Battery health state prediction method, cloud platform, readable storage medium and program product

Abstract

The invention relates to a battery health state prediction method, a cloud platform, a readable storage medium and a program product, and the method comprises the steps: obtaining the operation condition data of a target battery on a vehicle, carrying out the data screening according to the operation condition data, and obtaining the target charging data in a preset time period; performing health state estimation on the target battery according to the target charging data by using a Kalman filtering model, and determining initial health states of the target battery at different temperatures within a preset time period; acquiring temperature data of the target battery in a charging state within a preset time period from the target charging data; and performing unified temperature correction on the initial health states at different temperatures according to the preset temperature data and the temperature data, and determining the health state of the target battery at each time point at the same temperature. According to the method, the calculation of the initial health state of the target battery can be more focused on the change of key parameters in the charging process, and the prediction precision of the health state of the battery is improved.

Description

Technical Field

[0001] This application relates to the field of lithium-ion battery technology, and in particular to a battery health state prediction method, a cloud platform, a readable storage medium, and a program product. Background Technology

[0002] Lithium-ion batteries are highly favored in the electrochemical energy storage and electric vehicle industries due to their high energy density, long cycle life, and low cost. However, the lifespan of lithium-ion batteries gradually shortens with use, resulting in a decline in vehicle health. Therefore, accurate online health status estimation and prediction of lithium-ion batteries are crucial for intelligent management in electric vehicles, and are also of great significance for ensuring the power, economy, and safety of electric vehicle operation.

[0003] Currently, battery models are mainly used to predict the online health status of lithium-ion batteries. However, these models suffer from problems such as complex parameter identification and high computational cost. Therefore, improving the accuracy of predicting the health status of lithium-ion batteries has become a core issue that urgently needs to be addressed. Summary of the Invention

[0004] Therefore, it is necessary to provide a battery health status prediction method, cloud platform, readable storage medium, and program product to address the above-mentioned technical problems.

[0005] In a first aspect, this application provides a battery health state prediction method, characterized in that it is applied to an in-vehicle terminal and includes:

[0006] Acquire the operating condition data of the target battery on the vehicle, and filter the data based on the operating condition data to obtain the target charging data within a preset time period;

[0007] The health status of the target battery is estimated based on the target charging data using a Kalman filter model, and the initial health status of the target battery at different temperatures within a preset time period is determined.

[0008] Obtain the temperature data of the target battery under charging conditions within a preset time period from the target charging data;

[0009] Temperature data is sent to the cloud platform to instruct the cloud platform to perform unified temperature correction on the initial health status at different temperatures based on preset temperature data and temperature data, and to determine the health status of the target battery at various time points under the same temperature.

[0010] Receive and display the health status of the target battery at various time points under the same temperature.

[0011] In one embodiment, a Kalman filter model is used to estimate the health status of the target battery based on the target charging data, determining the initial health status of the target battery at different temperatures within a preset time period, including:

[0012] The input noise and output noise of the target charging data are quantized using a Kalman filter model to obtain the quantized target charging data.

[0013] The health status of the target battery is estimated based on the quantized target charging data using a Kalman filter model, and the initial health status of the target battery at different temperatures within a preset time period is determined.

[0014] In one embodiment, the target charging data is quantized using a Kalman filter model to quantize the input noise and output noise, resulting in quantized target charging data, including:

[0015] The parameters of the input noise are determined based on the accuracy of the sensors used to collect the operating data.

[0016] The Kalman filter model is configured based on the parameters of the input noise and the parameters of the output noise, and the configured Kalman filter model is used to quantize the input noise and output noise of the target charging data to obtain the quantized target charging data.

[0017] Secondly, this application also provides a battery health state prediction method, characterized in that it is applied to a cloud platform and includes:

[0018] Receive temperature data sent by the vehicle terminal; the temperature data is the temperature data of the target battery in the vehicle during the charging state within a preset time period.

[0019] Based on preset temperature data and temperature data, the initial health status at different temperatures is uniformly corrected, and the health status of the target battery at each time point under the same temperature is determined; the initial health status is estimated by the vehicle terminal using a Kalman filter model to estimate the health status of the target battery on the vehicle.

[0020] The health status of the target battery at various time points under the same temperature is sent to the vehicle terminal, so that the vehicle terminal can display the health status of the target battery at various time points under the same temperature.

[0021] In one embodiment, it further includes:

[0022] The health status data of the target battery at various time points under the same temperature are input into a trained long short-term memory neural network model to predict the health status and output the prediction results. The prediction results include the health degradation curve of the target battery and the vehicle mileage corresponding to the target battery reaching 80% health status.

[0023] In one embodiment, it further includes:

[0024] The health status data of the target battery at various time points under the same temperature are filtered, and the health status data corresponding to the most recent time point is selected as the target health status data.

[0025] The health status data of the target battery at various time points under the same temperature are input into a trained long short-term memory neural network model to predict the health status, and the prediction results are output, including:

[0026] The target health status data is input into a trained long short-term memory neural network model to predict the health status and output the prediction results.

[0027] In one embodiment, a unified temperature correction is performed on the initial health state at different temperatures based on preset temperature data and temperature data to determine the health state of the target battery at various time points under the same temperature, including:

[0028] The average temperature data is obtained by averaging the temperature data.

[0029] Determine the preset temperature data based on the calibration requirements;

[0030] The average temperature data and preset temperature data are substituted into the temperature correction model for calculation to determine the health status of the target battery at various time points under the same temperature.

