Battery soc correction method and system, cloud estimation method and device, and vehicle

Abstract

The application relates to a battery SOC correction method and system, a cloud estimation method and device, and a vehicle, which realize real-time and accurate estimation of a battery SOC. The method comprises the following steps: acquiring real-time operation parameters of a battery; inputting real-time temperature, real-time voltage and real-time current of the battery into a pre-constructed lightweight estimation model to obtain a first estimation value; performing ampere-hour integration on the real-time current of the battery to obtain a second estimation value; obtaining an SOC initial value according to the first estimation value and the second estimation value; when the operation condition of the battery belongs to a predetermined key condition, uploading the SOC initial value and a data segment containing the real-time operation parameters of the battery to the cloud to enable the cloud to determine whether to perform lightweight estimation network updating and to issue an SOC correction value; and when the updated lightweight estimation network and the SOC correction value issued by the cloud are received, performing lightweight estimation network updating, and obtaining a final SOC value of the battery according to the SOC initial value and the SOC correction value.

Description

Technical Field

[0001] This application relates to the field of batteries, specifically to a method and system for correcting battery SOC, a cloud-based estimation method, an apparatus, and a vehicle. Background Technology

[0002] Lithium iron phosphate (LiFePO4, or LFP) batteries are gradually becoming a key battery technology in the fields of electric vehicles and energy storage due to their cost-effectiveness, high safety and long cycle life.

[0003] In the current technical field, SOC estimation for batteries is usually based on the ampere-hour integration method. However, this method is not suitable for SOC estimation of LFP batteries because: the ampere-hour integration method relies on accurate measurement and time integration of the battery's charging and discharging current. LFP batteries have a very flat voltage plateau, unlike ternary lithium batteries which can periodically correct the integration results through a clear voltage inflection point. Therefore, current measurement errors, battery self-discharge, and complex electrochemical processes lead to a continuous accumulation of integration errors, causing the SOC estimate obtained by the ampere-hour integration method to gradually deviate from the actual value. Furthermore, the electrochemical performance of lithium iron phosphate batteries varies significantly under different operating conditions such as temperature and charge / discharge rates, making it difficult for traditional single ampere-hour integration estimation models to maintain high-accuracy SOC estimation under all operating conditions.

[0004] Therefore, existing technologies also include techniques for calibrating the State of Charge (SOC) estimated using the ampere-hour integration method by utilizing big data models in the cloud. Theoretically, using cloud-based big data for SOC calibration can improve the accuracy of SOC estimation; however, existing methods typically require uploading all data from the battery's operation to a cloud server. This not only places extremely high demands on communication network bandwidth but may also lead to data transmission delays or even interruptions, affecting the timeliness and accuracy of calibration. This results in an inability to respond in real-time to rapid dynamic changes in battery state, making it difficult to meet the demand for real-time and accurate SOC estimation during electric vehicle operation. Furthermore, vehicle-side devices (such as onboard BMS) have limited computing resources due to cost, power consumption, and space constraints; complex, high-precision big data models deployed in the cloud are difficult to directly deploy on these resource-constrained edge devices. Summary of the Invention

[0005] This application provides a method and system for correcting battery SOC, a cloud-based estimation method, device, and vehicle, enabling real-time and accurate estimation of battery SOC.

[0006] The technical solution of this application is as follows:

[0007] This application provides a cloud-based method for estimating battery SOC, including:

[0008] Obtain the initial SOC value uploaded by the vehicle and the data segment containing real-time operating parameters of the battery;

[0009] Based on the data segment containing the real-time operating parameters of the battery, the first estimation model is run to obtain the third SOC estimate;

[0010] Based on the data segment containing the real-time operating parameters of the battery, the second estimation model is run to obtain the fourth SOC estimate;

[0011] Based on the third SOC estimate and the fourth SOC estimate, the SOC calibration value is obtained;

[0012] Based on the SOC calibration value and the initial SOC value, determine whether to update the lightweight estimation model;

[0013] When it is determined that the lightweight estimation model needs to be updated, the lightweight estimation model is updated using the SOC calibration value and a data segment containing real-time operating parameters of the battery.

[0014] The updated lightweight estimation network is sent to the vehicle, and the difference between the SOC calibration value and the initial SOC value is sent to the vehicle as the SOC correction value.

[0015] Preferably, the step of obtaining the SOC calibration value based on the third SOC estimate and the fourth SOC estimate includes:

[0016] The third SOC estimate is multiplied by the third preset weighting coefficient to obtain the third product value;

[0017] The fourth SOC estimate is multiplied by the fourth preset weighting coefficient to obtain the fourth product value;

[0018] The third product value and the fourth product value are added together to obtain the SOC calibration value;

[0019] The sum of the third preset weighting coefficient and the fourth preset weighting coefficient is 1;

[0020] The third preset weighting coefficient and the fourth preset weighting coefficient are determined based on the aging degree of the battery and / or the current operating condition of the battery.

[0021] This application also provides a method for correcting battery SOC, including:

[0022] Obtain the real-time operating parameters of the battery; the real-time operating parameters of the battery include: real-time temperature, real-time voltage and real-time current of the battery;

[0023] The battery's real-time temperature, real-time voltage, and real-time current are input into a pre-built lightweight estimation model to obtain the first SOC estimate.

[0024] The real-time current of the battery is integrated in ampere-hours to obtain the second SOC estimate;

[0025] Based on the first SOC estimate and the second SOC estimate, the initial SOC value is obtained;

[0026] When the battery's operating condition is identified as a predetermined critical condition, the initial SOC value and a data segment containing the battery's real-time operating parameters are uploaded to the cloud for the cloud to determine whether to perform a lightweight estimation network update and distribute the SOC correction value.

[0027] Upon receiving the updated lightweight estimation network and SOC correction value from the cloud, the lightweight estimation network is updated, and the final SOC value of the battery is obtained based on the initial SOC value and the SOC correction value.

[0028] If the updated lightweight estimation network and SOC correction value are not received from the cloud, the initial SOC value will be used as the final SOC value of the battery.

[0029] In this process, the cloud-based method for estimating battery SOC is used to determine the updated lightweight estimation network and the SOC correction value.

[0030] Preferably, the battery operates under at least one of the following conditions:

[0031] The rate of change of battery voltage exceeds a first predetermined rate of change;

[0032] The battery's temperature change rate exceeds the second predetermined change rate;

[0033] The battery undergoes a charge / discharge state switch;

[0034] The battery's operating condition was determined to be a predetermined critical condition.

