention can be adjusted, combined, or deleted according to actual needs.

[0129] The units in the system of this invention can be merged, divided, or deleted according to actual needs.

[0130] In this invention, the terms "installation," "connection," "linking," and "fixing" should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; "linking" can be a direct connection or an indirect connection through an intermediate medium. Those skilled in the art can understand the specific meaning of these terms in this invention according to the specific circumstances.

[0131] The shapes of the components in the accompanying drawings are schematic and may differ from their actual shapes. The drawings are only used to illustrate the principles of the present invention and are not intended to limit the present invention.

[0132] Although the invention has been disclosed in detail with reference to the accompanying drawings, it should be understood that these descriptions are merely exemplary and not intended to limit the application of the invention. The scope of protection of the invention is defined by the appended claims and may include various variations, modifications, and equivalents made to the invention without departing from the scope and spirit of the invention.

Claims

1. A remote SOH estimation and cloud-based health early warning system for large-capacity commercial vehicle battery packs, characterized in that, The system architecture includes an in-vehicle layer, a communication layer, and a cloud platform layer. The in-vehicle layer deploys a Battery Management System (BMS), and the communication layer integrates a wireless communication module and a charging station interface module. The cloud platform layer obtains battery operation data uploaded by the BMS through the wireless communication module and estimates the State of Harmony (SOH) of the battery pack. It also obtains standard metering data of the charging pile through the charging station interface module and calibrates the State of Charge (SOC) of the battery pack to dynamically correct the SOH estimation results. For a single battery pack, the specific steps for SOH estimation by the cloud platform layer are as follows: S101: Real-time acquisition of battery operation data uploaded by BMS and raw SOH data of each individual cell. The battery operation data includes the cell temperature distribution matrix of the target battery pack. Based on the cell temperature distribution matrix, a temperature difference matrix of the target battery pack is constructed. S102, Based on the temperature difference matrix, determine whether the temperature difference of the target battery pack reaches the preset temperature difference threshold. If it does not reach the threshold, directly calculate the comprehensive SOH value of the battery pack using the original SOH data. If it does reach the threshold, generate the corresponding weight coefficient matrix based on the temperature difference matrix, perform weighted correction on the original SOH data using the weight coefficient matrix, and calculate the comprehensive SOH value of the battery pack using the corrected SOH data. S103, determine whether the comprehensive SOH value meets the early warning conditions; if it does, generate an early warning signal. For a single battery pack, the specific steps for SOC calibration at the cloud platform layer are as follows: S201: After the target vehicle is connected to the charging station, the charging pile metering data uploaded by the BMS is obtained in real time. The charging pile metering data includes the charging start voltage, charging end voltage and total charging capacity of the battery pack, and the initial SOC of the battery pack is obtained based on the charging start voltage. S202 calculates the actual SOC of the battery pack at the end of charging based on the initial SOC and total charging capacity, and calculates the deviation between the actual SOC and the real-time SOC estimate at the BMS. The deviation is used to dynamically update the maximum available capacity parameter of the battery in the battery pack SOH estimation model. At the same time, a global calibration factor is generated based on the maximum available capacity parameter of the battery before and after the update. The global calibration factor and the current weight coefficient matrix are used to perform weighted mapping update of the maximum available capacity parameter of the battery in the SOH estimation model of each individual cell, so as to realize the online calibration and optimization of the BMS capacity model parameters.

2. The large-capacity commercial vehicle battery pack SOH remote estimation and cloud-based health early warning system as described in claim 1, characterized in that, S101 specifically includes: During vehicle operation, the BMS in the onboard layer acquires the temperature data of each individual battery cell in real time and summarizes it to form a cell temperature distribution matrix. This cell temperature distribution matrix, along with the currently calculated raw SOH data, is then sent to the cloud platform layer via a wireless communication module integrated in the communication layer. The SOH estimation module in the cloud platform layer receives the cell temperature distribution matrix and the raw SOH data, and constructs a temperature difference matrix for the target battery pack based on the cell temperature distribution matrix, represented as follows: , where matrix element T i -T j This represents the temperature difference between the i-th and j-th individual cells. The temperature difference matrix is... The matrix, n represents the total number of individual battery cells in the battery pack that participate in temperature monitoring.

3. The large-capacity commercial vehicle battery pack SOH remote estimation and cloud-based health early warning system as described in claim 2, characterized in that, Specifically, S102 includes: The SOH estimation module determines whether the temperature difference of the target battery pack reaches a preset temperature difference threshold based on the temperature difference matrix. If the maximum temperature difference value in the temperature difference matrix is ​​less than or equal to the preset temperature difference threshold T... th The overall SOH value of the battery pack is then calculated directly using the original SOH data, and is expressed as: ； In the formula, The overall SOH value of the target battery pack. To preset weights, This represents the minimum initial SOH value among all individual cells in the battery pack. This represents the arithmetic mean of the original SOH values ​​of all individual cells in the battery pack.

