Method, device and equipment for detecting temperature anomaly of power battery and medium

Abstract

The invention provides a power battery temperature anomaly detection method, device and equipment and a medium, and the method comprises the steps: obtaining a current actual temperature value, detected by a temperature sensor in an electric vehicle, of a power battery, and inputting target operation data into a pre-trained temperature prediction model, obtaining a current predicted temperature value for the temperature sensor; based on a weight factor corresponding to the temperature sensor, correcting the difference between the current predicted temperature value and the current actual temperature value to obtain a target predicted difference; and acquiring a plurality of historical predicted temperature values and a plurality of historical actual temperature values in a historical time period in real time through a sliding window algorithm, calculating a mean square error, and judging whether the temperature of the power battery is abnormal or not according to the target predicted difference and the mean square error. According to the method and the device, the reliability and the efficiency of temperature anomaly detection of the power battery are improved.

Description

Technical Field

[0001] The present application relates to the technical field of power battery temperature monitoring, and in particular to a method, device, equipment and medium for detecting abnormal power battery temperature. Background Art

[0002] With the rapid development of new energy vehicles, power batteries, as the core components of new energy vehicles, have a crucial impact on the safety and reliability of the vehicle. The consistency of battery temperature is an important indicator to measure the health of the battery pack.

[0003] At present, most battery temperature monitoring methods rely on data from on-board sensors. This technology includes setting a fixed temperature threshold and analyzing the temperature rise trend. However, this early warning method based on simple trend analysis can only provide a relative comparison between different sampling points and lacks an absolute judgment standard. Therefore, how to improve the accuracy of power battery temperature anomaly detection has become a technical issue that cannot be underestimated. Summary of the invention

[0004] In view of this, the purpose of this application is to provide a method, device, equipment and medium for detecting abnormal power battery temperature, predict the temperature of the power battery through a temperature prediction model, correct the difference between the predicted temperature and the actual temperature through the weight factor corresponding to the temperature sensor, and make an abnormal judgment on the temperature of the power battery in combination with the historical temperature prediction and the historical temperature actual situation. The reliability and efficiency of abnormal power battery temperature detection are improved, thereby improving the accuracy and real-time monitoring of electric vehicle battery temperature, providing more accurate battery health monitoring for electric vehicles.

[0005] In a first aspect, an embodiment of the present application provides a method for detecting abnormal temperature of a power battery, the detection method comprising:

[0006] Acquiring real-time operating data of the electric vehicle, and performing data preprocessing on the real-time operating data to obtain target operating data;

[0007] Acquire the current actual temperature value of the power battery detected by the temperature sensor in the electric vehicle, and input the target operation data into a pre-trained temperature prediction model to obtain the current predicted temperature value for the temperature sensor;

[0008] Based on the weight factor corresponding to the temperature sensor, the difference between the current predicted temperature value and the current actual temperature value is corrected to obtain a target predicted difference; wherein the weight factor corresponding to the temperature sensor is determined by an attribute factor of the temperature sensor;

[0009] Acquire multiple historical predicted temperature values ​​and multiple historical actual temperature values ​​of the power battery in a historical time period in real time through a sliding window algorithm, and calculate a mean square error based on the multiple historical predicted temperature values ​​and the multiple historical actual temperature values;

[0010] It is determined whether the power battery has a temperature abnormality based on the target prediction difference and the mean square error.

[0011] Furthermore, the weight factor corresponding to the temperature sensor is determined by the following steps:

[0012] Acquire the attribute factors of the temperature sensor and the adjustment coefficient corresponding to each attribute factor; wherein the attribute factors include position attribute factors, flow channel attribute factors and impedance attribute factors;

[0013] The product between each attribute factor and the corresponding adjustment coefficient is determined, and each product is summed to obtain the weight factor.

[0014] Furthermore, the method of correcting the difference between the current predicted temperature value and the current actual temperature value based on the weight factor corresponding to the temperature sensor to obtain a target predicted difference includes:

[0015] The product of the difference between the current predicted temperature value and the current actual temperature value and the weight factor is determined as the target predicted difference.

[0016] Further, judging whether the power battery has a temperature abnormality based on the target prediction difference and the mean square error includes:

[0017] Calculating a difference ratio based on the target prediction difference and the mean square error;

[0018] Determine whether the difference ratio is greater than or equal to a ratio threshold;

[0019] If so, it is considered that the power battery has a temperature abnormality.

[0020] Furthermore, the real-time operation data is preprocessed to obtain target operation data, including:

[0021] Performing data cleaning on the real-time operation data to obtain processed operation data;

[0022] The processed operation data is normalized to obtain the target operation data.