[0035] Preferably, the step of obtaining the initial SOC value based on the first SOC estimate and the second SOC estimate includes:

[0036] The first SOC estimate is multiplied by the first preset weighting coefficient to obtain the first product value;

[0037] Multiply the second SOC estimate by the second preset weighting coefficient to obtain the second product value;

[0038] Add the first product value and the second product value to obtain the initial SOC value;

[0039] The sum of the first preset weighting coefficient and the second preset weighting coefficient is 1;

[0040] The first preset weighting coefficient and the second preset weighting coefficient are determined based on the real-time temperature of the battery.

[0041] Preferably, the step of obtaining the final SOC value of the battery based on the initial SOC value and the corrected SOC value includes:

[0042] The initial SOC value is added to the corrected SOC value to obtain the final SOC value of the battery.

[0043] Preferably, when the battery's operating condition is identified as a predetermined critical operating condition, uploading the initial SOC value and a data segment containing the battery's real-time operating parameters to the cloud includes:

[0044] When the battery's SOC is greater than or equal to a preset SOC and the battery's discharge rate is less than or equal to a preset discharge rate, the data segment is uploaded to the cloud at a first preset frequency.

[0045] When the battery's SOC is less than the preset SOC or the battery's discharge rate is greater than the preset discharge rate, the data segment is uploaded to the cloud at the second preset frequency.

[0046] The first preset frequency is less than the second preset frequency.

[0047] This application also provides a battery SOC correction system, comprising: a controller and an on-board connected terminal connected to each other; the controller is used for:

[0048] Obtain the real-time operating parameters of the battery; the real-time operating parameters of the battery include: real-time temperature, real-time voltage and real-time current of the battery;

[0049] The battery's real-time temperature, real-time voltage, and real-time current are input into a pre-built lightweight estimation model to obtain the first SOC estimate.

[0050] The real-time current of the battery is integrated in ampere-hours to obtain the second SOC estimate;

[0051] Based on the first SOC estimate and the second SOC estimate, the initial SOC value is obtained;

[0052] When the battery's operating condition is identified as a predetermined critical condition, the initial SOC value and a data segment containing the battery's real-time operating parameters are uploaded to the cloud via the vehicle-mounted network terminal, so that the cloud can determine whether to perform a lightweight estimation network update and issue a corrected SOC value.

[0053] When the vehicle-mounted connected terminal receives the updated lightweight estimation network and SOC correction value from the cloud, it updates the lightweight estimation network and obtains the final SOC value of the battery based on the initial SOC value and the SOC correction value.

[0054] If the updated lightweight estimation network and SOC correction value are not received from the cloud through the vehicle-mounted network terminal, the initial SOC value will be used as the final SOC value of the battery.

[0055] In this process, the cloud-based method for estimating battery SOC is used to determine the updated lightweight estimation network and the SOC correction value.

[0056] This application also provides a cloud-based device for estimating battery SOC, including:

[0057] The acquisition module is used to acquire the initial SOC value uploaded by the vehicle and a data segment containing real-time operating parameters of the battery.

[0058] The first estimation module is used to run the first estimation model based on the data segment containing the real-time operating parameters of the battery to obtain the third SOC estimation value.

[0059] The second estimation module is used to run the second estimation model based on the data segment containing the real-time operating parameters of the battery to obtain the fourth SOC estimate.

[0060] The SOC calibration value estimation module is used to obtain the SOC calibration value based on the third SOC estimation value and the fourth SOC estimation value.

[0061] The model update judgment module is used to determine whether to update the vehicle-side lightweight estimation model based on the SOC calibration value and the initial SOC value.

[0062] The model update module is used to update the lightweight estimation model by means of the SOC calibration value and a data segment containing real-time operating parameters of the battery when it is determined that the vehicle-side lightweight estimation model needs to be updated.

[0063] The distribution module is used to distribute the updated lightweight estimation network to the vehicle and distribute the difference between the SOC calibration value and the initial SOC value as the SOC correction value to the vehicle.

[0064] This application also provides a vehicle including the aforementioned battery SOC correction system.

[0065] The beneficial effects of this invention are as follows:

[0066] Due to the limited computing power of the vehicle-mounted controller, a lightweight estimation model is deployed on the vehicle side. This lightweight model, along with ampere-hour integration, handles SOC estimation under normal operating conditions, enabling real-time SOC estimation. For certain critical operating conditions, data segments at relevant time points are uploaded to the cloud. A complex estimation model deployed in the cloud calibrates the vehicle-mounted estimation results, achieving accurate SOC estimation. Furthermore, only data segments relevant to critical operating conditions are uploaded to the cloud, rather than all data as in traditional methods. This significantly reduces the amount of data uploaded from the vehicle to the cloud, effectively reducing communication bandwidth pressure. Attached Figure Description

[0067] Figure 1 This is a flowchart illustrating the battery SOC correction method in an embodiment of this application.

[0068] Figure 2 This is a flowchart illustrating step S104 in an embodiment of this application;

[0069] Figure 3 This is a flowchart illustrating the cloud-based estimation method for battery SOC in an embodiment of this application.

[0070] Figure 4 This is a comparison diagram showing the estimation error between the battery SOC correction method in this application embodiment and the traditional SOC estimation method;

[0071] Figure 5 This is a schematic diagram of the vehicle structure in an embodiment of this application. Detailed Implementation

[0072] Reference Figure 1 This application provides a method for correcting battery SOC, including:

[0073] S101, Obtain the real-time operating parameters of the battery; the real-time operating parameters of the battery include: real-time temperature, real-time voltage and real-time current of the battery;

[0074] S102, input the real-time temperature, real-time voltage and real-time current of the battery into the pre-built lightweight estimation model to obtain the first SOC estimate;

[0075] S103, perform ampere-hour integration on the real-time current of the battery to obtain the second SOC estimate;

[0076] S104, Based on the first SOC estimate and the second SOC estimate, obtain the initial SOC value;

[0077] S105, when it is identified that the battery's operating condition belongs to a predetermined critical operating condition, the initial SOC value and the data segment containing the real-time operating parameters of the battery are uploaded to the cloud for the cloud to determine whether to perform a lightweight estimation network update and distribute the SOC correction value.

[0078] S106: Upon receiving the updated lightweight estimation network and SOC correction value from the cloud, the lightweight estimation network is updated, and the final SOC value of the battery is obtained based on the initial SOC value and the SOC correction value.

[0079] S107, if the updated lightweight estimation network and SOC correction value are not received from the cloud, the initial SOC value is used as the final SOC value of the battery.

[0080] The battery SOC correction method in this application embodiment is applied to the vehicle end, specifically, for example, the vehicle's battery management system (BMS). The BMS, for example, uses a high-performance and resource-optimized STM32H743 MCU as the main control chip. This chip integrates rich peripherals and a high-speed processing core, capable of meeting the complex data processing needs of the vehicle end.

