Battery pre-regulation thermal management method and device based on multi-source data prediction

Abstract

The invention relates to the technical field of battery thermal management, in particular to a battery pre-regulation thermal management method and device based on multi-source data prediction. The method comprises the following steps: acquiring environment data and battery state data; adopting a temperature prediction model to predict future temperature data of the battery; according to the future temperature data and the current temperature data of the battery, determining the expected arrival time of the extreme weather; and pre-adjustment starting time is determined according to the expected arrival time of the extreme weather, the temperature change condition and the safety margin time. According to the method, the prediction model for coupling the environmental data and the battery state is established, the potential temperature change trend of the battery is predicted, the dependence on the real-time temperature is avoided, and the response efficiency is improved. Meanwhile, the pre-adjustment starting time is determined according to the expected arrival time of extreme weather, the temperature change condition and the safety margin time, so that dynamic adjustment of the pre-adjustment starting time of the thermal management system is achieved, and the problem that thermal management passive response lags behind or the whole-time operation cost is high is solved.

Description

Technical Field

[0001] The present invention relates to the technical field of battery thermal management, and in particular to a battery pre-regulation thermal management method and device based on multi-source data prediction. Background Art

[0002] Existing battery thermal management systems use real-time feedback or full-time operation to achieve temperature control in battery energy storage power stations. When using real-time feedback, temperature sensors distributed throughout the battery pack monitor the temperature of each cell or module in real time, and the thermal management system is regulated based on this real-time temperature data. However, this real-time feedback approach struggles to respond quickly to extreme temperature fluctuations, resulting in a delayed passive response that can lead to battery performance degradation or the risk of thermal runaway.

[0003] When using full-time operation, the system will continue to operate regardless of external temperature shocks to maintain the battery temperature within the preset range. Although this avoids the risk of battery performance degradation or thermal runaway, full-time operation will increase the overall energy consumption cost of the system. Summary of the Invention

[0004] In view of this, the present invention provides a battery pre-regulation thermal management method and device based on multi-source data prediction to solve the problems of passive response lag or high cost in the battery thermal management adjustment method in the prior art.

[0005] In a first aspect, the present invention provides a battery pre-regulation thermal management method based on multi-source data prediction, the method comprising: acquiring environmental data and battery status data; using a temperature prediction model to predict the future temperature data of the battery based on the environmental data and the battery status data; determining the expected arrival time of extreme weather based on the future temperature data of the battery and the current temperature data in the environmental data; and determining the pre-regulation start time based on the expected arrival time of extreme weather, temperature changes, and safety margin time.

[0006] This invention establishes a predictive model that couples environmental data with battery status to predict potential battery temperature trends, eliminating reliance on real-time temperature and improving response efficiency. Furthermore, the pre-conditioning start time is determined based on the expected arrival time of extreme weather, temperature fluctuations, and safety margin. This allows for dynamic adjustment of the thermal management system's pre-conditioning start time, avoiding the problems of delayed passive thermal management responses and high full-time operating costs.

[0007] In an optional embodiment, the method further includes: starting a heating mode when the predicted future temperature data of the battery is lower than the target temperature; starting a cooling mode when the predicted future temperature data of the battery is higher than the target temperature; starting a backup heat source when the difference between the real-time monitored battery temperature data and the predicted future temperature data of the battery is higher than a first preset value and the duration is higher than a second preset value, or when the difference between the maximum and minimum temperatures inside the monitored battery is higher than a third preset value.

[0008] In the present invention, whether to start the heating mode or the cooling mode is determined based on the relationship between the predicted temperature data and the target temperature, thereby realizing thermal management of the battery and achieving normal operation of the battery; at the same time, whether to start the backup heat source is determined based on the temperature difference, thereby balancing the battery temperature and improving battery performance.

[0009] In an optional embodiment, environmental data and battery status data are obtained, including: obtaining first environmental data for real-time monitoring and second environmental data for forecast; fusing the first environmental data and the second environmental data using a wavelet transform-support vector machine combined noise reduction algorithm or an improved Kalman filter algorithm, wherein the process noise covariance matrix of the improved Kalman filter algorithm is determined according to the data credibility, and the improved Kalman filter algorithm includes a temperature gradient constraint condition; obtaining SOC, SOH and charge and discharge current collected by the battery management system as battery status data.

[0010] In the present invention, data fusion is performed by obtaining monitored environmental data and forecasted environmental data as the data basis for model prediction, thereby improving the accuracy of model prediction and thus improving the accuracy of extreme weather determination; in addition, by introducing the process noise covariance matrix determined according to data credibility and the temperature gradient constraint condition into the Kalman filter algorithm, the improved Kalman filter can, when processing data, on the one hand, reasonably adjust the calculation weight according to the data credibility, and on the other hand, constrain the calculation results to conform to the actual physical laws, thereby more accurately estimating state variables such as battery temperature, providing a more reliable basis for subsequent thermal management decisions, and helping to improve the performance and reliability of the battery thermal management system.