[0031] Thirdly, this application also provides a cloud platform, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0032] Receive temperature data sent by the vehicle terminal; the temperature data is the temperature data of the target battery in the vehicle during the charging state within a preset time period.

[0033] Based on preset temperature data and temperature data, the initial health status at different temperatures is uniformly corrected, and the health status of the target battery at each time point under the same temperature is determined; the initial health status is estimated by the vehicle terminal using a Kalman filter model to estimate the health status of the target battery on the vehicle.

[0034] The health status of the target battery at various time points under the same temperature is sent to the vehicle terminal, so that the vehicle terminal can display the health status of the target battery at various time points under the same temperature.

[0035] Fourthly, this application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, performs the following steps:

[0036] Receive temperature data sent by the vehicle terminal; the temperature data is the temperature data of the target battery in the vehicle during the charging state within a preset time period.

[0037] Based on preset temperature data and temperature data, the initial health status at different temperatures is uniformly corrected, and the health status of the target battery at each time point under the same temperature is determined; the initial health status is estimated by the vehicle terminal using a Kalman filter model to estimate the health status of the target battery on the vehicle.

[0038] The health status of the target battery at various time points under the same temperature is sent to the vehicle terminal, so that the vehicle terminal can display the health status of the target battery at various time points under the same temperature.

[0039] Fifthly, this application also provides a computer program product, including a computer program that, when executed by a processor, performs the following steps:

[0040] Receive temperature data sent by the vehicle terminal; the temperature data is the temperature data of the target battery in the vehicle during the charging state within a preset time period.

[0041] Based on preset temperature data and temperature data, the initial health status at different temperatures is uniformly corrected, and the health status of the target battery at each time point under the same temperature is determined; the initial health status is estimated by the vehicle terminal using a Kalman filter model to estimate the health status of the target battery on the vehicle.

[0042] The health status of the target battery at various time points under the same temperature is sent to the vehicle terminal, so that the vehicle terminal can display the health status of the target battery at various time points under the same temperature.

[0043] The aforementioned battery health state prediction method, cloud platform, readable storage medium, and program product acquire operating condition data of the target battery in the vehicle, and filter the data based on the operating condition data to obtain target charging data within a preset time period. Using a Kalman filter model, the target battery's health state is estimated based on the target charging data, determining the initial health state of the target battery at different temperatures within the preset time period. Temperature data of the target battery under charging conditions within the preset time period is obtained from the target charging data. The temperature data is sent to the cloud platform, instructing the platform to perform unified temperature correction on the initial health state at different temperatures based on the preset temperature data and the temperature data, and to determine the health state of the target battery at each time point under the same temperature. The method receives and displays the health state of the target battery at each time point under the same temperature. By analyzing the target charging data using a Kalman filter model and filtering the operating condition data within the preset time period, the method effectively reduces the interference of non-charging conditions (such as discharging and resting) on ​​the health state estimation, making the calculation of the target battery's initial health state more focused on the changes in key parameters during the charging process, thus improving the prediction accuracy of the battery health state. At the same time, temperature correction is performed for different temperatures to avoid prediction deviations caused by temperature fluctuations, further improving the accuracy of predicting battery health status. Attached Figure Description

[0044] To more clearly illustrate the technical solutions in the embodiments of this application or related technologies, the drawings used in the description of the embodiments of this application or related technologies will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0045] Figure 1 This is a diagram illustrating the application environment of a battery health state prediction method in one embodiment.

[0046] Figure 2 This is one of the flowcharts illustrating a battery health status prediction method in one embodiment;

[0047] Figure 3 This is a second flowchart illustrating a battery health status prediction method in one embodiment;

[0048] Figure 4 This is the third flowchart of a battery health status prediction method in one embodiment;

[0049] Figure 5 This is the fourth flowchart of a battery health status prediction method in one embodiment;

[0050] Figure 6This is the fifth flowchart of a battery health status prediction method in one embodiment;

[0051] Figure 7 This is a flowchart of a battery health status prediction method in one embodiment, number six.

[0052] Figure 8 This is a graph showing the predicted future SOH of the battery in one embodiment;

[0053] Figure 9 Here is a graph showing the prediction error of future SOH for electric vehicles in one embodiment;

[0054] Figure 10 This is the seventh flowchart of a battery health status prediction method in one embodiment;

[0055] Figure 11 This is the eighth flowchart of a battery health status prediction method in one embodiment;

[0056] Figure 12 This is the ninth flowchart of a battery health status prediction method in one embodiment. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0058] It should be noted that the terms "first," "second," etc., used in this application can be used to describe various elements, but these elements are not limited by these terms. These terms are only used to distinguish the first element from the second element. The terms "comprising" and "having," and any variations thereof, used in this application, are intended to cover non-exclusive inclusion. The term "multiple" used in this application refers to two or more. The term "and / or" used in this application refers to one of the embodiments, or any combination of multiple embodiments.

[0059] Lithium-ion batteries are highly favored in the electrochemical energy storage and electric vehicle industries due to their high energy density, long cycle life, and low cost. However, the lifespan of lithium-ion batteries gradually shortens with use, resulting in a decline in vehicle health. Therefore, accurate online health status estimation and prediction of lithium-ion batteries are crucial for intelligent management in electric vehicles, and are also of great significance for ensuring the power, economy, and safety of electric vehicle operation.

[0060] Currently, battery models are mainly used to predict the online health status of lithium-ion batteries. However, these models suffer from problems such as complex parameter identification and high computational cost. Therefore, improving the accuracy of predicting the health status of lithium-ion batteries has become a core issue that urgently needs to be addressed.