[0081] In step S101, the real-time temperature, real-time voltage, and real-time current of the battery are all acquired through physical hardware. The real-time voltage, real-time current, and real-time temperature timing data are obtained as input at a sampling rate of 1Hz. This data is preprocessed by a signal conditioning circuit to remove noise and outliers, ensuring the accuracy of the input data. For example, the battery management system incorporates an ASE chip. The ASE chip acquires the real-time current of the battery through a current sensor. The current sensor uses a high-precision Hall effect sensor to accurately measure current changes during battery charging and discharging. It also acquires the real-time voltage of the battery through a voltage sensor. The voltage sensor uses a voltage divider resistor network combined with a high-precision ADC to achieve high-precision measurement of the battery voltage. Finally, it acquires the real-time temperature of the battery through a temperature sensor. The temperature sensor is a digital temperature sensor that directly outputs a digital signal, reducing signal conversion errors and ensuring accurate acquisition of battery temperature information.

[0082] In this embodiment, the lightweight estimation model in step S102 is an LTSM network. This LTSM network accelerates the inference process using an FPGA built into the battery management system (BMS). The FPGA performs parallel computation on the LSTM network through hardware programming, significantly improving the inference speed of the LSTM network and ensuring an accurate first SOC estimate is output in a short time.

[0083] To address the limited computing resources on the vehicle side, knowledge distillation technology was used to transfer key knowledge from a complex 10-layer deep reinforcement learning model in the cloud to a 2-layer LSTM network. This significantly reduced the model's complexity and computational load while ensuring that the model's accuracy loss was less than 0.5%. The complex deep reinforcement learning model in the cloud served as the teacher model, and the LSTM network served as the student model. By adjusting the structure and parameters of the student model, it was possible to mimic the teacher model's output as closely as possible while maintaining low complexity. After multiple iterations of training, the LSTM network size was compressed to below 100KB-150KB, and the inference latency was reduced to less than 10ms-15ms, while maintaining an accuracy loss of less than 0.5% compared to the deep reinforcement learning model. This met the vehicle's requirements for lightweight and real-time performance.

[0084] Reference Figure 2 In this embodiment of the application, step S104 specifically includes:

[0085] S1041, Multiply the first SOC estimate by the first preset weight coefficient to obtain the first product value;

[0086] S1042, Multiply the second SOC estimate by the second preset weighting coefficient to obtain the second product value;

[0087] S1043, add the first product value and the second product value to obtain the initial value of SOC;

[0088] The sum of the first preset weighting coefficient and the second preset weighting coefficient is 1;

[0089] The first preset weighting coefficient and the second preset weighting coefficient are determined based on the real-time temperature of the battery.

[0090] Specifically, by using a pre-established lookup table of real-time battery temperature and weighting coefficients, the values ​​of the first preset weighting coefficient k1 and the second preset weighting coefficient k2 are quickly and accurately determined based on the battery's real-time temperature. This temperature-adaptive weighting adjustment mechanism can dynamically optimize the SOC estimation formula according to the changes in battery characteristics at different temperatures, further improving the accuracy of the estimation.

[0091] In this embodiment of the application, the process of establishing the real-time temperature and weighting coefficient lookup table of the battery includes:

[0092] First, a variety of representative battery samples were selected, covering different models, batches, and service lives, to ensure a comprehensive reflection of battery performance in real-world applications. Charge-discharge experiments were conducted on N batteries under varying ambient temperatures. A series of discrete temperature points were set, for example, from -40℃ to 150℃, at 5℃ or 10℃ intervals, covering extreme low and high temperature conditions that the batteries might encounter. During the charge-discharge process at each temperature point, various battery parameters were precisely measured and recorded, including but not limited to real-time voltage, current, precise SOC measurements, and the corresponding real-time temperature.

[0093] Next, the SOC of each battery at each temperature point T is estimated using the two estimation methods in S102 and S103, resulting in the corresponding first SOC estimate and second SOC estimate. The first SOC estimate is then used to estimate the SOC value. ocv Second SOC estimate SOC Ah Compare with the actual measured SOC value. true By comparison, the first SOC error and the second SOC error of each battery at each temperature point T are calculated, and then the average first SOC error of N batteries at each temperature point T is obtained. ε OCV ( T ) and the mean error of the second SOC ε Ah ( T ), where:

[0094]

[0095] .

[0096] Finally, based on the first SOC mean SOC ocv Second SOC error mean SOC Ah The first preset weighting coefficient k1 and the second preset weighting coefficient k2 corresponding to each temperature point T are calculated according to the following formula:

[0097]

[0098] By repeating the above process, you can remember the k1 and k2 data corresponding to each of the test temperature points.

[0099] The various temperature points obtained from the experiment, along with the weighting coefficients k1 and k2 determined at each temperature point, are compiled into a table. One column of the table records the real-time temperature value of the battery, and the other two columns record the corresponding k1 and k2 values, respectively.

[0100] To more accurately obtain weighting coefficients based on real-time temperature in practical applications, interpolation processing of the generated lookup table may be necessary. For example, when the actual measured temperature falls between two adjacent temperature points in the lookup table, the corresponding weighting coefficients can be estimated using linear interpolation or more complex interpolation methods (such as spline interpolation). Simultaneously, changes in the weighting coefficients should be smoothed to avoid abrupt changes in the weighting coefficients due to small temperature variations, thus ensuring the stability of the SOC estimation.

[0101] In step S105, if the battery's operating conditions meet at least one of the following conditions:

[0102] The rate of change of battery voltage exceeds a first predetermined rate of change;

[0103] The battery's temperature change rate exceeds the second predetermined change rate;

[0104] The battery undergoes a charge / discharge state switch;

[0105] The battery's operating condition was determined to be a predetermined critical condition.

[0106] Because these predetermined critical operating conditions are prone to causing battery SOC errors, and the vehicle's computing power is limited and the voltage plateau of lithium iron phosphate batteries is flat, making it impossible to self-correct integral errors; therefore, only at highly sensitive moments such as sudden changes in battery voltage, sudden temperature changes, or switching between charge and discharge states, the data segments that are most likely to expose SOC errors are uploaded to the cloud, relying on the more complex and accurate models in the cloud to reduce such SOC errors.

[0107] Specifically, in the embodiments of this application, the first predetermined rate of change is, for example, 0.1 V / min, and the second predetermined rate of change is, for example, 2°C / min.

[0108] In this application embodiment, the data segment containing the real-time operating parameters of the battery not only includes the instantaneous data that meets the predetermined critical operating condition trigger time, but also covers the continuous voltage, current and temperature sequence from a period of time before the trigger time (such as the first 30 seconds) to a period of time after the trigger time (such as the last 10 seconds), so as to comprehensively record the dynamic changes of the battery state under critical operating conditions.