[0011] In an optional embodiment, the temperature prediction model includes a pre-trained Transformer time series model, a pre-trained LSTM model, or a physical model and a data-driven hybrid model; the pre-trained LSTM model includes an input layer, an LSTM layer, a Dropout layer, a fully connected layer, and an output layer, and the input data of the input layer includes SOC, SOH, charge and discharge current, ambient temperature, temperature change rate, wind speed, solar radiation intensity, and relative humidity; the loss function of the pre-trained long short-term memory network model includes a temperature gradient penalty term.

[0012] In the present invention, by using a variety of data as model input for temperature prediction, the influence of multiple factors can be comprehensively considered, making the model adaptable to different usage scenarios and environmental conditions, thereby improving the generalization ability and adaptability of the model.

[0013] In an optional embodiment, the temperature gradient penalty term is expressed using the following formula:

[0014] W new =W old +η·(T real -T pred )·X input

[0015] Where W new represents the updated weight, W old represents the weight before update, η represents the learning rate, T real Indicates the real-time monitored battery temperature data, T pred represents the battery temperature data predicted by the temperature prediction model, X input A feature vector representing the input to the input layer.

[0016] In the present invention, by adopting the above formula to set the temperature gradient penalty term in the loss function of the model, a closed-loop feedback calibration mechanism of the model is realized, thereby improving the accuracy of the model.

[0017] In an optional embodiment, when the predicted future temperature data of the battery is less than the target temperature, the heating mode is started, including: when the difference between the predicted future temperature data of the battery and the target temperature is greater than or equal to a fourth preset value, the first power heating mode is started; when the difference between the predicted future temperature data of the battery and the target temperature is greater than or equal to a fifth preset value and less than the fourth preset value, the second power heating mode is started, and the second power is less than the first power; when the difference between the predicted future temperature data of the battery and the target temperature is less than the fifth preset value, the pulse heating mode is started; when the predicted future temperature data of the battery is greater than the target temperature, after starting the cooling mode, the method also includes: determining the flow rate of the coolant in the cooling mode according to the fuzzy PID method.

[0018] In this invention, the power of the heating mode is changed according to temperature changes, thereby achieving precise regulation of battery temperature and reducing energy consumption. At the same time, the flow rate of the coolant is determined by a fuzzy PID algorithm, achieving dynamic flow adjustment.

[0019] In an optional embodiment, the pre-adjustment start time is determined using the following formula:

[0020]

[0021] Where, T startIndicates the pre-regulation start time, T event Indicates the expected arrival time of extreme weather, ΔT req Indicates the target temperature rise, represents the rate of temperature change, t safe represents the safety margin time, α represents the correction coefficient, n represents the number of historical samples, T pred,i Indicates the predicted battery temperature corresponding to the i-th sample, T real,i Indicates the battery temperature monitored by the i-th sample.

[0022] In the present invention, the pre-adjustment start time is determined by using the above formula, thereby achieving accurate determination of the start time and improving the efficiency of responding to extreme weather.

[0023] In a second aspect, the present invention provides a battery pre-regulation thermal management device based on multi-source data prediction, the device comprising: a data acquisition module for acquiring environmental data and battery status data; a temperature prediction module for using a temperature prediction model to predict the future temperature data of the battery based on the environmental data and the battery status data; an extreme weather determination module for determining the expected arrival time of extreme weather based on the future temperature data of the battery and the current temperature data in the environmental data; and a pre-regulation start-up module for determining the pre-regulation start-up time based on the expected arrival time of extreme weather, temperature changes, and safety margin time.

[0024] In a third aspect, the present invention provides a computer device comprising: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to thereby execute the battery pre-conditioning thermal management method based on multi-source data prediction of the above-mentioned first aspect or any corresponding embodiment thereof.

[0025] In a fourth aspect, the present invention provides a computer-readable storage medium having computer instructions stored thereon, the computer instructions being used to enable a computer to execute the battery pre-conditioning thermal management method based on multi-source data prediction of the above-mentioned first aspect or any corresponding embodiment thereof.

[0026] In a fifth aspect, the present invention provides a computer program product comprising computer instructions for causing a computer to execute the battery pre-conditioning thermal management method based on multi-source data prediction according to the first aspect or any corresponding embodiment thereof. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the specific embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0028] Figure 1 is a flow chart of a battery pre-conditioning thermal management method based on multi-source data prediction according to an embodiment of the present invention;

[0029] Figure 2 is a flow chart of another battery pre-conditioning thermal management method based on multi-source data prediction according to an embodiment of the present invention;

[0030] Figure 3 is a structural block diagram of a battery pre-conditioning thermal management device based on multi-source data prediction according to an embodiment of the present invention;

[0031] Figure 4 Schematic diagram of the hardware structure of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0032] To make the purpose, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without making creative efforts shall fall within the scope of protection of the present invention.

[0033] According to an embodiment of the present invention, an embodiment of a battery pre-conditioning thermal management method based on multi-source data prediction is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0034] In this embodiment, a battery pre-conditioning thermal management method based on multi-source data prediction is provided, which can be used in electronic devices such as computers, mobile phones, tablet computers, etc. Figure 1 FIG. 1 is a flow chart of a battery pre-conditioning thermal management method based on multi-source data prediction according to an embodiment of the present invention. Figure 1 As shown, the process includes the following steps:

[0035] Step S101, obtain environmental data and battery status data. Specifically, the environmental data may include environmental data monitored in real time by deployed sensors, and may also include environmental data obtained through weather forecasts. For the method of obtaining environmental data using sensors, environmental data can be collected at preset time intervals, such as at time intervals of 5 minutes; when obtaining environmental data through weather forecasts, data can be obtained through the API interface of the Meteorological Bureau, or through the Internet. This embodiment does not specifically limit the method of obtaining weather forecast data. In addition, the environmental data obtained by the weather forecast includes not only the environmental data of the current time, but also the environmental data for a period of time in the future.