[0023] Furthermore, the temperature prediction model is trained by the following steps:

[0024] Acquire multiple groups of vehicle operation sample data and temperature sensor sample data corresponding to each group of vehicle operation sample data;

[0025] Input each set of vehicle operation sample data into the original prediction model, and obtain temperature sensor prediction data corresponding to each set of vehicle operation sample data output by the original prediction model;

[0026] Compare the temperature sensor sample data and the temperature sensor prediction data corresponding to each set of vehicle operation sample data, and calculate the loss function of the original prediction model under the current state;

[0027] Based on the loss function of the original prediction model, the model parameters of the original prediction model are continuously adjusted until the original prediction model reaches a convergence state, thereby obtaining the temperature prediction model.

[0028] In a second aspect, an embodiment of the present application further provides a device for detecting abnormal temperature of a power battery, the detection device comprising:

[0029] An operation data acquisition module is used to acquire real-time operation data of the electric vehicle and perform data preprocessing on the real-time operation data to obtain target operation data;

[0030] A temperature prediction module, used to obtain the current actual temperature value of the power battery detected by the temperature sensor in the electric vehicle, and input the target operation data into a pre-trained temperature prediction model to obtain a current predicted temperature value for the temperature sensor;

[0031] a difference correction module, configured to correct the difference between the current predicted temperature value and the current actual temperature value based on a weight factor corresponding to the temperature sensor, so as to obtain a target predicted difference; wherein the weight factor corresponding to the temperature sensor is determined by an attribute factor of the temperature sensor;

[0032] A mean square error calculation module, used to obtain in real time a plurality of historical predicted temperature values ​​and a plurality of historical actual temperature values ​​of the power battery in a historical time period through a sliding window algorithm, and calculate a mean square error according to the plurality of historical predicted temperature values ​​and the plurality of historical actual temperature values;

[0033] The abnormality detection module is used to determine whether the power battery has a temperature abnormality based on the target prediction difference and the mean square error.

[0034] Furthermore, the detection device further includes a weight factor determination module, and the weight factor determination module is used to determine the weight factor corresponding to the temperature sensor through the following steps:

[0035] Acquire the attribute factors of the temperature sensor and the adjustment coefficient corresponding to each attribute factor; wherein the attribute factors include position attribute factors, flow channel attribute factors and impedance attribute factors;

[0036] The product between each attribute factor and the corresponding adjustment coefficient is determined, and each product is summed to obtain the weight factor.

[0037] In a third aspect, an embodiment of the present application further provides an electronic device, comprising: a processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor communicates with the memory through the bus, and when the machine-readable instructions are executed by the processor, the steps of the method for detecting abnormal power battery temperature as described above are performed.

[0038] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the method for detecting abnormal power battery temperature as described above are executed.

[0039] The embodiments of the present application provide a method, device, equipment and medium for detecting abnormal temperature of a power battery, which obtains real-time operating data of an electric vehicle, and performs data preprocessing on the real-time operating data to obtain target operating data; then, obtains the current actual temperature value of the power battery detected by the temperature sensor in the electric vehicle, and inputs the target operating data into a pre-trained temperature prediction model to obtain the current predicted temperature value for the temperature sensor; based on the weight factor corresponding to the temperature sensor, corrects the difference between the current predicted temperature value and the current actual temperature value to obtain a target predicted difference; wherein the weight factor corresponding to the temperature sensor is determined by the attribute factor of the temperature sensor; a plurality of historical predicted temperature values ​​and a plurality of historical actual temperature values ​​of the power battery in a historical time period are obtained in real time by a sliding window algorithm, and a mean square error is calculated based on the plurality of historical predicted temperature values ​​and the plurality of historical actual temperature values; finally, based on the target predicted difference and the mean square error, it is determined whether the power battery has a temperature abnormality.

[0040] This application predicts the temperature of the power battery through a temperature prediction model, corrects the difference between the predicted temperature and the actual temperature through the weight factor corresponding to the temperature sensor, and makes an abnormal judgment on the temperature of the power battery in combination with the historical temperature prediction and the historical temperature actual situation. It improves the reliability and efficiency of power battery temperature anomaly detection, thereby improving the accuracy and real-time monitoring of electric vehicle battery temperature, and provides more accurate battery health monitoring for electric vehicles.

[0041] In order to make the above-mentioned objects, features and advantages of the present application more obvious and easy to understand, preferred embodiments are specifically cited below and described in detail with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0043] Figure 1 A flow chart of a method for detecting abnormal power battery temperature provided in an embodiment of the present application;

[0044] Figure 2 This is an example diagram of temperature sampling changes of a power battery provided in an embodiment of the present application;

[0045] Figure 3 A schematic diagram of a method for detecting abnormal power battery temperature provided in an embodiment of the present application;

[0046] Figure 4 A schematic diagram of the structure of a power battery temperature abnormality detection device provided in an embodiment of the present application;

[0047] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0048] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application usually described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, each other embodiment obtained by those skilled in the art without making creative work belongs to the scope of protection of the present application.