[0109] Furthermore, in this embodiment, when the battery is under conditions requiring extremely high SOC estimation accuracy, such as low SOC or high-rate discharge, the data upload frequency is reduced to ensure the accuracy of SOC estimation. Low SOC is, for example, less than 20%, and high rate discharge is, for example, greater than 2C. When the battery is under conditions requiring extremely high SOC estimation accuracy, such as low SOC (e.g., less than 20%) or high-rate discharge (e.g., greater than 2C), the data upload frequency is significantly increased from the default 30 minutes / time to 5 minutes / time. In this state, the vehicle collects and uploads data segments for predetermined key operating conditions more frequently, enabling the cloud to obtain battery state change information more promptly, thereby calibrating the SOC more frequently and meeting the demand for high-precision SOC estimation under complex operating conditions.

[0110] In the embodiments of the application, differential privacy technology is used to protect data to ensure data security and privacy during transmission. Before uploading data segments from the vehicle, noise is added to data segments under predetermined critical operating conditions. The intensity of the noise is dynamically adjusted according to the sensitivity of the data and the required level of privacy protection. Simultaneously, the MQTT protocol is selected for data transmission. This protocol has advantages such as lightweight design, low bandwidth consumption, and support for asynchronous communication, making it particularly suitable for achieving stable data transmission between the vehicle and the cloud in scenarios with complex network conditions and limited bandwidth. The communication module between the vehicle and the cloud uses an NB-IoT module, which supports low-power wide-area network (LPWAN) communication and is suitable for scenarios where devices are widely distributed in energy storage systems, enabling stable communication under low bandwidth and low power consumption conditions.

[0111] Furthermore, when the battery is in a quiescent state and the State of Charge (SOC) is stable (e.g., SOC decreases by 0.01% within 10 minutes), communication between the vehicle and the cloud is suspended to reduce system power consumption and extend device operating time. At this time, the vehicle's Battery Management System (BMS) retains only basic battery parameter monitoring functions, disables data upload and reception, and enters a low-power sleep state. When a change in battery status is detected, such as the battery voltage change rate exceeding a first predetermined rate, the battery temperature change rate exceeding a second predetermined rate, or a charge / discharge state switch, the BMS is automatically woken up to resume normal data acquisition and communication functions.

[0112] For the cloud, after receiving the data segments of predetermined key operating conditions uploaded by the vehicle, it first decrypts and preprocesses the data to remove added noise and restore the original data format. Then, the processed data segments are input into the hybrid model engine, where the first and second estimation models obtain the third and fourth SOC estimates. The outputs of the two estimation models are merged using a dynamic weighted fusion algorithm to generate the final SOC calibration value. The SOC calibration value is compared with the initial SOC value uploaded by the vehicle, and the error between the two is calculated. If the error exceeds a set percentage (e.g., 2%) for three consecutive times, the current lightweight estimation model on the vehicle needs to be updated.

[0113] When updating the current lightweight estimation model in the cloud, multi-node data aggregation and analysis technology is used to collect data segments uploaded from multiple vehicle-side devices to deeply optimize the weights of the LSTM model. Distributed computing and parallel processing technologies are used to accelerate data processing and model training. During optimization, algorithms such as stochastic gradient descent are employed to continuously adjust the weights and biases of the LSTM model to minimize the error between the first SOC estimate and the actual SOC value. After optimization, the LSTM model is pruned and quantized to remove redundant connections and parameters, and the parameters are represented in a low-precision data format to generate a new lightweight estimation model version, which is periodically distributed to the vehicle-side.

[0114] In the embodiments of this application, multi-node data refers to the real-time voltage, real-time current, and real-time temperature of each cell unit.

[0115] For the cloud, when it determines that the current lightweight estimation model on the vehicle needs to be updated, it needs to perform weighted processing based on the third SOC estimation value and the fourth SOC estimation value to obtain the SOC calibration value, and then send the SOC calibration value to the vehicle.

[0116] In other words, refer to Figure 3 In this embodiment of the application, the following SOC estimation logic is executed in the cloud:

[0117] S201, Obtain the initial SOC value uploaded by the vehicle and the data segment containing the real-time operating parameters of the battery;

[0118] S202, Based on the data segment containing the real-time operating parameters of the battery, run the first estimation model to obtain the third SOC estimate;

[0119] S203, Based on the data segment containing the real-time operating parameters of the battery, run the second estimation model to obtain the fourth SOC estimate;

[0120] S204, Based on the third SOC estimate and the fourth SOC estimate, obtain the SOC calibration value;

[0121] S205, Based on the SOC calibration value and the initial SOC value, determine whether to update the vehicle-side lightweight estimation model;

[0122] S206, When it is determined that the lightweight estimation model of the vehicle needs to be updated, the lightweight estimation model is updated using the SOC calibration value and the data segment containing the real-time operating parameters of the battery.

[0123] S207, the updated lightweight estimation network is sent to the vehicle, and the difference between the SOC calibration value and the initial SOC value is sent to the vehicle as the SOC correction value.

[0124] In step S204, the cloud performs weighted processing on the third and fourth SOC estimates to obtain the required SOC calibration value; specifically, the following steps are used:

[0125] The third SOC estimate is multiplied by the third preset weighting coefficient to obtain the third product value;

[0126] The fourth SOC estimate is multiplied by the fourth preset weighting coefficient to obtain the fourth product value;

[0127] The third product value and the fourth product value are added together to obtain the SOC calibration value;

[0128] The sum of the third preset weighting coefficient and the fourth preset weighting coefficient is 1;

[0129] The third and fourth preset weighting coefficients are determined based on the battery's aging level and / or current operating condition to obtain the most accurate SOC calibration value. For example, the first estimation model is an electrochemical model, and the second estimation model is a deep reinforcement learning model; in the early stages of battery aging, the electrochemical model has a higher weight; while in the later stages of battery aging, the weight of the deep reinforcement learning model gradually increases.

[0130] In this embodiment, battery aging includes the number of charge-discharge cycles and changes in battery internal resistance. The number of charge-discharge cycles and changes in battery internal resistance can be periodically uploaded from the vehicle to the cloud.

[0131] The current operating conditions of the battery include its current temperature and current charge / discharge rate. The current charge / discharge rate is calculated based on the current current transmitted from the vehicle. The charge / discharge rate (C) is defined as the ratio of the charge / discharge current (I) to the battery's rated capacity (Ce), i.e., C = I / Ce. The vehicle's Battery Management System (BMS) monitors the battery's charge / discharge current in real time, and since the battery's rated capacity is known, the cloud can calculate the current charge / discharge rate using simple methods given the vehicle's charge / discharge current.