[0036] When the environmental data obtained includes environmental data obtained in multiple ways, the environmental data obtained in multiple ways can be integrated first to improve the accuracy of the data. In addition, for battery status data, the required battery status data can be collected through the battery management system. In this embodiment, the collected battery status data is used to predict the battery temperature, so data related to the battery temperature is used, such as SOC (State of Charge), SOH (State of Health) and charge and discharge current.

[0037] In step S102, a temperature prediction model is used to predict future battery temperature data based on the environmental data and battery status data. Specifically, this temperature prediction model enables prediction of future data based on current data. Therefore, a model for processing time series data is used as the temperature prediction model. For example, a deep learning model such as a Transformer model or a long short-term memory network can be used as the temperature prediction model. Furthermore, before using a deep learning model for prediction, the model can be trained to enable accurate temperature prediction.

[0038] Step S103 , determining the expected arrival time of extreme weather according to the future temperature data of the battery and the current temperature data in the environmental data.

[0039] Step S104 : determining a pre-adjustment start time according to the expected arrival time of extreme weather, temperature changes, and a safety margin time.

[0040] Specifically, in order to achieve better control over thermal management and avoid the problems of delayed passive response of thermal management or high full-time operating costs in related technologies, the occurrence of extreme weather is used as an indicator for the start of thermal management in this embodiment. Among them, in this embodiment, the predicted temperature change rate is greater than or equal to 3°C / h and the duration is >2h, which is used as an indicator of the arrival of extreme weather. The predicted temperature change rate can be determined with the help of the temperature data in the acquired environmental data and the predicted temperature data, thereby further determining the expected arrival time of extreme weather. Specifically, when temperature data is obtained by sensors and weather forecasts, the fused temperature data obtained by the two methods can be used, or the predicted temperature change rate can be calculated using the temperature data obtained by the two methods, and the final value is obtained by weighted average.

[0041] Furthermore, when using the estimated arrival time of extreme weather to determine the pre-conditioning start time, the temperature change rate and safety margin are further considered. This safety margin can be understood as additional time to ensure sufficient time for battery temperature adjustment after thermal management is initiated based on the pre-conditioning start time. In practical applications, this safety margin can be determined using historical temperature data.

[0042] The battery preconditioning thermal management method based on multi-source data prediction provided by the present invention establishes a prediction model that couples environmental data with battery status to predict potential battery temperature trends, eliminating reliance on real-time temperature and improving response efficiency. The preconditioning start time is also determined based on the expected arrival time of extreme weather, temperature fluctuations, and safety margin. This enables dynamic adjustment of the thermal management system's preconditioning start time, avoiding the problems of delayed passive thermal management responses or high full-time operating costs.

[0043] In this embodiment, a battery pre-conditioning thermal management method based on multi-source data prediction is provided. The process includes the following steps:

[0044] Step S201: Acquire environmental data and battery status data.

[0045] Specifically, the above step S201 includes:

[0046] Step S2011: Acquire real-time monitored first environmental data and forecasted second environmental data. The first environmental data can be determined using data acquired by temperature and humidity sensors deployed near the battery, and the second environmental data includes weather forecast environmental data. Furthermore, in this embodiment, the environmental data used for temperature prediction includes ambient temperature, temperature change rate, wind speed, solar radiation intensity, and relative humidity. Thus, the environmental data acquired here includes temperature data, humidity data, wind speed data, and solar radiation intensity data. This environmental data can include both sensor-collected data and forecasted data.

[0047] Step S2012: The first environmental data and the second environmental data are fused using a wavelet transform-support vector machine combined noise reduction algorithm or an improved Kalman filter algorithm. The process noise covariance matrix of the improved Kalman filter algorithm is determined based on the data credibility, and the improved Kalman filter algorithm includes a temperature gradient constraint. The method of using the wavelet transform-support vector machine combined noise reduction algorithm for data fusion during data fusion can be implemented with reference to related technologies. In this embodiment, the improved Kalman filter algorithm is primarily used for data fusion.

[0048] Specifically, when using the improved Kalman filter algorithm for data fusion, an adaptive process noise covariance matrix Q is introduced. This matrix is used to describe the statistical characteristics of the system process noise. In an embodiment, this matrix is determined based on the credibility of the first and second environmental data to be fused. If the credibility of the second environmental data of the weather forecast is higher, Q of the second environmental data is set to 0.01, and if the credibility of the first environmental data is lower, Q of the first environmental data is set to 0.05. By adaptively adjusting Q, the algorithm can better adapt to the characteristics of data from different sources, improving the accuracy of filtering and prediction.