[0061] In view of the above-mentioned technical problems, this application provides a battery health status prediction method, and the following embodiments will specifically illustrate the battery health status prediction method.

[0062] The battery health status prediction method provided in this application embodiment can be applied to, for example... Figure 1 The internal structure diagram of the battery health status prediction system shown can be as follows: Figure 1 As shown.

[0063] The battery health status prediction system includes a battery management platform 101, an in-vehicle terminal 102, a cloud platform 103, and a display device 104. The in-vehicle terminal 102 is connected to both the battery management platform 101 and the cloud platform 103. The in-vehicle terminal 102 processes the operating condition data obtained from the battery management platform 101 and sends the processed data to the cloud platform 103 for further processing. The cloud platform 103 performs final processing to obtain the health status of the target battery, and then sends the target battery's health status to the display device 104 for display.

[0064] In one exemplary embodiment, such as Figure 2 As shown, a battery health state prediction method is provided, which can be applied to... Figure 1 Taking the vehicle-mounted terminal as an example, the explanation includes:

[0065] S201: Obtain the operating condition data of the target battery on the vehicle, and filter the data based on the operating condition data to obtain the target charging data within a preset time period.

[0066] Among them, the operating condition data refers to a series of detailed parameters of the target battery, such as the battery state of charge at the start of charging, the battery state of charge (SOC) at the end of charging, the cumulative charging amount, the charge / discharge status flag, voltage, current, temperature (including battery core temperature and ambient temperature), and data that can reflect the charging status of the target battery.

[0067] In this embodiment, the battery of any one of multiple vehicles is selected as the target battery. The Battery Management System (BMS) collects the operating condition data of the target battery in real time at a preset frequency. After collection, the operating condition data undergoes multi-dimensional data cleaning. For example, the vehicle terminal can remove abnormal values ​​of parameters such as current and voltage by setting a threshold (e.g., a current mutation rate greater than 20% is considered abnormal) or delete duplicate records using timestamp deduplication. Missing data is then filled using interpolation. Abnormal values ​​(e.g., current mutations) or duplicate values ​​(e.g., redundant data) are removed from the data. For example, the vehicle terminal can remove abnormal values ​​by setting a threshold (e.g., the range of valid data). Subsequently, the vehicle terminal filters charging data from the cleaned operating condition data: First, the vehicle terminal identifies the complete charging segment of the target battery through the charging / discharging status flag. Then, it further filters out valid charging segments where the battery charge at the start of charging is less than 40% and the battery charge at the end of charging is greater than or equal to 80%. If a charging data segment does not meet the above filtering conditions, it is discarded. Finally, by setting the time period according to user needs (such as 7 days), the vehicle terminal can automatically extract all target charging data that meet the filtering criteria within that time period, providing data support for subsequent health assessment of the target battery.

[0068] S202, using a Kalman filter model to estimate the health status of the target battery based on the target charging data, and determining the initial health status of the target battery at different temperatures within a preset time period.

[0069] In this embodiment, the vehicle terminal receives target charging data within a preset time period that meets the screening criteria (effective charging segments where the battery charge at the start of charging is less than 40% and the battery charge at the end of charging is greater than or equal to 80%), and inputs the target charging data into a Kalman filter model. Before inputting the target charging data into the Kalman filter model, the vehicle terminal first calculates the current battery capacity at the corresponding time using a two-point method for each effective data segment, and maps the current battery capacity at the corresponding time to the specific collection time to generate an initial health state set. Subsequently, the initial health state set is input into the Kalman filter model. First, a predicted value of the health state is generated by combining the prediction equation with the prior law of battery capacity decay (such as the number of cycles and the influence of temperature). Then, the observed health value of the target battery is fused with the predicted value using an update equation, and the filtering noise is dynamically adjusted to correct the error. Finally, the filtered SOH value of the target battery is output. Finally, by associating the collection time and ambient temperature data corresponding to each SOH value, the vehicle terminal can determine the initial health state of the target battery at different temperatures within the preset time period.

[0070] S203: Obtain the temperature data of the target battery during the charging state within a preset time period from the target charging data.

[0071] In this embodiment, the vehicle-mounted terminal can collect temperature data of the target battery during charging using temperature sensors in the vehicle, and transmit the collected temperature data to the BMS management system. When predicting and estimating the health status of the target battery, the vehicle-mounted terminal can obtain the temperature data of the target battery during charging within a preset time period from the target battery's charging data.

[0072] S204, send temperature data to the cloud platform to instruct the cloud platform to perform unified temperature correction on the initial health status at different temperatures based on preset temperature data and temperature data, and determine the health status of the target battery at various time points under the same temperature.

[0073] In this embodiment, since temperature affects the health status of the target battery—for example, the battery's discharge capacity decreases significantly at low temperatures and increases at high temperatures—directly comparing the battery's health status at different temperatures would lead to inaccurate assessments. Therefore, the vehicle terminal can send the target battery's temperature data during charging to a cloud platform. The cloud platform can then perform a unified temperature correction on the initial health status at different temperatures based on preset temperatures and temperature data, converting the initial health status of the target battery at different temperatures into the initial health status of the target battery under the same preset temperature data. This allows the health status of the target battery at various time points within the same temperature range to be determined.

[0074] S205 receives and displays the health status of the target battery at various time points under the same temperature.