[0132] In this embodiment, the first estimation model in the cloud is, for example, an electrochemical model, specifically a simplified version of the P2D model. The electrochemical model is based on the physicochemical principles within the battery, combined with the battery's actual operating parameters (such as electrode material characteristics and electrolyte conductivity), and by solving a series of partial differential equations, it provides a preliminary estimate of the battery's State of Charge (SOC), offering a general range and boundary constraints for SOC based on physical laws.

[0133] In this embodiment, the first estimation model in the cloud is specifically based on the principle of estimating SOC using the Nernst equation. The Nernst equation establishes a quantitative relationship between electrode potential and the concentration of substances participating in the electrode reaction, playing a crucial role in estimating the SOC of secondary ion batteries. The general expression of the Nernst equation is: ,in:

[0134] E represents the actual measured electrode potential, which can be indirectly determined by the real-time voltage uploaded to the cloud. The electrode potential reflects the actual potential of the electrodes in the current state of the battery and is a key measurable parameter in the Nernst equation.

[0135] E0 is the standard electrode potential, which depends on the properties of the electrode material itself. This value is preset in the cloud.

[0136] R is the gas constant, with a value of 8.314 J·mol⁻¹. -1 ·K -1 .

[0137] T represents the absolute temperature of the battery, a value preset in the cloud.

[0138] n is the number of electrons transferred in the electrode reaction, and this value is preset in the cloud.

[0139] F is the Faraday constant, with a value of 96485 C·mol⁻¹. -1 .

[0140] Assuming the total amount of electrode material is known, the ion content in the current electrode can be determined based on the calculated ratio of oxidized to reduced concentrations, thereby further calculating the battery's state of charge (SOC).

[0141] When applying the Nernst equation to secondary ion batteries in vehicles, the electrode potential E indirectly obtained from the battery voltage, along with known parameters such as the standard electrode potential E0 and the battery's absolute temperature T, allows for the calculation of the ratio of oxidized to reduced state concentrations based on the Nernst equation. Since the state of charge (SOC) of a secondary ion battery is directly related to the amount of ion insertion / extraction in the electrodes, and this amount is closely linked to the concentration ratio of oxidized and reduced substances, this method of estimating SOC based on the Nernst equation utilizes the fundamental thermodynamic principles of the electrochemical reactions within the secondary ion battery. By measuring and calculating the relationship between electrode potential and substance concentration, it provides a reliable method for SOC estimation based on physical laws, and in this electrochemical model, it is used to determine the approximate range and boundary constraints of the SOC.

[0142] In this embodiment, the second estimation model is a deep reinforcement learning model. The deep reinforcement learning model takes historical operating condition data and currently uploaded data segments as input. Through unsupervised learning from a large amount of historical operating condition data, it mines the complex nonlinear relationships in the battery operation process, and performs a more accurate nonlinear fitting of the SOC. In this embodiment, the actual training and estimation process of the deep reinforcement learning model includes:

[0143] 1) First, collect a massive amount of historical battery operating data. This data covers various operating parameters of the battery under different operating conditions, including but not limited to time-varying sequence data such as real-time voltage, real-time current, and real-time temperature, as well as the corresponding high-precision SOC measurement values ​​(SOC). actual These data come from a wide range of sources and may include operational records of different types of secondary ion batteries in various application scenarios, such as data collection from electric vehicles under stationary, driving, high-temperature, and low-temperature conditions.

[0144] 2) The collected data often contains problems such as noise and missing values, so preprocessing is required.

[0145] 3) Use filtering algorithms, such as Gaussian filtering or median filtering, to remove random noise from voltage, current and temperature data, making the data smoother and reducing the interference of noise on subsequent analysis.

[0146] 4) For data points with missing values, use linear interpolation or machine learning-based interpolation methods (such as the K-nearest neighbor algorithm) to fill in the missing values ​​based on the time series characteristics of the data, ensuring data integrity.

[0147] 5) Map feature data of different ranges (e.g., voltage may be in the range of a few volts, while current may be in the range of milliamperes to amperes) to the same normalization interval, usually [0, 1] or [-1, 1]. Normalization can speed up the convergence of the model and improve its stability and generalization ability.

[0148] In this embodiment, the deep reinforcement learning model employs a Long Short-Term Memory (LSTM) network structure, which has powerful feature extraction capabilities. LSTM is particularly suitable for processing time-series data of battery operation because it can effectively capture long-term dependencies in the data, overcoming the gradient vanishing or exploding problems of traditional methods when processing long-term series.

[0149] LSTM networks control the inflow, retention, and output of information through gating mechanisms (input gate, forget gate, and output gate). The input gate determines how much information from the current input data will be added to the cell state. It is implemented using a sigmoid layer and a tanh layer. The sigmoid layer outputs a value between 0 and 1, representing the importance weights of each dimension of the input data, while the tanh layer creates a new vector of candidate values. Multiplying the two yields the information added to the cell state. The forget gate determines which information to discard from the cell state. Also implemented using a sigmoid layer, its output value is between 0 and 1, where 0 indicates complete discarding of the information and 1 indicates complete retention. Through the forget gate, LSTM can selectively forget information from past cell states, thus better adapting to new input data. The output gate determines which parts of the cell state will be output. It consists of a sigmoid layer and a tanh layer. The sigmoid layer determines which parts of the cell state will be output, and the tanh layer transforms the cell state and multiplies it with the output of the sigmoid layer to obtain the final output.

[0150] When processing battery operating data, the LSTM network sequentially inputs time-series data such as real-time voltage, real-time current, and real-time temperature. Through a gating mechanism, the LSTM network can capture the complex patterns and trends of these parameters changing over time, transforming them into more abstract feature representations. For example, it can learn the unique response patterns of battery voltage and current to changes in SOC at different temperatures. These feature representations contain key information about the battery's operation, laying the foundation for subsequent exploration of complex nonlinear relationships.

[0151] The goal of an LSTM model is to learn an optimal policy to accurately predict SOC, making the predicted value as close as possible to the actual SOC measurement. This policy is learned through interaction with the environment; at each time step, the model selects an action (the predicted SOC value) based on the current environmental state (i.e., the extracted feature representation). The environment will provide a reward (r) based on this action. The design of the reward function is crucial, as it guides the model to learn in the right direction. For example, the reward function can be defined as the predicted SOC value. Compared with actual measured value The reciprocal r of the error between: Where ε is a very small positive number (to prevent the denominator from being zero), the smaller the error, the greater the reward. The model continuously adjusts its parameters to maximize the long-term cumulative reward, thereby gradually learning the complex nonlinear relationship between battery operating parameters and SOC.

[0152] To measure the difference between predicted and actual values, the model uses a loss function. A loss function such as the mean squared error (MSE) loss function is expressed as:

[0153]

[0154] Where n0 is the total number of data points in the training dataset. and They are the first i The predicted SOC value and the actual SOC value for each data point. The MSE loss function can intuitively reflect the degree of deviation between the predicted and actual values; the greater the deviation, the greater the loss value.