[0049] First, the application scenarios to which the present application is applicable are introduced. The present application can be applied to the field of power battery temperature monitoring technology.

[0050] With the rapid development of new energy vehicles, power batteries, as the core components of new energy vehicles, have a crucial impact on the safety and reliability of the vehicle. The consistency of battery temperature is an important indicator to measure the health of the battery pack.

[0051] Research has found that currently, most battery temperature monitoring methods rely on data from onboard sensors. This technology includes setting a fixed temperature threshold and analyzing the temperature rise trend. However, this early warning method based on simple trend analysis can only provide a relative comparison between different sampling points and lacks an absolute judgment standard. Therefore, how to improve the accuracy of power battery temperature anomaly detection has become a technical issue that cannot be underestimated.

[0052] Based on this, the embodiment of the present application provides a method for detecting abnormal power battery temperature, which improves the reliability and efficiency of power battery temperature abnormality detection, thereby improving the accuracy and real-time monitoring of electric vehicle battery temperature, and providing electric vehicles with more accurate battery health monitoring.

[0053] See also Figure 1 , Figure 1 This is a flow chart of a method for detecting abnormal power battery temperature provided by an embodiment of the present application. Figure 1 As shown in, the detection method provided in the embodiment of the present application includes:

[0054] S101, acquiring real-time operating data of the electric vehicle, and performing data preprocessing on the real-time operating data to obtain target operating data.

[0055] For the above step S101, in the specific implementation, the real-time operation data of the electric vehicle is obtained from the cloud in real time. Here, as an example, the real-time operation data may include the vehicle status, charging status, driving status and other data of the electric vehicle, which is not specifically limited in this application. Then, the acquired real-time operation data is preprocessed to obtain the target operation data. Here, when collecting different real-time operation data, it may be necessary to obtain data from different hardware or software interfaces depending on the data source.

[0056] As an optional embodiment, with respect to the above step S101, performing data preprocessing on the real-time operation data to obtain target operation data includes:

[0057] Step 1011, clean up the real-time operation data to obtain processed operation data.

[0058] In the specific implementation of step 1011, the real-time operation data collected in step S101 is cleaned to obtain processed operation data to improve the quality of the data and make it more suitable for analysis. Specifically, data cleaning includes removing outliers, duplicate values ​​and missing values. For invalid operation data (for example, data outside the normal range), mean substitution or linear interpolation is used to fill it.

[0059] Step 1012: normalize the processed operation data to obtain the target operation data.

[0060] Regarding the above step 1012, during the specific implementation, the processed operating data obtained in step 1011 is normalized, and the operating data is scaled to a fixed range, usually between [0-1], to obtain the target operating data, ensuring that the data remains consistent and comparable during the prediction process.

[0061] S102, obtaining the current actual temperature value of the power battery detected by the temperature sensor in the electric vehicle, and inputting the target operation data into a pre-trained temperature prediction model to obtain the current predicted temperature value for the temperature sensor.

[0062] Here, the temperature sensor is an on-board sensor used to detect the temperature of the power battery in the electric vehicle. The temperature sensor may be one or more, and this application does not make any specific limitation on this.

[0063] With respect to the above step S102, the current actual temperature value of the power battery detected by the temperature sensor in the electric vehicle is obtained, and the target operation data is input into a pre-trained temperature prediction model to obtain the current predicted temperature value for the temperature sensor.

[0064] Here, specifically, the temperature prediction model adopts a single-layer long short-term memory network (LSTM) structure, combined with a data-driven optimization strategy and a simplified model training process, which effectively improves the real-time and accuracy of monitoring. This method standardizes the temperature data by dynamically adjusting the data window and the minimum and maximum scaler to ensure the consistency and comparability of the data, while reducing the complexity and time of model training. The single-layer LSTM network architecture makes the model structure simpler and more efficient. The single-layer design reduces the complexity of the model, reduces the training time and consumption of computing resources, and reduces the risk of overfitting. Compared with multi-layer LSTM networks, single-layer networks can maintain good prediction capabilities and computational efficiency with relatively less data. Hidden layer optimization: Set a limited number of hidden units to balance the computational complexity and accuracy of the model. The hidden layer parameters of the model are carefully adjusted to ensure that the model is lightweight and responsive while capturing the long-term dependencies of time series data. The LSTM model consists of an LSTM layer and a fully connected layer. The input data is passed to the LSTM layer, and its output is converted into the final prediction result through the fully connected layer.