[0049] In addition, when adopting the improved Kalman filter algorithm, a temperature gradient constraint is further introduced. This constraint is used to limit the temperature change amplitude of adjacent time periods to no more than the historical extreme value. Specifically, during the operation of the improved Kalman filter algorithm, once the calculated temperature change of adjacent time periods exceeds the historical extreme value, the calculation result will be corrected to return it to a reasonable range. This can prevent temperature mutations that do not conform to actual physical laws due to abnormal data fluctuations or calculation errors, thereby ensuring the rationality and stability of the temperature estimation results. It should be noted that the improved Kalman filter algorithm is based on the improvement of the adaptive matrix and constraint conditions on the basis of related technologies. For other processing processes, they can be implemented with reference to the relevant Kalman filter algorithm, and this embodiment will not be repeated here.

[0050] Furthermore, in addition to using a wavelet transform-support vector machine combined noise reduction algorithm or a modified Kalman filter algorithm to fuse the first and second environmental data, a federated learning framework can also be used to implement multi-site data collaborative training, thereby achieving the fusion of the two environmental data. The specific training methods can be referenced in related art for federated learning frameworks and will not be elaborated here.

[0051] Step S2013: Acquire the SOC, SOH, and charge and discharge current collected by the battery management system as battery status data.

[0052] Step S202 uses a temperature prediction model to predict the battery's future temperature based on the environmental data and battery status data. The temperature prediction model includes a pre-trained Transformer time series model, a pre-trained LSTM model, or a hybrid model that combines a physical model with data. The pre-trained LSTM model includes an input layer, an LSTM layer, a dropout layer, a fully connected layer, and an output layer. The input layer inputs data including state of charge (SOC), state of ohm (SOH), charge and discharge current, ambient temperature, temperature change rate, wind speed, solar radiation intensity, and relative humidity. The loss function of the pre-trained long-short-term memory network model includes a temperature gradient penalty term.

[0053] Specifically, when using a hybrid model of physical models and data-driven models for temperature prediction, the physical model of the battery and the data-driven model of the battery can be constructed separately (which can be constructed using algorithms of related technologies, such as neural networks, support vector machines, and random forests, etc.), and then the physical model and the data-driven model can be used for temperature prediction respectively. Finally, the temperature data predicted by the two are fused to obtain the final temperature prediction data.

[0054] The method of using a pre-trained Transformer time series model and a pre-trained LSTM model for temperature prediction is similar. This embodiment mainly uses a pre-trained LSTM model for temperature prediction. Specifically, the LSTM model of this embodiment includes an input layer, an LSTM layer, a Dropout layer, a fully connected layer, and an output layer, wherein the input layer is used to receive environmental data and battery status data. The data received in this embodiment includes SOC, SOH, charge and discharge current, ambient temperature, temperature change rate, wind speed, solar radiation intensity, and relative humidity. Therefore, the input layer is provided with 8 input units, which respectively receive 8 types of special input data; 64 LSTM units, i.e., 64 neurons, are provided in the LSTM layer. Each LSTM unit can perform complex nonlinear transformation and information processing on the input sequence. The 64 units work in parallel, which enhances the network's ability to learn complex patterns and long-term dependencies in the input data.

[0055] Dropout is a regularization technique used to prevent neural networks from overfitting. In this embodiment, the Dropout layer is set to 0.2, which means that during the training process, this layer will randomly set the activation values of some neurons output by the LSTM layer to 0 with a probability of 0.2 (or 20%). This is equivalent to randomly "dropping" some neurons, so that the network cannot overly rely on certain specific neurons during each training, thereby enhancing the network's generalization ability and avoiding overfitting problems where the model performs well on the training set but performs poorly on new data such as the test set.

[0056] The fully connected layer has 32 neurons. Each neuron in the fully connected layer is connected to all output neurons in the dropout layer. These neurons perform linear operations such as weighted summation on the features processed by dropout. They also introduce nonlinearity through activation functions (such as ReLU) to map the data to a new feature space, further refining and compressing the features in preparation for the output layer. In this embodiment, a single neuron is used in the output layer to output the predicted temperature.

[0057] In addition, before using the LSTM model for temperature prediction, it needs to be trained first. Specifically, historical relevant data can be used as a training set to complete the training process. During training, the hyperparameters used in this embodiment are epoch = 200 and batch_size = 32. Among them, epoch represents the number of training rounds. In each round of training, the model predicts the input data based on the current parameters, and then calculates the error between the prediction result and the true label. Then, the parameters of the model are adjusted through the back propagation algorithm to reduce the error. After 200 rounds of training in this embodiment, the model will gradually learn the patterns and rules in the data, thereby improving the accuracy of the prediction. Batch_size represents the batch size. Specifically, during the training process, due to the limitations of memory and computing resources, it is usually impossible to input the entire training data set into the model for training at one time. Therefore, this embodiment divides the training set into 32 small batches, and only processes one small batch of data at a time.