[0155] The model uses backpropagation to calculate the gradient of the loss function with respect to the model parameters. Starting with the loss function, backpropagation uses a chain rule to propagate the gradient from the output layer back to each layer of the network, calculating the contribution of each parameter to the loss function. Then, an optimizer (such as Stochastic Gradient Descent (SGD) or Adam) updates the model parameters based on the calculated gradients, continuously reducing the loss function. Taking the Adam optimizer as an example, it combines the advantages of momentum and RMSProp algorithms, adaptively adjusting the learning rate of each parameter, enabling faster convergence to the optimal solution during training. As training progresses, the model gradually adjusts its parameters to better fit the nonlinear relationship between battery operating parameters and State of Charge (SOC), thus enabling more accurate predictions of SOC.

[0156] In summary, deep reinforcement learning models, through a series of steps including battery operation data collection, preprocessing, feature extraction, policy learning, and model optimization, uncover the complex nonlinear relationships in the battery operation process, achieving high-precision nonlinear fitting of battery SOC, and providing key support for edge-cloud collaborative battery SOC dynamic calibration systems.

[0157] In this embodiment of the application, for example, an AWS EC2 P4 instance or an Alibaba Cloud ECS g6 instance is used as a server in the cloud. This instance is equipped with a powerful GPU, which can accelerate the training and computation of deep learning models.

[0158] On cloud servers, model services can be deployed based on the PyTorch deep learning framework or the TensorFlow deep learning framework. By utilizing PyTorch's automatic differentiation and distributed training features or TensorFlow's powerful model training and optimization capabilities, efficient training and optimization of hybrid models can be achieved.

[0159] A data storage system and a data management system are built on a cloud server to store and manage lightweight estimation model parameters from the vehicle and data segments uploaded from multiple vehicle terminals, ensuring secure data storage and fast retrieval. The data management system uses Alibaba Cloud database services, supporting large-scale data storage and fast retrieval, and can efficiently manage key operating condition data segments and model parameters from multiple energy storage units.

[0160] Reference Figure 4 Using the method described in the embodiments of this application, the batteries of multiple electric vehicles were tested under low temperature conditions of -20℃. The results showed that the SOC estimation error was significantly reduced from 8.2% of the traditional method to 2.1%, which greatly improved the accuracy of SOC estimation under low temperature conditions and provided more reliable data support for energy management and range prediction of electric vehicles under low temperature conditions.

[0161] Thanks to the dynamic data segment upload mechanism, the data upload volume is reduced by 76% compared to the traditional method of uploading all data to the cloud. By uploading only data segments for predetermined critical operating conditions, the communication bandwidth pressure is effectively reduced. In actual testing, the communication overhead is less than 50KB per hour, achieving efficient data transmission under low bandwidth conditions. Furthermore, optimized configuration of the MQTT protocol ensures the stability and reliability of data transmission, achieving a data transmission success rate of over 99% in complex network environments.

[0162] This embodiment achieves dynamic SOC calibration through vehicle-to-cloud collaboration, ensuring that the SOC estimation error of LFP batteries is controlled within 2.5% under different temperature and charge / discharge rate conditions, significantly outperforming traditional estimation methods (which typically have errors of 5%-10%). A dynamic data segment upload mechanism reduces data upload volume by 80% compared to traditional full-data upload methods. In actual testing, communication overhead is less than 30KB per hour, achieving efficient data transmission under low bandwidth conditions. Optimized configuration of the NB-IoT module and MQTT protocol ensures the stability and reliability of data transmission. In complex network environments, the data transmission success rate reaches over 98%. In the energy storage system, when the battery is under predetermined critical operating conditions such as low SOC (less than 15%) or high charge / discharge rate (greater than 2C), it automatically switches to a high-frequency calibration mode (data upload frequency increased to once every 10 minutes) to ensure high accuracy and real-time performance of SOC estimation. When the battery is in a static state and the SOC is stable, it automatically enters a sleep mode to reduce system power consumption and extend device operating time.

[0163] This application also provides a battery SOC correction system, including: a controller and an on-board connected terminal connected to each other; the controller is used for:

[0164] Obtain the real-time operating parameters of the battery; the real-time operating parameters of the battery include: real-time temperature, real-time voltage and real-time current of the battery;

[0165] The battery's real-time temperature, real-time voltage, and real-time current are input into a pre-built lightweight estimation model to obtain the first SOC estimate.

[0166] The real-time current of the battery is integrated in ampere-hours to obtain the second SOC estimate;

[0167] Based on the first SOC estimate and the second SOC estimate, the initial SOC value is obtained;

[0168] When the battery's operating condition is identified as a predetermined critical condition, the initial SOC value and a data segment containing the battery's real-time operating parameters are uploaded to the cloud via the vehicle-mounted network terminal, so that the cloud can determine whether to perform a lightweight estimation network update and issue a corrected SOC value.

[0169] When the vehicle-mounted connected terminal receives the updated lightweight estimation network and SOC correction value from the cloud, it updates the lightweight estimation network and obtains the final SOC value of the battery based on the initial SOC value and the SOC correction value.

[0170] If the updated lightweight estimation network and SOC correction value are not received from the cloud through the vehicle-mounted connected terminal, the initial SOC value is used as the final SOC value of the battery.

[0171] In this embodiment, the controller is, for example, a battery management system (BMS) or a vehicle control unit (VCU) controller. Depending on the actual product requirements, either a multi-domain controller or a separate domain controller from the vehicle can be used as the aforementioned controller. The controller can autonomously collect battery current, temperature, and voltage data using the ASE chip described in the preceding embodiments (the data collection principle of the ASE chip is well-known in the prior art).

[0172] In addition, the vehicle-mounted connected terminal refers to the T-Box (Telematics Box) on the vehicle that is responsible for information interaction and communication with the cloud.

[0173] This application also provides a cloud-based device for estimating battery SOC, including:

[0174] The acquisition module is used to acquire the initial SOC value uploaded by the vehicle and a data segment containing real-time operating parameters of the battery.

[0175] The first estimation module is used to run the first estimation model based on the data segment containing the real-time operating parameters of the battery to obtain the third SOC estimation value.

[0176] The second estimation module is used to run the second estimation model based on the data segment containing the real-time operating parameters of the battery to obtain the fourth SOC estimate.

[0177] The SOC calibration value estimation module is used to obtain the SOC calibration value based on the third SOC estimation value and the fourth SOC estimation value.

[0178] The model update judgment module is used to determine whether to update the vehicle-side lightweight estimation model based on the SOC calibration value and the initial SOC value.