[0075] In this embodiment of the application, after the cloud platform completes the detection of the health status of the target battery at various time points under the same temperature based on the data information sent by the vehicle terminal, the cloud platform sends the health status of the target battery at various time points under the same temperature back to the vehicle terminal, which then displays it and can provide suggestions to the user based on the health status.

[0076] The aforementioned battery health state prediction method, cloud platform, readable storage medium, and program product acquire operating condition data of the target battery in the vehicle, and filter the data based on the operating condition data to obtain target charging data within a preset time period. Using a Kalman filter model, the target battery's health state is estimated based on the target charging data, determining the initial health state of the target battery at different temperatures within the preset time period. Temperature data of the target battery under charging conditions within the preset time period is obtained from the target charging data. The temperature data is sent to the cloud platform, instructing the platform to perform unified temperature correction on the initial health state at different temperatures based on the preset temperature data and the temperature data, and to determine the health state of the target battery at each time point under the same temperature. The method receives and displays the health state of the target battery at each time point under the same temperature. By analyzing the target charging data using a Kalman filter model and filtering the operating condition data within the preset time period, the method effectively reduces the interference of non-charging conditions (such as discharging and resting) on ​​the health state estimation, making the calculation of the target battery's initial health state more focused on the changes in key parameters during the charging process, thus improving the prediction accuracy of the battery health state. At the same time, temperature correction is performed for different temperatures to avoid prediction deviations caused by temperature fluctuations, further improving the accuracy of predicting battery health status.

[0077] In an exemplary embodiment, the step S202 described above, "using a Kalman filter model to estimate the health status of the target battery based on the target charging data, and determining the initial health status of the target battery at different temperatures within a preset time period," is as follows: Figure 3 As shown, it includes:

[0078] S301 uses a Kalman filter model to quantize the input and output noise of the target charging data, obtaining the quantized target charging data.

[0079] The input noise is the process noise; the output noise is the measurement noise.

[0080] In this embodiment, when the vehicle-mounted terminal uses a Kalman filter model to predict the initial health state of the target battery, the Kalman filter model will have input noise and output noise. To more accurately predict the initial health state of the target battery, the vehicle-mounted terminal quantifies the uncertainty caused by the input and output noise to the Kalman filter model. The input noise can be represented by parameters. The output noise can be represented by parameters. To express, and The smaller the accuracy, the higher the output noise of the Kalman filter model. The larger; and The higher the accuracy, the lower the output noise of the Kalman filter model. The smaller. and They are independent of each other. Regarding input noise... The calculation can be expressed by relation (1), which is shown below:

[0081] (1);

[0082] in, Let V be the process noise variance, and let V be the uncertainty of the current capacity.

[0083] For output noise The calculation can be expressed by relation (2), which is shown below:

[0084] (2);

[0085] in, This represents the measurement noise floor, i.e., the error in measuring electrical quantity. The noise weighting coefficient can be calculated using equation (3), which is shown below:

[0086] (3);

[0087] in, and For the belonging function, and ;for The choice of the attribution function needs to meet three requirements: First, the interval partitioning should be such that... The range from 0% to 100% is divided into several smaller intervals, and weights are determined based on these intervals, with an auxiliary reference value set at 40%; secondly... The value varies between 0% and 40%. Monotonically increasing, and The maximum value is taken at 40% of the auxiliary reference value; third, The attribution function must be a convex function. For The choice of the attribution function also needs to meet three requirements: First, the interval partitioning should be such that... The range from 0% to 100% is divided into several smaller intervals, and weights are determined based on these intervals, with an auxiliary reference value set at 98%; secondly... Value follows Monotonically decreasing, and The minimum value is taken at 98% of the auxiliary reference value; third, The attribution function must be a convex function.

[0088] S302 uses a Kalman filter model to estimate the health status of the target battery based on the quantized target charging data, and determines the initial health status of the target battery at different temperatures within a preset time period.

[0089] In this embodiment of the application, the vehicle terminal can use the Kalman filter model to estimate the health status of the target battery based on the quantized target charging data. The discrete state equation and system output equation of the Kalman filter model are shown in relation (4) and relation (5).

[0090] (4);

[0091] (5);

[0092] in, Indicates the current capacity to be estimated. As a system state vector, it cannot be directly obtained through measurement; The initial charge level serves as a measurable system output. and These are input noise (process noise) and output noise (observation noise), which cannot be obtained through measurement.

[0093] The iterative formulas of the Kalman filter model are state estimation time update, estimation variance time update, Kalman gain update, state estimation measurement update, and estimation variance measurement update, respectively. Among them, the state estimation time update can be represented by relation (6), the estimation variance time update can be represented by relation (7), the Kalman gain update can be represented by relation (8), the state estimation measurement update can be represented by relation (9), and the error estimation variance measurement update can be represented by relation (10). The relationship from relation (6) to relation (10) is as follows:

[0094] (6);

[0095] (7);

[0096] (8);

[0097] (9);

[0098] (10);

[0099] in, This represents the optimal posterior state estimate after the previous state correction; This means that, under the assumption that the state of battery health (SOH) changes slowly between adjacent sampling points, the best prediction result of the previous state is used as the prior state estimate of the current state. It is the optimal posterior state estimate after the current state is corrected. U represents the estimated variance. This represents the prior state estimate of the current state. The estimated variance, This represents the estimated variance of the posterior estimate of the previous state; Represents the posterior state estimate of the current state. The estimated variance. , These represent the input noise respectively. variance and output noise The variance. This represents the Kalman gain of the current state. This represents the actual observed value obtained from the current state, which is the SOH calculated using the two-point method. In each iteration, the Kalman filter estimates the state vector and the estimated variance twice. , It is based on the previous moment. , Calculated , The output of the measurement system is then measured based on the current time. , It was estimated.