[0058] Specifically, the loss function used in this embodiment for LSTM model training includes MAE (Mean Absolute Error) and a temperature gradient penalty term. The temperature gradient penalty term implements feedback optimization of the model by establishing an error compensation matrix. In the error compensation matrix, the model weights are updated by using a reinforcement learning algorithm. The error compensation matrix can be specifically expressed as follows:

[0059] W new =W old +η·(T real -T pred )·Xinput

[0060] Where W new represents the updated weight, W old represents the weight before update, η represents the learning rate, T real Indicates the real-time monitored battery temperature data, T pred represents the battery temperature data predicted by the temperature prediction model, X input The feature vector representing the input of the input layer includes 8 normalized feature vectors. The learning rate is set to 0.001 in this embodiment. The setting of the learning rate can control the parameter update amplitude.

[0061] Step S203: Determine the expected arrival time of extreme weather according to the future temperature data of the battery and the current temperature data in the environmental data; Figure 1 Step S103 of the illustrated embodiment will not be described in detail here.

[0062] Step S204: Determine the pre-adjustment start time based on the expected arrival time of extreme weather, temperature changes, and safety margin time. Specifically, the pre-adjustment start time is determined using the following formula:

[0063]

[0064]

[0065] Where, T start Indicates the pre-regulation start time, T event Indicates the expected arrival time of extreme weather, ΔT req Indicates the target temperature rise, represents the rate of temperature change, t safe represents the safety margin time, α represents the correction coefficient, n represents the number of historical samples, T pred,i Indicates the predicted battery temperature corresponding to the i-th sample, T real,i Indicates the battery temperature monitored by the i-th sample. Among them, the target temperature rise ΔT req This can be set based on the battery type, for example, 10°C for lithium iron phosphate batteries. The target temperature rise specifically represents the difference between the desired safe operating temperature and the current temperature. Historical samples can include both historically monitored temperature data and predicted temperature data.

[0066] In this embodiment, a battery pre-conditioning thermal management method based on multi-source data prediction is provided, and the method includes the following steps:

[0067] Step S301: Obtain environmental data and battery status data; see Figure 1 Step S101 of the illustrated embodiment will not be described in detail here.

[0068] Step S302: Use the temperature prediction model to predict the battery's future temperature data based on the environmental data and battery status data. Figure 1 Step S102 of the illustrated embodiment will not be described in detail here.

[0069] Step S303: Determine the expected arrival time of extreme weather according to the future temperature data of the battery and the current temperature data in the environmental data; Figure 1 Step S103 of the illustrated embodiment will not be described in detail here.

[0070] Step S304: Determine the pre-conditioning start time based on the expected arrival time of extreme weather, temperature changes, and safety margin time. Figure 1 Step S104 of the illustrated embodiment will not be described in detail here.

[0071] Step S305: When the predicted future battery temperature is lower than the target temperature, a heating mode is activated. Specifically, when the difference between the predicted future battery temperature and the target temperature is greater than or equal to a fourth preset value, a first power heating mode is activated. When the difference between the predicted future battery temperature and the target temperature is greater than or equal to a fifth preset value and lower than the fourth preset value, a second power heating mode is activated, with the second power lower than the first power. When the difference between the predicted future battery temperature and the target temperature is lower than a fifth preset value, a pulse heating mode is activated.

[0072] Among them, the fourth preset value and the fifth preset value can be set according to actual needs. For example, the fourth preset value is set to 15°C, and the fifth preset value is set to 5°C. The target temperature can be set according to the optimal temperature range of the battery, for example, it can be set to 25±3°C. Therefore, when the difference ΔT1 (ΔT1=Ttarget-Tpred) between the predicted future temperature data of the battery Tpred and the target temperature Ttarget is greater than or equal to 15°C, the high power (such as 3kW / m 2 ) heating mode; when 5℃≤ΔT1<15℃, start medium power mode (such as 1.5kW / m 2 ); when ΔT < 5°C, a pulse heating mode (e.g., a duty cycle of 30%) is activated. During heating, a PTC (positive temperature coefficient thermistor) ceramic heater array may be used for heating, or electromagnetic induction heating may be used for heating, which is not specifically limited in this embodiment.

[0073] Step S306: When the predicted future battery temperature data is greater than the target temperature, the cooling mode is activated. When the cooling mode is used, if the cooling is achieved by circulating ethylene glycol solution through the liquid cooling plate, the flow rate of the coolant is determined using a fuzzy PID method. Specifically, the flow rate is determined using the following formula:

[0074]

[0075] Where Q represents the coolant flow rate (unit: m 3 / s); K p is the proportional coefficient; K i is the integral coefficient (s -1 );K d is the differential coefficient (s); e(t) is the temperature error function, defined as: e(t) = T batt (t)-T target , T batt (t) represents the real-time temperature of the battery collected by the sensor arranged in the battery module, T target Indicates the target temperature, which is dynamically adjusted according to the battery chemical characteristics, such as 15°C during low-temperature charging and 35°C during high-temperature discharge. Indicates the cumulative amount of error over time; Indicates the error change rate (℃ / s).

[0076] In addition, in addition to cooling by circulating ethylene glycol solution through a liquid cooling plate, a cooling method using a compressor coupled with a phase change material can also be used. This embodiment does not limit the specific cooling method.