[0065] As an optional embodiment, according to the detection method provided in the present application, the temperature prediction model is trained through the following steps:

[0066] I: Acquire multiple groups of vehicle operation sample data and temperature sensor sample data corresponding to each group of vehicle operation sample data.

[0067] For the above step I, in the specific implementation, multiple groups of vehicle operation sample data and temperature sensor sample data corresponding to each group of vehicle operation sample data are obtained. Here, specifically, in order to ensure the consistency and comparability of the data during the training process and the authenticity of the model prediction process, the vehicle operation sample data also needs to be cleaned and normalized.

[0068] Specifically, when dynamically constructing a sample data set, the time window is flexibly adjusted, and the time window length (look_back parameter) is adjusted based on the actual application scenario to adapt to different application environments. The adjustment of the time window enables the model to flexibly adjust the composition of the input data set according to different temperature change patterns. The sliding window method with adjustable window length is used to divide the data set. For example, through the look_back parameter, the window size can be adjusted from 3 to 10 days.

[0069] II: Input each set of vehicle operation sample data into the original prediction model, and obtain the temperature sensor prediction data corresponding to each set of vehicle operation sample data output by the original prediction model.

[0070] III: Compare the temperature sensor sample data and the temperature sensor prediction data corresponding to each group of vehicle operation sample data, and calculate the loss function of the original prediction model under the current state.

[0071] IV: Based on the loss function of the original prediction model, the model parameters of the original prediction model are continuously adjusted until the original prediction model reaches a convergence state to obtain the temperature prediction model.

[0072] For the above steps II to IV, in the specific implementation, each group of vehicle operation sample data is input into the original prediction model, and the temperature sensor prediction data corresponding to each group of vehicle operation sample data output by the original prediction model is obtained. For each group of vehicle operation sample data, the temperature sensor sample data corresponding to the group of vehicle operation sample data and the temperature sensor prediction data are compared. By comparing the temperature sensor sample data corresponding to the group of vehicle operation sample data and the temperature sensor prediction data, it is judged whether the prediction of the original prediction model is accurate. When the temperature sensor sample data and the temperature sensor prediction data are different, it is considered that the prediction of the original prediction model is inaccurate. At this time, it is necessary to calculate the loss function of the original prediction model in the current state. The method of calculating the loss function is described in detail in the prior art, and will not be described in detail here. Then the model parameters of the original prediction model are continuously adjusted. The original prediction model will continuously minimize the loss through iteration. At each step of the iteration, the loss value of the original prediction model will be calculated. When the loss value of the original prediction model cannot reach the loss threshold, the model parameters of the original prediction model are continuously updated, and the new parameters will calculate the new loss value, so that the loss value shows a trend of fluctuating downward during the iteration process. Finally, when the loss value reaches smoothness, that is, when the loss value of the trained original prediction model is not greater than the loss threshold, that is, when the loss value does not decrease significantly compared with the loss value calculated last time, it is considered that the original prediction model has reached a convergence state. At this time, the prediction of the original prediction model is more accurate. At this time, the training is terminated and the temperature prediction model is obtained.

[0073] Here, when training the temperature prediction model, the number of training cycles is reasonably adjusted so that the model can complete training quickly while ensuring prediction accuracy, adapting to the real-time prediction needs in different scenarios. A batch training strategy is used, and the size of each batch is adjusted according to the amount of data and the availability of computing resources. A batch-based training strategy is used in combination with a gradient optimization algorithm to achieve efficient updating of model parameters. By using the Adam optimizer, the model maintains stable convergence while maintaining a high training speed.

[0074] S103: Based on the weight factor corresponding to the temperature sensor, correct the difference between the current predicted temperature value and the current actual temperature value to obtain a target predicted difference.

[0075] Here, the weight factor corresponding to the temperature sensor is determined by the attribute factor of the temperature sensor. Specifically, when the difference is corrected by sampling the position of the thermal management pipeline, the connection impedance, and the contact between the battery and the outside world, the attribute factor of the temperature sensor may include the position attribute factor, the flow channel attribute factor, and the impedance attribute factor.

[0076] In the specific implementation of step S103, the weight factor corresponding to the temperature sensor is obtained, and based on the weight factor, the difference between the current predicted temperature value and the current actual temperature value is corrected to obtain the target predicted difference. In this way, the predicted anomaly will be dynamically adjusted according to the specific attributes of each sensor.

[0077] As an optional embodiment, according to the detection method provided in the embodiment of the present application, the weight factor corresponding to the temperature sensor is determined by the following steps:

[0078] A: Obtain the attribute factor of the temperature sensor and the adjustment coefficient corresponding to each attribute factor.