[0100] In an exemplary embodiment, the above-mentioned S301, "quantizing the input noise and output noise of the target charging data using a Kalman filter model to obtain quantized target charging data," is as follows: Figure 4 As shown, it includes:

[0101] S401 determines the input noise parameters based on the accuracy of the sensors used to collect operating condition data.

[0102] In this embodiment of the application, the input noise The parameter 'a' represents the process noise variance, indicating the uncertainty of the current capacity. The value of 'a' depends on the accuracy of the target battery's sensors, such as the accuracy of the current sensor and the accuracy of the voltage sensor.

[0103] S402: Configure the Kalman filter model according to the parameters of the input noise and the parameters of the output noise, and use the configured Kalman filter model to quantize the input noise and output noise of the target charging data to obtain the quantized target charging data.

[0104] In this embodiment of the application, the method involves configuring a Kalman filter model based on the parameters of the input noise and the parameters of the output noise, and using the configured Kalman filter model to quantize the input noise and output noise of the target charging data to obtain the quantized target charging data. This method is similar to the aforementioned... Figures 2-3 The methods described in any implementation are basically the same, and for details, please refer to the foregoing explanation, which will not be repeated here.

[0105] In summary, based on all the above embodiments, a method for predicting battery health status is also provided, such as... Figure 5 As shown, applied to an in-vehicle terminal, the method includes:

[0106] S501: Obtain the operating condition data of the target battery on the vehicle, and filter the data based on the operating condition data to obtain the target charging data within a preset time period.

[0107] S502 determines the input noise parameters based on the accuracy of the sensors used to collect operating condition data;

[0108] S503 configures the Kalman filter model according to the parameters of the input noise and the parameters of the output noise, and uses the configured Kalman filter model to quantize the input noise and output noise of the target charging data to obtain the quantized target charging data.

[0109] S504 uses a Kalman filter model to estimate the health status of the target battery based on the quantized target charging data, and determines the initial health status of the target battery at different temperatures within a preset time period.

[0110] S505: Obtain the temperature data of the target battery under charging state within a preset time period from the target charging data;

[0111] S506 sends temperature data to the cloud platform to instruct the cloud platform to perform unified temperature correction on the initial health status at different temperatures based on preset temperature data and temperature data, and to determine the health status of the target battery at various time points under the same temperature.

[0112] S507 receives and displays the health status of the target battery at various time points under the same temperature.

[0113] The methods described in each of the above steps have been described in the foregoing embodiments. For details, please refer to the foregoing descriptions. They will not be repeated here.

[0114] In one exemplary embodiment, based on the battery health state prediction method described in any of the foregoing embodiments, another battery health state prediction method is also provided, applied to a cloud platform, such as... Figure 6 As shown, it includes:

[0115] S101 receives temperature data sent by the vehicle-mounted terminal.

[0116] The temperature data refers to the temperature data of the target battery on the vehicle during the charging process within a preset time period.

[0117] In this embodiment of the application, the cloud platform receives temperature data of the target battery in the vehicle under charging state from the vehicle terminal within a preset time period (such as 7 days) via the vehicle network communication protocol.

[0118] S102, perform unified temperature correction on the initial health status at different temperatures based on preset temperature data and temperature data, and determine the health status of the target battery at each time point under the same temperature.

[0119] The initial health state is estimated by the vehicle terminal using a Kalman filter model to assess the health state of the target battery on the vehicle.

[0120] In this embodiment, the cloud platform receives temperature data of the target battery in the vehicle under charging state from the vehicle terminal within a preset time period (e.g., 7 days) via the vehicle network communication protocol. Then, it combines the initial health state obtained by the vehicle terminal using the Kalman filter model to estimate the health state of the target battery. Finally, it performs a unified temperature correction on the initial health state at different temperatures through a temperature correction model, and finally determines the health state of the target battery at each time point under the same temperature.

[0121] S103, send the health status of the target battery at various time points under the same temperature to the vehicle terminal, so as to instruct the vehicle terminal to display the health status of the target battery at various time points under the same temperature.

[0122] In this embodiment, after obtaining the health status of the target battery at various time points under the same temperature, the cloud platform sends the health status of the battery at various time points under the same temperature to the vehicle terminal through the vehicle network communication protocol. The vehicle terminal can display the health status of the target battery at various time points under the same temperature and provide suggestions to the user based on the health status.

[0123] In one exemplary embodiment, Figure 6 The specific implementation methods of S101 to S102 in the scheme described in the embodiment are as follows: Figure 7 As shown, it also includes:

[0124] S104: Input the health status data of the target battery at various time points under the same temperature into the trained long short-term memory neural network model to predict the health status and output the prediction results.

[0125] The prediction results include the target battery's health degradation curve and the vehicle's mileage corresponding to the target battery reaching 80% health status.