[0179] The model update module is used to update the lightweight estimation model by means of the SOC calibration value and a data segment containing real-time operating parameters of the battery when it is determined that the vehicle-side lightweight estimation model needs to be updated.

[0180] The distribution module is used to distribute the updated lightweight estimation network to the vehicle and distribute the difference between the SOC calibration value and the initial SOC value as the SOC correction value to the vehicle.

[0181] The specific principle of the cloud-based SOC estimation device mentioned above refers to the working principle of the cloud-based SOC estimation method mentioned above.

[0182] Figure 5This is a block diagram illustrating a vehicle 200 according to an exemplary embodiment. For example, vehicle 200 may be a hybrid vehicle, a non-hybrid vehicle, an electric vehicle, a fuel cell vehicle, or other types of vehicle. Vehicle 200 may be an autonomous vehicle, a semi-autonomous vehicle, or a non-autonomous vehicle.

[0183] Reference Figure 5 The vehicle 200 may include various subsystems, such as an infotainment system 210, a perception system 220, a decision control system 230, a drive system 240, and a computing platform 250. The vehicle 200 may also include more or fewer subsystems, and each subsystem may include multiple components. Furthermore, each subsystem and component of the vehicle 200 can be interconnected via wired or wireless means. In some embodiments, the infotainment system 210 may include a communication system, an entertainment system, and a navigation system, etc.

[0184] The perception system 220 may include several types of sensors for sensing information about the environment surrounding the vehicle 200. For example, the perception system 220 may include a global positioning system (which may be a GPS system, a BeiDou system, or another positioning system), an inertial measurement unit (IMU), lidar, millimeter-wave radar, ultrasonic radar, and a camera device.

[0185] The decision control system 230 may include a computing system, a vehicle controller, a steering system, a throttle, and a braking system. The drive system 240 may include components that provide power to the vehicle 200. In one embodiment, the drive system 240 may include an engine, an energy source, a transmission system, and wheels. The engine may be one or a combination of internal combustion engines, electric motors, and compressed air engines. The engine is capable of converting energy provided by the energy source into mechanical energy.

[0186] Some or all of the functions of vehicle 200 are controlled by computing platform 250. Computing platform 250 may include at least one processor 251 and memory 252, and processor 251 may execute instructions 253 stored in memory 252.

[0187] Processor 251 can be any conventional processor, such as a commercially available CPU. The processor may also include, for example, a Graphics Processing Unit (GPU), a Field Programmable Gate Array (FPGA), a System on Chip (SOC), an Application Specific Integrated Circuit (ASIC), or a combination thereof.

[0188] The memory 252 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk or optical disk.

[0189] In addition to instruction 253, memory 252 can also store data, such as road maps, route information, vehicle position, direction, speed, and other data. The data stored in memory 252 can be used by computing platform 250.

[0190] In this embodiment of the disclosure, the processor 251 may execute instructions 253 to complete all or part of the steps of the vehicle control method described above.

[0191] This disclosure also provides a computer-readable storage medium having stored thereon computer program instructions that, when executed by a processor, implement the steps of the vehicle control method provided in this disclosure.

[0192] Furthermore, the term “exemplary” is used herein to mean serving as an example, instance, or illustration. Any aspect or design described herein as “exemplary” is not necessarily to be construed as advantageous compared to other aspects or designs. Rather, the use of the term “exemplary” is intended to present the concept in a concrete manner. As used herein, the term “or” is intended to mean an inclusive “or” rather than an exclusive “or.” That is, unless otherwise specified or clear from the context, “X applies A or B” is intended to mean any of the natural inclusive arrangements. That is, “X applies A or B” satisfies any of the foregoing instances if X applies A; X applies B; or both X applies A and B. Additionally, unless otherwise specified or clear from the context to refer to the singular form, the articles “a” and “an” as used in this application and the appended claims are generally understood to mean “one or more.”

[0193] Similarly, although this disclosure has been shown and described with respect to one or more implementations, equivalent variations and modifications will occur to those skilled in the art upon reading and understanding this specification and the accompanying drawings. This disclosure includes all such modifications and variations and is limited only by the scope of the claims. In particular, with respect to the various functions performed by the components described above (e.g., elements, resources, etc.), unless otherwise indicated, the terminology used to describe such components is intended to correspond to any component (functionally equivalent) that performs the specific function of the described component, even if structurally not equivalent to the disclosed structure. Furthermore, although specific features of this disclosure may have been disclosed with respect to only one of several implementations, such features may be combined with one or more other features of other implementations, as may be desired and advantageous for any given or particular application. Moreover, with regard to the terms “comprising,” “owning,” “having,” “having,” or variations thereof as used in the detailed description or claims, such terms are intended to be inclusive in a manner similar to the term “including.”

[0194] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the following claims.

[0195] It should be understood that this disclosure is not limited to the precise structures described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this disclosure is limited only by the appended claims.

[0196] It should be noted that the terms "first," "second," etc., used in the specification, claims, and accompanying drawings of this disclosure are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this disclosure described herein can be implemented in orders other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this disclosure. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this disclosure as detailed in the appended claims.

[0197] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "illustrative embodiment," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0198] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of preferred embodiments of this disclosure includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the function involved, as will be understood by those skilled in the art to which embodiments of this disclosure pertain.

[0199] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a system including a processing module, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of computer-readable media include: an electrical connection having one or more wires (control method), a portable computer disk drive (magnetic device), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic device, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which programs can be printed, because programs can be obtained electronically, for example, by optically scanning the paper or other media, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0200] It should be understood that various parts of the embodiments of this disclosure can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0201] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0202] Furthermore, the functional units in the various embodiments of this disclosure can be integrated into a single processing module, or each unit can exist physically separately, or two or more units can be integrated into a single module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. The aforementioned storage medium can be a read-only memory, a hard disk, or an optical disk, etc.

[0203] Although embodiments of the present disclosure have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present disclosure. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present disclosure.

Claims

1. A cloud-based method for estimating battery SOC, characterized in that, The method comprises: obtaining the SOC initial value uploaded when it is identified that the operation condition of the battery belongs to a predetermined critical condition according to the battery voltage change rate, the battery temperature change rate and / or the charge-discharge state of the battery, and the data segment containing the real-time operation parameters of the battery; obtaining the third SOC estimation value by running the first estimation model according to the data segment containing the real-time operation parameters of the battery; the first estimation model is an electrochemical model; obtaining the fourth SOC estimation value by running the second estimation model according to the data segment containing the real-time operation parameters of the battery; the second estimation model is a deep reinforcement learning model; obtaining the SOC calibration value according to the third SOC estimation value and the fourth SOC estimation value; determining whether to update the lightweight estimation model of the vehicle end according to the SOC calibration value and the SOC initial value; updating the lightweight estimation model by the SOC calibration value and the data segment containing the real-time operation parameters of the battery when it is determined that the lightweight estimation model of the vehicle end needs to be updated; downloading the updated lightweight estimation network to the vehicle end, and downloading the difference between the SOC calibration value and the SOC initial value to the vehicle end as the SOC correction value.