[0079] Specifically, the attribute factors of the temperature sensor include a position attribute factor, a flow channel attribute factor, and an impedance attribute factor. The adjustment coefficient is used to adjust the contribution of different attributes to the final weight.

[0080] Regarding the above step A, during the specific implementation, the attribute factors of the temperature sensor and the adjustment coefficient corresponding to each attribute factor are obtained.

[0081] B: Determine the product between each attribute factor and the corresponding adjustment coefficient, and sum each product to obtain the weight factor.

[0082] Regarding the above step B, in specific implementation, each attribute factor is multiplied by the corresponding adjustment coefficient to obtain the product between each attribute factor and the corresponding adjustment coefficient, and each product is summed to obtain the weight factor corresponding to the temperature sensor.

[0083] Here, specifically, the weight factor corresponding to temperature sensor i is defined as w i , according to the properties of position, flow channel and impedance, the weight factor w is calculated by the following formula i :

[0084] w i =α·Location_Factor(i)+β·Flow_Channel_Factor(i)+γ

[0085] Impedance_Factor(i)

[0086] Among them, Location_Factor(i) represents the location attribute factor of temperature sensor i, α represents the adjustment coefficient of the location attribute factor, Flow_Channel_Factor(i) represents the flow channel attribute factor of temperature sensor i, β represents the adjustment coefficient of the flow channel attribute factor, Impedance_Factor(i) represents the impedance attribute factor of temperature sensor i, and γ represents the adjustment coefficient of the impedance attribute factor.

[0087] As an optional embodiment, when there are multiple temperature sensors, a weighted sum may be performed on the weight factors corresponding to each temperature sensor to determine the final weight factors of the multiple temperature sensors, and the weight factors may be used to correct the prediction differences.

[0088] Thus, according to the above steps A-B, weight factors can be defined according to the location of the sensor, the flow channel properties and the impedance factor, the temperature difference detection standard can be adjusted, and the weights of temperature prediction and abnormality detection can be adjusted. In this way, the temperature prediction difference can be corrected in combination with the specific properties of the temperature sensor to ensure the accuracy of abnormality detection.

[0089] Specifically, with respect to the above step S103, the difference between the current predicted temperature value and the current actual temperature value is corrected based on the weight factor corresponding to the temperature sensor to obtain a target predicted difference, including:

[0090] The product of the difference between the current predicted temperature value and the current actual temperature value and the weight factor is determined as the target predicted difference.

[0091] For the above steps, in the specific implementation, the difference between the current predicted temperature value and the current actual temperature value is corrected by the weight factor corresponding to the temperature sensor, and the product of the difference between the current predicted temperature value and the current actual temperature value and the weight factor is determined as the target prediction difference. The temperature prediction difference is corrected according to the properties of the sensor to ensure the accuracy of anomaly detection. Specifically, the target prediction difference is calculated by the following formula:

[0092] ΔT corrected,i =w i ·(T pred,i -T actual,i )

[0093] Where, ΔT corrected,i represents the target prediction difference, T pred,i and T actual,i are the current predicted temperature value and the current actual temperature value of temperature sensor i respectively.

[0094] S104, obtaining in real time a plurality of historical predicted temperature values ​​and a plurality of historical actual temperature values ​​of the power battery within a historical time period through a sliding window algorithm, and calculating a mean square error according to the plurality of historical predicted temperature values ​​and the plurality of historical actual temperature values.

[0095] For the above step S104, in the specific implementation, for the above step S104, in the specific implementation, a plurality of historical predicted temperature values ​​and a plurality of historical actual temperature values ​​of the power battery in the historical time period are obtained in real time through the sliding window algorithm, and the mean square error is calculated based on the plurality of historical predicted temperature values ​​and the plurality of historical actual temperature values. Here, the window length of the sliding window can be 24 hours, which is not specifically limited in this application. In this way, the error between the predicted temperature and the actual temperature in the past period of time is calculated in real time through the sliding window algorithm, the ratio of the temperature difference is dynamically adjusted, and a statistical baseline is established for anomaly detection. A fixed time window length is set, the temperature prediction and actual data in the window are compared, and the mean square error is calculated to determine the prediction accuracy.

[0096] S105 , judging whether the power battery has a temperature abnormality based on the target prediction difference and the mean square error.

[0097] Regarding the above step S105, in a specific implementation, it is determined whether the power battery has a temperature abnormality based on the target prediction difference calculated in step S103 and the mean square error calculated in step S104.

[0098] As an optional embodiment, with respect to the above step S105, judging whether the power battery has a temperature abnormality based on the target prediction difference and the mean square error includes:

[0099] The difference ratio is calculated according to the target prediction difference and the mean square error; and it is determined whether the difference ratio is greater than or equal to a ratio threshold; if so, it is considered that the power battery has a temperature abnormality.