[0126] In this embodiment, before inputting the health status data of the target battery at various time points under the same temperature into the trained Long Short-Term Memory (LSTM) neural network model, the cloud platform first inputs the health status data of the target battery at various time points under the same temperature into a fuzzy Kalman filter algorithm for SOH filtering, thereby eliminating noise. Subsequently, the cloud platform uses the historical SOH values ​​of the target battery to train the LSTM neural network model, with training data and validation data accounting for 80% and 20% of the entire dataset, respectively. Furthermore, the cloud platform needs to set the initial parameters of the LSTM neural network model, such as 32 neurons, 32 batch size, 100 iterations, a learning rate of 0.1, and a sliding window size of 10. The trained LSTM neural network model can map the SOH value of the target battery to the corresponding vehicle mileage and output prediction results, for example, such as... Figure 4 As shown in the figure, the blue line represents the historical data input to the model, the red dashed line represents the model's predicted SOH curve, and the orange curve represents the vehicle's actual SOH curve. The figure shows that even with a relatively dynamic SOH estimation result, the Long Short-Term Memory (LSTM) neural network model, based on long-term data, can still accurately reflect the SOH decay trend, demonstrating strong generalization ability. It can also accurately predict the mileage at which 80% SOH is achieved, with a maximum error of 0.6% SOH. Figure 8 and Figure 9 As shown.

[0127] In an exemplary embodiment, the step S104 above, "inputting the health status data of the target battery at various time points under the same temperature into a trained long short-term memory neural network model to predict the health status and outputting the prediction result," is as follows: Figure 10 As shown, it includes:

[0128] S201, the health status data of the target battery at various time points under the same temperature are filtered, and the health status data corresponding to the most recent time point is selected as the target health status data.

[0129] The most recent time point can be filtered by the user according to the time rules.

[0130] In this embodiment, during battery health status monitoring and evaluation, the cloud platform can filter the State of Health (SOH) values ​​of the target battery at various time points under the same temperature based on user-preset time filtering rules (such as the last 10 days). For example, if the target battery has recorded key indicators such as charging capacity and discharging efficiency daily under a constant temperature of 25°C for the past 10 days, the cloud platform will automatically filter out historical data older than 10 days and retain only the SOH values ​​of the battery at various time points within the last 10 days of charging as the core basis for analyzing the target health status.

[0131] S202, input the target health status data into the trained long short-term memory neural network model to predict the health status and output the prediction result.

[0132] In this embodiment of the application, after obtaining the target health status data, the cloud platform inputs it into a trained long short-term memory neural network model to predict the health status of the target battery and outputs the prediction results. The prediction results include the health degradation curve of the target battery's SOH and the vehicle mileage corresponding to the target battery's health status reaching 80%.

[0133] When training a Long Short-Term Memory (LSTM) neural network model on a cloud platform, the first step is to use the forget gate within the LTM model to determine which information will be saved or discarded. This decision is made by the forget gate. sigmoid activation function Complete. The input to the forget gate is the current input vector. Hidden state from the previous moment The forget gate is calculated and outputs a value between 0 and 1. An output of 1 indicates that the state value is completely retained, while 0 indicates that the value is completely discarded. The calculation method of the forget gate is shown in relation (11);

[0134] (11);

[0135] in, Indicates at time step The output vector of the forget gate; This represents the input vector at the current time step; This indicates the hidden state in the previous moment; Indicates processing the current input The weight matrix output by the forget gate; This indicates the processing of the hidden state from the previous time step. The weight matrix output by the forget gate; This represents the bias term for the forget gate.

[0136] Subsequently, the cloud platform uses the input gate in the Long Short-Term Memory neural network model to decide whether to update the state using the current input. Specifically, this step includes two parts: forming a new candidate state vector using the hyperbolic tangent activation function tanh. Input gate sigmoid activation function Determine the new state vector Which information can be updated to the cell state? The calculation formulas for these two parts are shown in relation (12) and relation (13);

[0137] (12);

[0138] (13);

[0139] in, Indicates at time step The output vector of the input gate; Indicates processing the current input The weight matrix to the input gate output vector; This indicates the processing of the hidden state from the previous time step. The weight matrix to the input gate output vector; This represents the bias term of the input gate; Indicates at time step The candidate state vector; Indicates processing the current input The weight matrix to the candidate state vector; This indicates the processing of the hidden state from the previous time step. The weight matrix to the candidate state vector; This represents the bias term of the candidate state vector.

[0140] By combining relation (12) and relation (13), the internal unit state of the previous time step can be obtained. Update to the current state It can be represented by relation (14), which is shown below:

[0141] (14);

[0142] in, Indicates the current state of the cell; This indicates the internal state of the cell at the previous moment; Indicates at time step The output vector of the input gate; Indicates at time step The candidate state vector; Indicates at time step The output vector of the forget gate; This represents the multiplication of the principal elements of a vector, also known as the Hadamard product.

[0143] Finally, from the output gate Decide which information to extract from the cell status Transform into the current hidden layer state Output. Specifically, internal cell states. The value is transformed to the range of -1 to 1 using a hyperbolic tangent activation function tanh layer, and then combined with the sigmoid activation function. Multiplying the output vectors of the activated output gates can transfer the information of the cell state to the hidden layer. The calculation formula can be expressed by relation (15) and relation (16), which are shown below:

[0144] (15);

[0145] (16);

[0146] in, Indicates at time step The output vector of the output gate; Indicates processing the current input The weight matrix to the output vector of the output gate; This indicates the processing of the hidden state from the previous time step. The weight matrix leading to the output vector of the output gate; This represents the bias term of the input gate; Indicates at time step The hidden state.

[0147] In an exemplary embodiment, the above-mentioned S102 step of "performing unified temperature correction on the initial health state at different temperatures based on preset temperature data and temperature data, and determining the health state of the target battery at various time points at the same temperature" is as follows: Figure 11 As shown, it includes:

[0148] S301, calculates the average temperature data by averaging the temperature data at different time points within a preset time period under the charging state.