[0077] Step S307: When the difference between the real-time monitored battery temperature data and the predicted future battery temperature data is greater than a first preset value and the duration is greater than a second preset value, or when the difference between the maximum and minimum temperatures inside the monitored battery is greater than a third preset value, the backup heat source is activated. Specifically, the first preset value, the second preset value, and the third preset value can be determined based on actual needs. For example, the first preset value can be 3°C, the second preset value can be 15 minutes, and the third preset value can be 5°C. Therefore, the activation conditions of the backup heat source can be expressed as:

[0078] |Real-time temperature - predicted temperature|≥3℃ and duration>15 minutes;

[0079] Alternatively, the temperature difference of the single battery is ≥5°C.

[0080] As a specific application example of the embodiment of the present invention, Figure 2 As shown, the battery pre-conditioning thermal management method based on multi-source data prediction includes the following steps:

[0081] 1. Multi-source data fusion.

[0082] (1) Obtain forecast data for the next 72 hours every hour through the Meteorological Bureau API interface;

[0083] (2) Locally deployed temperature and humidity sensors collect environmental data at 5-minute intervals;

[0084] (3) Using the improved Kalman filter algorithm, setting the process noise Q = 0.01 and the observation noise R = 0.1 to achieve data fusion;

[0085] (4) Establish extreme weather warning indicators: trigger a warning when the predicted temperature change rate is ≥3℃ / h and the duration is >2h.

[0086] 2. Battery thermal load prediction.

[0087] (1) The input layer of the LSTM neural network contains 8 nodes: SOC, SOH, charge and discharge current, ambient temperature, temperature change rate, wind speed, solar radiation intensity, and relative humidity;

[0088] (2) Network structure: input layer (8)-LSTM layer (64)-Dropout layer (0.2)-fully connected layer (32)-output layer (1);

[0089] (3) Training parameters: epoch = 200, batch_size = 32, and the loss function uses MAE + temperature gradient penalty term.

[0090] 3. Dynamic time window decision.

[0091] The pre-adjustment start time is calculated based on the temperature change rate, which is specifically expressed by the following formula:

[0092]

[0093] Where: T start is the pre-adjustment start time, T event is the estimated arrival time of extreme weather, ΔT req is the target temperature rise, ΔT / Δt is the temperature change rate, t safe is the safety margin time.

[0094] The safety margin time t safe Dynamic adjustment based on historical errors is specifically expressed using the following formula:

[0095]

[0096] Where α is the correction coefficient, n is the number of historical samples, T pred,i To predict the temperature, T real,i is the actual temperature.

[0097] 4. Multi-objective optimization execution.

[0098] (1) Heating mode: adopts graded PTC heating strategy.

[0099] When ΔT≥15℃, high power mode (3kW / m 2 );

[0100] When 5℃≤ΔT<15℃, start the medium power mode (1.5kW / m 2 );

[0101] When ΔT<5°C, the pulse heating mode (duty cycle 30%) is started.

[0102] (2) Cooling mode: variable flow liquid cooling control. The flow rate is determined by the fuzzy PID algorithm shown in the following formula:

[0103]

[0104] In addition, it should be noted that when selecting specific heating and cooling methods, the power generation surplus and time-of-use electricity prices can be further considered to select the optimal method.

[0105] 5. Real-time monitoring and feedback optimization.

[0106] Reinforcement learning is used to establish the error compensation matrix shown in the following formula as the temperature gradient penalty term:

[0107] W new =W old +η·(T real -T pred )·X input

[0108] Among them, the learning rate η=0.001, the input vector X input Normalized to [0,1].

[0109] Backup heat source startup conditions:

[0110] |Real-time temperature - predicted temperature|≥3℃ and duration>15 minutes;

[0111] or,

[0112] The temperature difference of single battery is ≥5℃.

[0113] This embodiment also provides a battery pre-conditioning thermal management device based on multi-source data prediction. This device is used to implement the above-mentioned embodiments and preferred embodiments, and details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware, is also possible and contemplated.

[0114] This embodiment provides a battery pre-conditioning thermal management device based on multi-source data prediction, such as Figure 3As shown, including:

[0115] A data acquisition module 31 is used to acquire environmental data and battery status data;

[0116] A temperature prediction module 32 is configured to use a temperature prediction model to predict future battery temperature data based on environmental data and battery status data;

[0117] An extreme weather determination module 33 is configured to determine an estimated arrival time of extreme weather based on future temperature data of the battery and current temperature data in the environmental data;

[0118] The pre-adjustment start module 34 is used to determine the pre-adjustment start time according to the expected arrival time of extreme weather, temperature changes and safety margin time.

[0119] In an optional embodiment, the device also includes: a heating mode start-up module, which is used to start the heating mode when the predicted future temperature data of the battery is less than the target temperature; a cooling mode start-up module, which is used to start the cooling mode when the predicted future temperature data of the battery is greater than the target temperature; and a backup heat source start-up module, which is used to start the backup heat source when the difference between the real-time monitored battery temperature data and the predicted future battery temperature data is greater than a first preset value and the duration is greater than a second preset value, or when the difference between the maximum temperature and the minimum temperature inside the monitored battery is greater than a third preset value.