[0100] Here, the ratio threshold may be pre-set to 0.2, and this application does not make any specific limitation on this.

[0101] For the above steps, in the specific implementation, the difference ratio is calculated based on the target prediction difference and the mean square error. When the difference ratio is greater than or equal to the ratio threshold, it is considered that the power battery has a temperature anomaly. In this way, the difference ratio is calculated based on the difference between the mean square error and the actual temperature in different time windows to help determine the temperature anomaly. When the temperature difference exceeds the dynamic threshold within a certain time window, this situation is marked as abnormal.

[0102] See also Figure 2 , Figure 2 This is an example diagram of the temperature sampling changes of a power battery provided in an embodiment of the present application. Figure 2 As shown in , a sliding window algorithm is used to calculate the mean and standard deviation of the battery temperature prediction error, and a dynamic statistical baseline is established. The deviation percentage of each temperature prediction is shown in the sub-figure title. Figure 2 The embodiment in the example uses a deviation of more than 10% as a temperature anomaly. At the same time, combined with the vehicle's operating status, charging status and environmental conditions, the temperature deviation is intelligently evaluated to provide fault diagnosis and early warning. When an abnormality is identified, the system automatically adjusts the thermal management strategy or reminds maintenance personnel to conduct inspections. Figure 2 In the embodiment, it was found that the temperature samples numbered 1 / 2 / 3 / 4 / 11 / 13 / 14 / 15 / 16 were abnormal. After detailed analysis, the water path through which the temperature samples flowed was blocked and the abnormality was confirmed.

[0103] See also Figure 3 , Figure 3 A schematic diagram of a method for detecting abnormal power battery temperature provided in an embodiment of the present application. Figure 3 As shown in the figure, the vehicle's operating data is first obtained through the cloud, and then the battery temperature is predicted through an efficient algorithm. At the same time, the temperature is sampled through the thermal management pipeline position, connection impedance, and battery contact correction difference. The dynamic sliding window calculates the ratio of the predicted and actual temperature differences. Combined with temperature monitoring, environmental analysis, and abnormal judgment of thermal management strategies, the power battery temperature is detected for abnormalities. If there is no abnormality, continuous monitoring is performed. If there is an abnormality, the vehicle-side alarm is triggered, and preventive strategies such as power limiting are implemented.

[0104] The method for detecting abnormal temperature of a power battery provided in an embodiment of the present application comprises the following steps: first, obtaining real-time operating data of an electric vehicle, and performing data preprocessing on the real-time operating data to obtain target operating data; then, obtaining the current actual temperature value of the power battery detected by a temperature sensor in the electric vehicle, and inputting the target operating data into a pre-trained temperature prediction model to obtain a current predicted temperature value for the temperature sensor; based on a weight factor corresponding to the temperature sensor, correcting the difference between the current predicted temperature value and the current actual temperature value to obtain a target predicted difference; wherein the weight factor corresponding to the temperature sensor is determined by an attribute factor of the temperature sensor; obtaining in real time multiple historical predicted temperature values ​​and multiple historical actual temperature values ​​of the power battery in a historical time period through a sliding window algorithm, and calculating a mean square error based on the multiple historical predicted temperature values ​​and the multiple historical actual temperature values; finally, judging whether the power battery has a temperature abnormality based on the target predicted difference and the mean square error.

[0105] This application predicts the temperature of the power battery through a temperature prediction model, corrects the difference between the predicted temperature and the actual temperature through the weight factor corresponding to the temperature sensor, and makes an abnormal judgment on the temperature of the power battery in combination with the historical temperature prediction and the historical temperature actual situation. It improves the reliability and efficiency of power battery temperature anomaly detection, thereby improving the accuracy and real-time monitoring of electric vehicle battery temperature, and provides more accurate battery health monitoring for electric vehicles.

[0106] See also Figure 4 , Figure 4 This is a schematic diagram of the structure of a power battery temperature abnormality detection device provided in an embodiment of the present application. Figure 4 As shown in , the detection device 400 includes:

[0107] The operation data acquisition module 401 is used to acquire the real-time operation data of the electric vehicle and perform data preprocessing on the real-time operation data to obtain target operation data;

[0108] The temperature prediction module 402 is used to obtain the current actual temperature value of the power battery detected by the temperature sensor in the electric vehicle, and input the target operation data into a pre-trained temperature prediction model to obtain the current predicted temperature value for the temperature sensor;

[0109] A difference correction module 403, used to correct the difference between the current predicted temperature value and the current actual temperature value based on the weight factor corresponding to the temperature sensor to obtain a target predicted difference; wherein the weight factor corresponding to the temperature sensor is determined by the attribute factor of the temperature sensor;

[0110] A mean square error calculation module 404 is used to obtain multiple historical predicted temperature values ​​and multiple historical actual temperature values ​​of the power battery in a historical time period in real time through a sliding window algorithm, and calculate a mean square error based on the multiple historical predicted temperature values ​​and the multiple historical actual temperature values;

[0111] The abnormality detection module 405 is used to determine whether the power battery has a temperature abnormality based on the target prediction difference and the mean square error.