2. The method of claim 1, wherein, The step of obtaining the SOC calibration value according to the third SOC estimation value and the fourth SOC estimation value comprises: multiplying the third SOC estimation value by a third preset weight coefficient to obtain a third product value; multiplying the fourth SOC estimation value by a fourth preset weight coefficient to obtain a fourth product value; adding the third product value and the fourth product value to obtain the SOC calibration value; the sum of the third preset weight coefficient and the fourth preset weight coefficient is 1; the third preset weight coefficient and the fourth preset weight coefficient are determined according to the aging degree of the battery and / or the current condition of the battery.

3. A method of correcting a state of charge of a battery, characterized by, The method comprises: obtaining the real-time operation parameters of the battery; the real-time operation parameters of the battery include the real-time temperature, the real-time voltage and the real-time current of the battery; inputting the real-time temperature, the real-time voltage and the real-time current of the battery into the pre-constructed lightweight estimation model to obtain the first SOC estimation value; integrating the real-time current of the battery to obtain the second SOC estimation value; obtaining the SOC initial value according to the first SOC estimation value and the second SOC estimation value; uploading the SOC initial value and the data segment containing the real-time operation parameters of the battery to the cloud end when it is identified that the operation condition of the battery belongs to a predetermined critical condition, for the cloud end to determine whether to update the lightweight estimation network and to download the SOC correction value; updating the lightweight estimation network and obtaining the final SOC value of the battery according to the SOC initial value and the SOC correction value when the updated lightweight estimation network and the SOC correction value downloaded by the cloud end are received; taking the SOC initial value as the final SOC value of the battery when the updated lightweight estimation network and the SOC correction value downloaded by the cloud end are not received; wherein the cloud end determines the updated lightweight estimation network and the SOC correction value by using the battery SOC cloud estimation method of any one of claims 1-2.

4. The method of correcting the SOC of a battery according to claim 3, characterized by, If the operation condition of the battery meets at least one of the following conditions: a voltage change rate of the battery exceeds a first predetermined change rate; a temperature change rate of the battery exceeds a second predetermined change rate; the battery undergoes a charge-discharge state switching; determining that an operating condition of the battery belongs to a predetermined critical condition.

5. The method of correcting the SOC of a battery according to claim 3, characterized by, The step of obtaining an initial SOC value according to the first SOC estimation value and the second SOC estimation value comprises: multiplying the first SOC estimation value by a first preset weight coefficient to obtain a first product value; multiplying the second SOC estimation value by a second preset weight coefficient to obtain a second product value; adding the first product value and the second product value to obtain the initial SOC value; a sum of the first preset weight coefficient and the second preset weight coefficient is 1; the first preset weight coefficient and the second preset weight coefficient are determined according to a real-time temperature of the battery.

6. The method of correcting the SOC of a battery according to claim 3, wherein The step of obtaining a final SOC value of the battery according to the initial SOC value and the SOC correction value comprises: adding the initial SOC value and the SOC correction value to obtain the final SOC value of the battery.

7. The method of correcting the SOC of a battery according to claim 3, characterized by, When it is identified that the operating condition of the battery belongs to the predetermined critical condition, uploading the initial SOC value and a data segment containing real-time operating parameters of the battery to the cloud comprises: when the SOC of the battery is greater than or equal to a preset SOC and a discharge rate of the battery is less than or equal to a preset discharge rate, uploading the data segment to the cloud at a first preset frequency; when the SOC of the battery is less than the preset SOC or the discharge rate of the battery is greater than the preset discharge rate, uploading the data segment to the cloud at the first preset frequency; the first preset frequency is less than the second preset frequency.

8. A system for revising a state of charge of a battery, characterized by comprising: comprises: a controller and a vehicle-mounted network terminal connected to each other; the controller is configured to: obtain real-time operating parameters of the battery; the real-time operating parameters of the battery comprise a real-time temperature, a real-time voltage and a real-time current of the battery; input the real-time temperature, the real-time voltage and the real-time current of the battery into a pre-constructed lightweight estimation model to obtain a first SOC estimation value; integrate the real-time current of the battery to obtain a second SOC estimation value; obtain an initial SOC value according to the first SOC estimation value and the second SOC estimation value; when it is identified that the operating condition of the battery belongs to the predetermined critical condition, upload the initial SOC value and a data segment containing the real-time operating parameters of the battery to the cloud through the vehicle-mounted network terminal, for the cloud to determine whether to perform lightweight estimation network updating and issue a SOC correction value; when the updated lightweight estimation network and the SOC correction value issued by the cloud are received through the vehicle-mounted network terminal, perform lightweight estimation network updating, and obtain a final SOC value of the battery according to the initial SOC value and the SOC correction value; when the updated lightweight estimation network and the SOC correction value issued by the cloud are not received through the vehicle-mounted network terminal, take the initial SOC value as the final SOC value of the battery; wherein the cloud determines the updated lightweight estimation network and the SOC correction value by using the battery SOC cloud estimation method of any one of claims 1-3.

9. A cloud-based estimation device for battery SOC, characterized in that, comprises: The acquisition module is configured to acquire an SOC initial value uploaded when it is identified that the operation condition of the battery belongs to a predetermined critical condition according to a battery voltage change rate, a battery temperature change rate, and / or a battery charge-discharge state, and a data segment containing real-time operation parameters of the battery; The first estimation module is configured to obtain a third SOC estimation value by running a first estimation model according to the data segment containing the real-time operation parameters of the battery; the first estimation model is an electrochemical model; The second estimation module is configured to obtain a fourth SOC estimation value by running a second estimation model according to the data segment containing the real-time operation parameters of the battery; The second estimation model is a deep reinforcement learning model; The SOC calibration value estimation module is configured to obtain an SOC calibration value according to the third SOC estimation value and the fourth SOC estimation value; The model update judgment module is configured to determine whether to update the lightweight estimation model at the vehicle end according to the SOC calibration value and the SOC initial value; The model update module is configured to update the lightweight estimation model by using the SOC calibration value and the data segment containing the real-time operation parameters of the battery when it is determined that the lightweight estimation model at the vehicle end needs to be updated; The delivery module is configured to deliver the updated lightweight estimation network to the vehicle end, and deliver a difference between the SOC calibration value and the SOC initial value as an SOC correction value to the vehicle end.

10. A vehicle characterized by comprising: The correction system of the battery SOC includes the system as claimed in claim 8. The correction system of the battery SOC includes the system as claimed in claim 8.

Images