[0149] In this embodiment, the cloud platform can obtain average temperature data by averaging the temperature data at different charging points within a preset time period. For example, if the cloud platform collects temperature data of 25 degrees, 26 degrees, 27 degrees, 28 degrees, 29 degrees, 30 degrees, and 31 degrees within 10 minutes, then the average temperature is (25+26+27+28+29+30+31)÷7=28 degrees.

[0150] S302, determine the preset temperature data according to the calibration requirements.

[0151] In this embodiment, because the target battery health status data calculated by the cloud platform is obtained at different temperatures, and the target battery health status data is different at different temperatures, the cloud platform analyzes the temperature characteristics of the battery and determines preset temperature data, such as 25 degrees Celsius, according to the correction requirements. This ensures the accuracy and consistency of subsequent analysis of the target battery's health status.

[0152] S303 uses the average temperature data and preset temperature data to calculate the health status of the target battery at various time points under the same temperature by substituting them into the temperature correction model.

[0153] In this embodiment, the cloud platform first processes the data of the target charging segment using the two-point method to calculate the original state of health (SOH) observation value of the target battery and its corresponding original current capacity observation value for the corresponding time period. Then, the original observation value, which contains measurement noise and is directly calculated by the two-point method, is used as the observation input. The input is fed into a pre-built Kalman filter model. Then, the Kalman filter model, based on the system's dynamic characteristics and the statistical characteristics of observation noise, applies the observed input... Optimal estimation and smoothing are performed. Finally, the cloud platform can obtain not only the filtered, more accurate SOH value of the target battery from the output of the Kalman filter, but also the filtered, more accurate current capacity estimate of the target battery. The two-point method can be represented by relation (17), which is shown below:

[0154] (17);

[0155] in, Indicates the initial capacity of the target battery; This indicates the current capacity of the target battery after filtering.

[0156] The cloud platform utilizes a temperature correction model to filter the SOH results. By uniformly correcting the results to a temperature of 25℃, the health status of the target battery at various time points under the same temperature can be obtained. The temperature correction model can be represented by the relationship (18), which is shown below:

[0157] (18);

[0158] in, This indicates the SOH result after temperature correction; This represents the average value of all temperature sensors at all times during the i-th charging period; This indicates a temperature of 25℃; This represents the initial health state estimate of the target battery by the Kalman filter model.

[0159] In summary, based on all the above embodiments, a method for predicting battery health status is also provided, such as... Figure 12 As shown, applied to a cloud platform, the method includes:

[0160] S401 receives temperature data sent by the vehicle terminal;

[0161] S402, perform a mean calculation on the temperature data to obtain the average temperature data;

[0162] S403, determine the preset temperature data according to the calibration requirements;

[0163] S404, substitutes the average temperature data and preset temperature data into the temperature correction model for calculation, and determines the health status of the target battery at each time point under the same temperature.

[0164] S405, filter the health status data of the target battery at various time points under the same temperature, and select the health status data corresponding to the most recent time point as the target health status data.

[0165] S406, Input the target health status data into the trained long short-term memory neural network model to predict the health status and output the prediction result;

[0166] S407 sends the prediction results to the vehicle terminal to instruct the vehicle terminal to display the health status of the target battery at various time points under the same temperature.

[0167] The methods described in each of the above steps have been described in the foregoing embodiments. For details, please refer to the foregoing descriptions. They will not be repeated here.

[0168] It should be understood that although the steps in the flowcharts of the embodiments described above are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the flowcharts of the embodiments described above may include multiple steps or multiple stages. These steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least some of the steps or stages in other steps. It is understood that the steps in different embodiments can be freely combined as needed, and all non-contradictory solutions formed by such combinations are within the scope of protection of this application.

[0169] In one exemplary embodiment, a cloud platform is provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to perform the following steps:

[0170] Receive temperature data sent by the vehicle-mounted terminal;

[0171] Based on preset temperature data and temperature data, the initial health status at different temperatures is uniformly corrected, and the health status of the target battery at each time point under the same temperature is determined.

[0172] The health status of the target battery at various time points under the same temperature is sent to the vehicle terminal, so that the vehicle terminal can display the health status of the target battery at various time points under the same temperature.

[0173] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0174] The health status data of the target battery at various time points under the same temperature are input into a trained long short-term memory neural network model to predict the health status and output the prediction results.

[0175] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0176] The health status data of the target battery at various time points under the same temperature are filtered, and the health status data corresponding to the most recent time point is selected as the target health status data.

[0177] The target health status data is input into a trained long short-term memory neural network model to predict the health status and output the prediction results.

[0178] In one embodiment, the processor, when executing a computer program, also performs the following steps:

[0179] The average temperature data is obtained by averaging the temperature data.

[0180] Determine the preset temperature data based on the calibration requirements;

[0181] The average temperature data and preset temperature data are substituted into the temperature correction model for calculation to determine the health status of the target battery at various time points under the same temperature.

[0182] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor:

[0183] Receive temperature data sent by the vehicle-mounted terminal;

[0184] Based on preset temperature data and temperature data, the initial health status at different temperatures is uniformly corrected, and the health status of the target battery at each time point under the same temperature is determined.

[0185] The health status of the target battery at various time points under the same temperature is sent to the vehicle terminal, so that the vehicle terminal can display the health status of the target battery at various time points under the same temperature.

[0186] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0187] The health status data of the target battery at various time points under the same temperature are input into a trained long short-term memory neural network model to predict the health status and output the prediction results.