[0112] Furthermore, the detection device 400 further includes a weight factor determination module, and the weight factor determination module is used to determine the weight factor corresponding to the temperature sensor through the following steps:

[0113] Acquire the attribute factors of the temperature sensor and the adjustment coefficient corresponding to each attribute factor; wherein the attribute factors include position attribute factors, flow channel attribute factors and impedance attribute factors;

[0114] The product between each attribute factor and the corresponding adjustment coefficient is determined, and each product is summed to obtain the weight factor.

[0115] Further, when the difference correction module 403 is used to correct the difference between the current predicted temperature value and the current actual temperature value based on the weight factor corresponding to the temperature sensor to obtain the target predicted difference, the difference correction module 403 is also used to:

[0116] The product of the difference between the current predicted temperature value and the current actual temperature value and the weight factor is determined as the target predicted difference.

[0117] Further, when the abnormality detection module 405 is used to determine whether the power battery has a temperature abnormality based on the target prediction difference and the mean square error, the abnormality detection module 405 is also used to:

[0118] Calculating a difference ratio based on the target prediction difference and the mean square error;

[0119] Determine whether the difference ratio is greater than or equal to a ratio threshold;

[0120] If so, it is considered that the power battery has a temperature abnormality.

[0121] Furthermore, when the operation data acquisition module 401 is used to perform data preprocessing on the real-time operation data to obtain target operation data, the operation data acquisition module 401 is also used to:

[0122] Performing data cleaning on the real-time operation data to obtain processed operation data;

[0123] The processed operation data is normalized to obtain the target operation data.

[0124] Furthermore, the detection device 400 further includes a model training module, and the model training module is used to train the temperature prediction model through the following steps:

[0125] Acquire multiple groups of vehicle operation sample data and temperature sensor sample data corresponding to each group of vehicle operation sample data;

[0126] Input each set of vehicle operation sample data into the original prediction model, and obtain temperature sensor prediction data corresponding to each set of vehicle operation sample data output by the original prediction model;

[0127] Compare the temperature sensor sample data and the temperature sensor prediction data corresponding to each set of vehicle operation sample data, and calculate the loss function of the original prediction model under the current state;

[0128] Based on the loss function of the original prediction model, the model parameters of the original prediction model are continuously adjusted until the original prediction model reaches a convergence state, thereby obtaining the temperature prediction model.

[0129] See also Figure 5 , Figure 5 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. Figure 5 As shown in , the electronic device 500 includes a processor 510 , a memory 520 and a bus 530 .

[0130] The memory 520 stores machine-readable instructions executable by the processor 510. When the electronic device 500 is running, the processor 510 communicates with the memory 520 via the bus 530. When the machine-readable instructions are executed by the processor 510, the above-mentioned Figure 1 The steps of the method for detecting abnormal power battery temperature in the method embodiment shown and the specific implementation manner can be referred to the method embodiment, which will not be described in detail here.

[0131] The present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the computer program can execute the above-mentioned Figure 1 The steps of the method for detecting abnormal power battery temperature in the method embodiment shown and the specific implementation manner can be referred to the method embodiment, which will not be described in detail here.

[0132] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0133] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interfaces, and the indirect coupling or communication connection of devices or units can be electrical, mechanical or other forms.

[0134] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0135] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0136] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a non-volatile computer-readable storage medium that is executable by a processor. Based on this understanding, the technical solution of the present application can essentially be embodied in the form of a software product, or in other words, the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0137] Finally, it should be noted that the above-described embodiments are only specific implementation methods of the present application, which are used to illustrate the technical solutions of the present application, rather than to limit them. The protection scope of the present application is not limited thereto. Although the present application is described in detail with reference to the above-mentioned embodiments, ordinary technicians in the field should understand that any technician familiar with the technical field can still modify the technical solutions recorded in the above-mentioned embodiments within the technical scope disclosed in the present application, or can easily think of changes, or make equivalent replacements for some of the technical features therein; and these modifications, changes or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application, and should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.