[0120] In an optional embodiment, the data acquisition module is specifically used to: obtain the first environmental data monitored in real time and the second environmental data predicted; use a wavelet transform-support vector machine combined noise reduction algorithm or an improved Kalman filter algorithm to fuse the first environmental data and the second environmental data, the process noise covariance matrix of the improved Kalman filter algorithm is determined according to the data credibility, and the improved Kalman filter algorithm includes a temperature gradient constraint condition; obtain the SOC, SOH and charge and discharge current collected by the battery management system as battery status data.

[0121] In an optional embodiment, the temperature prediction model includes a pre-trained Transformer time series model, a pre-trained LSTM model, or a physical model and a data-driven hybrid model; the pre-trained LSTM model includes an input layer, an LSTM layer, a Dropout layer, a fully connected layer, and an output layer, and the input data of the input layer includes SOC, SOH, charge and discharge current, ambient temperature, temperature change rate, wind speed, solar radiation intensity, and relative humidity; the loss function of the pre-trained long short-term memory network model includes a temperature gradient penalty term.

[0122] In an optional embodiment, the temperature gradient penalty term is expressed using the following formula:

[0123] Wnew =W old +η·(T real -T pred )·X input

[0124] Where W new represents the updated weight, W old represents the weight before update, η represents the learning rate, T real Indicates the real-time monitored battery temperature data, T pred represents the battery temperature data predicted by the temperature prediction model, X input A feature vector representing the input to the input layer.

[0125] In an optional embodiment, the heating mode start module is specifically used to: start the first power heating mode when the difference between the predicted future temperature data of the battery and the target temperature is greater than or equal to a fourth preset value; start the second power heating mode when the difference between the predicted future temperature data of the battery and the target temperature is greater than or equal to a fifth preset value and less than the fourth preset value, and the second power is less than the first power; start the pulse heating mode when the difference between the predicted future temperature data of the battery and the target temperature is less than the fifth preset value; when the predicted future temperature data of the battery is greater than the target temperature, the device also includes: a flow control module, which is used to determine the flow rate of the coolant in the cooling mode according to the fuzzy PID method.

[0126] In an optional embodiment, the pre-adjustment start time is determined using the following formula:

[0127]

[0128] Where, T start Indicates the pre-regulation start time, T event Indicates the expected arrival time of extreme weather, ΔT req Indicates the target temperature rise, represents the rate of temperature change, t safe represents the safety margin time, α represents the correction coefficient, n represents the number of historical samples, T pred,i Indicates the predicted battery temperature corresponding to the i-th sample, T teal,i Indicates the battery temperature monitored by the i-th sample.

[0129] The further functional description of each of the above modules and units is the same as that of the above corresponding embodiments and will not be repeated here.

[0130] The embodiment of the present invention also provides a computer device having the above Figure 3 The battery pre-conditioning thermal management device shown is based on multi-source data prediction.

[0131] See also Figure 4 , Figure 4 is a structural diagram of a computer device provided by an optional embodiment of the present invention, such as Figure 4 As shown, the computer device includes: one or more processors 10, memory 20, and interfaces for connecting various components, including high-speed interfaces and low-speed interfaces. Various components utilize different buses to communicate with each other and can be installed on a common mainboard or installed in other ways as needed. The processor can process the instructions executed in the computer device, including instructions stored in the memory or on the memory to display the graphical information of the GUI on an external input / output device (such as, a display device coupled to the interface). In some optional embodiments, if necessary, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories. Equally, multiple computer devices can be connected, and each device provides part of the necessary operations (for example, as a server array, a group of blade servers, or a multi-processor system). Figure 4 A processor 10 is taken as an example.

[0132] The processor 10 may be a central processing unit, a network processor, or a combination thereof. The processor 10 may further include a hardware chip. The hardware chip may be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The programmable logic device may be a complex programmable logic device, a field programmable gate array, a general purpose array logic, or any combination thereof.

[0133] The memory 20 stores instructions that can be executed by at least one processor 10, so as to enable at least one processor 10 to execute the method shown in the above embodiment.

[0134] The memory 20 may include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function; the data storage area may store data created based on the use of a computer device for displaying a small program landing page, etc. In addition, the memory 20 may include a high-speed random access memory, and may also include a non-transient memory, such as at least one disk storage device, a flash memory device, or other non-transient solid-state storage device. In some optional embodiments, the memory 20 may optionally include a memory remotely located relative to the processor 10, and these remote memories may be connected to the computer device via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0135] The memory 20 may include a volatile memory, such as a random access memory; the memory may also include a non-volatile memory, such as a flash memory, a hard disk or a solid-state drive; the memory 20 may also include a combination of the above types of memory.

[0136] The computer device further includes a communication interface 30 for the computer device to communicate with other devices or a communication network.

[0137] The embodiment of the present invention also provides a computer-readable storage medium. The above-mentioned method according to the embodiment of the present invention can be implemented in hardware, firmware, or implemented as a computer code that can be recorded in a storage medium, or implemented as a computer code that is originally stored in a remote storage medium or a non-temporary machine-readable storage medium and downloaded through a network and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disk, a read-only storage memory, a random access memory, a flash memory, a hard disk or a solid-state drive, etc.; further, the storage medium can also include a combination of the above-mentioned types of memory. It can be understood that a computer, a processor, a microprocessor controller or programmable hardware includes a storage component that can store or receive software or computer code. When the software or computer code is accessed and executed by a computer, a processor or hardware, the method shown in the above embodiment is implemented.