4. The large-capacity commercial vehicle battery pack SOH remote estimation and cloud-based health early warning system as described in claim 3, characterized in that, S102 further includes: The SOH estimation module determines whether the temperature difference of the target battery pack reaches a preset temperature difference threshold based on the temperature difference matrix. If the maximum temperature difference value in the temperature difference matrix is ​​greater than the preset temperature difference threshold T, then... th Then, a corresponding weighting coefficient matrix is ​​generated based on the temperature difference matrix, specifically including: Based on temperature difference matrix Calculate the sum of the absolute values ​​of the temperature differences between each individual cell i and all other individual cells, expressed as: ； In the formula, It is the sum of the absolute values ​​of the temperature difference between the i-th individual cell and all other individual cells, which reflects the degree of alienation of the i-th individual cell in the overall heat distribution; Will Substituting into the exponential decay function, the weighting coefficient of each individual cell is calculated and expressed as: ； In the formula, The weighting coefficients for the i-th individual cell are used to obtain the final weighting coefficient matrix for the target battery pack. , ,..., } 5. The large-capacity commercial vehicle battery pack SOH remote estimation and cloud-based health early warning system as described in claim 4, characterized in that, S102 further includes: In obtaining the weighting coefficient matrix of each individual battery cell { , ,..., Afterwards, the original SOH data of each individual cell is weighted and normalized, and the vector corresponding to the original SOH data is defined as SOH. 0 =[ , ,..., Using the calculated weight coefficient matrix to adjust the vector SOH 0 The weighted adjustment is expressed as: ； In the formula, The corrected SOH value for the i-th individual cell. This is the estimated initial SOH value for the i-th individual cell.

6. The large-capacity commercial vehicle battery pack SOH remote estimation and cloud-based health early warning system as described in claim 1, characterized in that, In step S201, obtaining the initial SOC of the battery pack based on the charging start voltage specifically includes: First, calculate the open-circuit voltage based on the charging start voltage, charging start current, and the internal resistance of the battery pack. , is represented as: ； In the formula, I is the charging start voltage, and I is the charging start current. The internal resistance of the battery pack; The initial SOC value of the battery pack is obtained by looking up the table using the open-circuit voltage through a pre-established SOC-OCV curve table. This initial SOC value is then corrected for temperature to obtain the final initial SOC, expressed as: ； In the formula, The initial SOC of the battery pack, Represents the reference temperature The reference SOC value is obtained by mapping the open-circuit voltage. The preset reference temperature, The actual temperature of the battery pack is measured at the current moment, and k is the temperature correction factor for the battery pack at the current voltage point.

7. The large-capacity commercial vehicle battery pack SOH remote estimation and cloud-based health early warning system as described in claim 6, characterized in that, S202 specifically includes: The SOC calibration module in the cloud platform layer calculates the battery pack's true SOC at the end of charging based on the initial SOC and the total charging capacity, expressed as: ； In the formula, This represents the actual SOC value of the battery pack at the end of charging. The initial SOC value, Total charging capacity This is the charging efficiency coefficient. , The rated total energy storage capacity of the battery pack is specified at the factory; then the actual SOC value of the battery pack is... The data is sent to the vehicle-mounted BMS as a reference benchmark for deviation correction of the SOC calculation model, thereby triggering the BMS to smoothly correct and synchronize the current SOC estimate.

8. The large-capacity commercial vehicle battery pack SOH remote estimation and cloud-based health early warning system as described in claim 7, characterized in that, S202 further includes: Calculate the deviation between the actual SOC and the real-time SOC estimate from the BMS. , = - ,in This is the real-time SOC estimate from the BMS side; using this deviation... The maximum usable capacity parameter of the battery in the dynamically updated SOH estimation model of the battery pack is expressed as: ； ； ； In the formula, This refers to the updated maximum usable battery capacity parameter in the battery pack SOH estimation model. To provide the maximum usable capacity parameter of the battery before the update in the battery pack SOH estimation model, It is the reciprocal of the capacity calculated based on the actual increase in electricity consumption. The incremental capacity is calculated based on BMS, where K is a preset gain coefficient. For symbolic functions, The nonlinear mapping representing the intensity of the deviation. As an exponential factor, By adjusting Perform strong correction for large deviations and weak correction for small deviations.

9. The large-capacity commercial vehicle battery pack SOH remote estimation and cloud-based health early warning system as described in claim 7, characterized in that, S202 further includes: A global calibration factor is generated based on the battery's maximum available capacity parameters before and after the update, expressed as: This global calibration factor A correction ratio used to reflect the overall capacity of the battery pack; utilizing a global calibration factor. Using the currently obtained weight coefficient matrix, the maximum usable capacity parameter of the battery in the SOH estimation model of each individual cell is updated by weighted mapping to achieve parameter refinement, as shown below: ； In the formula, This represents the maximum usable capacity of the i-th individual cell after the update. This represents the maximum usable capacity of the i-th individual cell before the update. is the weighting coefficient for the i-th individual cell.

10. The large-capacity commercial vehicle battery pack SOH remote estimation and cloud-based health early warning system as described in claim 5, characterized in that, S103 specifically includes: The health early warning module in the cloud platform layer receives the comprehensive SOH value output by the SOH estimation module and compares the comprehensive SOH value with the first preset warning threshold and the second preset warning threshold respectively. If the comprehensive SOH value is greater than or equal to the first preset warning threshold, the warning condition is not met. If the comprehensive SOH value is less than the first preset warning threshold but greater than or equal to the second preset warning threshold, a first-level warning signal is generated. If the comprehensive SOH value is less than the second preset warning threshold, a second-level warning signal is generated.

Images