Claims

1. A method for detecting abnormal power battery temperature, characterized in that: The detection method comprises: Acquiring real-time operating data of the electric vehicle, and performing data preprocessing on the real-time operating data to obtain target operating data; Acquire the current actual temperature value of the power battery detected by the temperature sensor in the electric vehicle, and input the target operation data into a pre-trained temperature prediction model to obtain the current predicted temperature value for the temperature sensor; Based on the weight factor corresponding to the temperature sensor, the difference between the current predicted temperature value and the current actual temperature value is corrected to obtain a target predicted difference; wherein the weight factor corresponding to the temperature sensor is determined by an attribute factor of the temperature sensor; Acquire multiple historical predicted temperature values ​​and multiple historical actual temperature values ​​of the power battery in a historical time period in real time through a sliding window algorithm, and calculate a mean square error based on the multiple historical predicted temperature values ​​and the multiple historical actual temperature values; It is determined whether the power battery has a temperature abnormality based on the target prediction difference and the mean square error.

2. The detection method according to claim 1, characterized in that: The weight factor corresponding to the temperature sensor is determined by the following steps: Acquire the attribute factors of the temperature sensor and the adjustment coefficient corresponding to each attribute factor; wherein the attribute factors include position attribute factors, flow channel attribute factors and impedance attribute factors; The product between each attribute factor and the corresponding adjustment coefficient is determined, and each product is summed to obtain the weight factor.

3. The detection method according to claim 1, characterized in that: The correcting the difference between the current predicted temperature value and the current actual temperature value based on the weight factor corresponding to the temperature sensor to obtain a target predicted difference includes: The product of the difference between the current predicted temperature value and the current actual temperature value and the weight factor is determined as the target predicted difference.

4. The detection method according to claim 1, characterized in that: The determining whether the power battery has a temperature abnormality based on the target prediction difference and the mean square error includes: Calculating a difference ratio based on the target prediction difference and the mean square error; Determine whether the difference ratio is greater than or equal to a ratio threshold; If so, it is considered that the power battery has a temperature abnormality.

5. The detection method according to claim 1, characterized in that: The performing data preprocessing on the real-time operation data to obtain target operation data includes: Performing data cleaning on the real-time operation data to obtain processed operation data; The processed operation data is normalized to obtain the target operation data.

6. The detection method according to claim 1, characterized in that: The temperature prediction model is trained by the following steps: Acquire multiple groups of vehicle operation sample data and temperature sensor sample data corresponding to each group of vehicle operation sample data; Input each set of vehicle operation sample data into the original prediction model, and obtain temperature sensor prediction data corresponding to each set of vehicle operation sample data output by the original prediction model; Compare the temperature sensor sample data and the temperature sensor prediction data corresponding to each set of vehicle operation sample data, and calculate the loss function of the original prediction model under the current state; Based on the loss function of the original prediction model, the model parameters of the original prediction model are continuously adjusted until the original prediction model reaches a convergence state, thereby obtaining the temperature prediction model.

7. A device for detecting abnormal power battery temperature, characterized in that: The detection device comprises: An operation data acquisition module is used to acquire real-time operation data of the electric vehicle and perform data preprocessing on the real-time operation data to obtain target operation data; A temperature prediction module, used to obtain the current actual temperature value of the power battery detected by the temperature sensor in the electric vehicle, and input the target operation data into a pre-trained temperature prediction model to obtain a current predicted temperature value for the temperature sensor; a difference correction module, configured to correct the difference between the current predicted temperature value and the current actual temperature value based on a weight factor corresponding to the temperature sensor, so as to obtain a target predicted difference; wherein the weight factor corresponding to the temperature sensor is determined by an attribute factor of the temperature sensor; A mean square error calculation module, used to obtain in real time a plurality of historical predicted temperature values ​​and a plurality of historical actual temperature values ​​of the power battery in a historical time period through a sliding window algorithm, and calculate a mean square error according to the plurality of historical predicted temperature values ​​and the plurality of historical actual temperature values; The abnormality detection module is used to determine whether the power battery has a temperature abnormality based on the target prediction difference and the mean square error.

8. The detection device according to claim 7, characterized in that: The detection device further includes a weight factor determination module, which is used to determine the weight factor corresponding to the temperature sensor through the following steps: Acquire the attribute factors of the temperature sensor and the adjustment coefficient corresponding to each attribute factor; wherein the attribute factors include position attribute factors, flow channel attribute factors and impedance attribute factors; The product between each attribute factor and the corresponding adjustment coefficient is determined, and each product is summed to obtain the weight factor.

9. An electronic device, characterized in that: include: A processor, a memory and a bus, wherein the memory stores machine-readable instructions executable by the processor, and when the electronic device is running, the processor and the memory communicate through the bus, and the machine-readable instructions are executed by the processor to execute the steps of the method for detecting abnormal power battery temperature as described in any one of claims 1 to 6.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the method for detecting abnormal power battery temperature as claimed in any one of claims 1 to 6 are executed.

Images