[0138] A portion of the present invention may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present invention through the operation of the computer. Those skilled in the art should understand that the form in which the computer program instruction exists in a computer-readable medium includes, but is not limited to, a source file, an executable file, an installation package file, etc. Accordingly, the way in which the computer program instruction is executed by the computer includes, but is not limited to: the computer directly executes the instruction, or the computer compiles the instruction and then executes the corresponding compiled program, or the computer reads and executes the instruction, or the computer reads and installs the instruction and then executes the corresponding installed program. Here, the computer-readable medium may be any available computer-readable storage medium or communication medium that can be accessed by the computer.

[0139] Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention. Such modifications and variations are all within the scope defined by the appended claims.

Claims

1. A battery pre-conditioning thermal management method based on multi-source data prediction, characterized in that: The method comprises: Obtain environmental data and battery status data; Use a temperature prediction model to predict the battery's future temperature data based on environmental data and battery status data; Determine the expected arrival time of extreme weather based on future temperature data of the battery and current temperature data in environmental data; The pre-adjustment start time is determined according to the expected arrival time of the extreme weather, temperature changes, and safety margin time.

2. The method according to claim 1, characterized in that The method further comprises: When the predicted future battery temperature data is lower than the target temperature, the heating mode is started; When the predicted future battery temperature data is greater than the target temperature, the cooling mode is started; When the difference between the real-time monitored battery temperature data and the predicted future battery temperature data is greater than a first preset value and the duration is greater than a second preset value, or when the difference between the maximum and minimum temperatures inside the monitored battery is greater than a third preset value, the backup heat source is started.

3. The method according to claim 1, characterized in that Obtain environmental data and battery status data, including: Acquiring first environmental data monitored in real time and second environmental data predicted; The first environmental data and the second environmental data are fused using a wavelet transform-support vector machine combined noise reduction algorithm or an improved Kalman filter algorithm, wherein the process noise covariance matrix of the improved Kalman filter algorithm is determined according to the data credibility, and the improved Kalman filter algorithm includes a temperature gradient constraint condition; The SOC, SOH, and charge and discharge current collected by the battery management system are obtained as battery status data.

4. The method according to claim 1, wherein The temperature prediction model includes a pre-trained Transformer time series model, a pre-trained LSTM model, or a hybrid model of a physical model and a data-driven model; the pre-trained LSTM model includes an input layer, an LSTM layer, a Dropout layer, a fully connected layer, and an output layer, and the input data of the input layer includes SOC, SOH, charge and discharge current, ambient temperature, temperature change rate, wind speed, solar radiation intensity, and relative humidity; the loss function of the pre-trained long short-term memory network model includes a temperature gradient penalty term.

5. The method according to claim 4, characterized in that The temperature gradient penalty term is expressed as follows: W new =W old +η·(T real -T pred )·X input Where W new represents the updated weight, W old represents the weight before update, η represents the learning rate, T real Indicates the real-time monitored battery temperature data, T pred represents the battery temperature data predicted by the temperature prediction model, X input A feature vector representing the input to the input layer.

6. The method according to claim 2, characterized in that When the predicted future battery temperature data is lower than the target temperature, the heating mode is started, including: When the difference between the predicted future battery temperature data and the target temperature is greater than or equal to a fourth preset value, starting the first power heating mode; When the difference between the predicted future battery temperature data and the target temperature is greater than or equal to a fifth preset value and less than a fourth preset value, a second power heating mode is activated, where the second power is less than the first power; When the difference between the predicted future battery temperature data and the target temperature is less than a fifth preset value, starting the pulse heating mode; When the predicted future temperature data of the battery is greater than the target temperature, after starting the cooling mode, the method further includes: The flow rate of coolant in cooling mode is determined according to the fuzzy PID method.

7. The method according to claim 1, characterized in that The pre-conditioning start time is determined using the following formula: Where, T start Indicates the pre-regulation start time, T event Indicates the expected arrival time of extreme weather, ΔT req Indicates the target temperature rise, represents the rate of temperature change, t safe represents the safety margin time, α represents the correction coefficient, n represents the number of historical samples, T pred,i Indicates the predicted battery temperature corresponding to the i-th sample, T real,i Indicates the battery temperature monitored by the i-th sample.

8. A battery pre-conditioning thermal management device based on multi-source data prediction, characterized in that: The device comprises: Data acquisition module, used to obtain environmental data and battery status data; A temperature prediction module is used to predict the future temperature data of the battery based on environmental data and battery status data using a temperature prediction model; An extreme weather determination module is used to determine the expected arrival time of extreme weather based on the future temperature data of the battery and the current temperature data in the environmental data; The pre-regulation start module is used to determine the pre-regulation start time according to the expected arrival time of the extreme weather, temperature changes and safety margin time.

9. A computer device, characterized in that: include: A memory and a processor, wherein the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the battery pre-conditioning thermal management method based on multi-source data prediction according to any one of claims 1 to 7 by executing the computer instructions.

10. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute the battery pre-conditioning thermal management method based on multi-source data prediction according to any one of claims 1 to 7.

Images