[0188] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0189] The health status data of the target battery at various time points under the same temperature are filtered, and the health status data corresponding to the most recent time point is selected as the target health status data.

[0190] The target health status data is input into a trained long short-term memory neural network model to predict the health status and output the prediction results.

[0191] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0192] The average temperature data is obtained by averaging the temperature data.

[0193] Determine the preset temperature data based on the calibration requirements;

[0194] The average temperature data and preset temperature data are substituted into the temperature correction model for calculation to determine the health status of the target battery at various time points under the same temperature.

[0195] In one embodiment, a computer program product is provided, including a computer program that, when executed by a processor, performs the following steps:

[0196] Receive temperature data sent by the vehicle-mounted terminal;

[0197] Based on preset temperature data and temperature data, the initial health status at different temperatures is uniformly corrected, and the health status of the target battery at each time point under the same temperature is determined.

[0198] The health status of the target battery at various time points under the same temperature is sent to the vehicle terminal, so that the vehicle terminal can display the health status of the target battery at various time points under the same temperature.

[0199] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0200] The health status data of the target battery at various time points under the same temperature are input into a trained long short-term memory neural network model to predict the health status and output the prediction results.

[0201] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0202] The health status data of the target battery at various time points under the same temperature are filtered, and the health status data corresponding to the most recent time point is selected as the target health status data.

[0203] The target health status data is input into a trained long short-term memory neural network model to predict the health status and output the prediction results.

[0204] In one embodiment, when the computer program is executed by a processor, it also performs the following steps:

[0205] The average temperature data is obtained by averaging the temperature data.

[0206] Determine the preset temperature data based on the calibration requirements;

[0207] The average temperature data and preset temperature data are substituted into the temperature correction model for calculation to determine the health status of the target battery at various time points under the same temperature.

[0208] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the above methods. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile memory and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM). The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, artificial intelligence (AI) processors, etc., and are not limited to these.

[0209] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this application.

[0210] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of this patent application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this application should be determined by the appended claims.

Claims

1. A method for predicting battery health status, characterized in that, Applied to vehicle-mounted terminals, the method includes: The operating condition data of the target battery on the vehicle is obtained, and the data is filtered according to the operating condition data to obtain the target charging data within a preset time period. The health status of the target battery is estimated based on the target charging data using a Kalman filter model, and the initial health status of the target battery at different temperatures within the preset time period is determined. Obtain the temperature data of the target battery in the charging state within the preset time period from the target charging data; The temperature data is sent to the cloud platform to instruct the cloud platform to perform unified temperature correction on the initial health status at different temperatures based on preset temperature data and the temperature data, and to determine the health status of the target battery at each time point under the same temperature. Receive and display the health status of the target battery at various time points under the same temperature.

2. The method according to claim 1, characterized in that, The step of using a Kalman filter model to estimate the health status of the target battery based on the target charging data, and determining the initial health status of the target battery at different temperatures within the preset time period, includes: The target charging data is quantized by using a Kalman filter model to quantize the input noise and output noise, resulting in quantized target charging data. The health status of the target battery is estimated using a Kalman filter model based on the quantized target charging data, and the initial health status of the target battery at different temperatures within the preset time period is determined.

3. The method according to claim 2, characterized in that, The step of quantizing the input and output noise of the target charging data using a Kalman filter model to obtain quantized target charging data includes: The parameters of the input noise are determined based on the accuracy of the sensors used to collect the operating data. The Kalman filter model is configured according to the parameters of the input noise and the parameters of the output noise, and the configured Kalman filter model is used to quantize the input noise and output noise of the target charging data to obtain the quantized target charging data.

4. A method for predicting battery health status, characterized in that, Applied to a cloud platform, the method includes: Receive temperature data sent by the vehicle terminal; the temperature data is the temperature data of the target battery in the vehicle during the charging state within a preset time period. Based on preset temperature data and the temperature data, a unified temperature correction is performed on the initial health status at different temperatures, and the health status of the target battery at each time point under the same temperature is determined; the initial health status is estimated by the vehicle terminal using a Kalman filter model to estimate the health status of the target battery in the vehicle. The health status of the target battery at various time points under the same temperature is sent to the vehicle terminal to instruct the vehicle terminal to display the health status of the target battery at various time points under the same temperature.

5. The method according to claim 4, characterized in that, The method further includes: The health status data of the target battery at various time points under the same temperature are input into a trained long short-term memory neural network model to predict the health status and output the prediction results. The prediction results include the health degradation curve of the target battery and the mileage of the vehicle corresponding to the target battery reaching a health status of 80%.

6. The method according to claim 5, characterized in that, The method further includes: The health status data of the target battery at various time points under the same temperature are filtered, and the health status data corresponding to the most recent time point is selected as the target health status data. The step involves inputting the health status data of the target battery at various time points under the same temperature into a trained long short-term memory neural network model to predict its health status and outputting the prediction results, including: The target health status data is input into a trained long short-term memory neural network model to predict the health status and output the prediction results.

7. The method according to any one of claims 4 to 6, characterized in that, The step of uniformly correcting the initial health status at different temperatures based on preset temperature data and the temperature data, and determining the health status of the target battery at various time points under the same temperature, includes: The temperature data is averaged to obtain the average temperature data. The preset temperature data is determined according to the calibration requirements; The average temperature data and the preset temperature data are substituted into the temperature correction model for calculation to determine the health status of the target battery at various time points under the same temperature.

8. A cloud platform, comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of the method according to any one of claims 4 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 4 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the steps of the method according to any one of claims 4 to 7.

Images