Power battery safety state prediction method based on multi-level model structure

Abstract

A power battery safety state prediction method, apparatus and device based on a multilevel structure model, and a storage medium, the method comprising: according to a received prediction instruction, obtaining to-be-predicted data within a first preset duration before a current moment, the to-be-predicted data comprising power battery state information and vehicle parameter information; based on a preset multi-layer structure model, according to the power battery state information and the vehicle parameter information, safety state information of the power battery within a first preset duration after the current moment is predicted, and the preset multi-layer structure model comprises a battery state model, a driving behavior evaluation model and a power battery safety prediction model; the technical problems that safety monitoring is carried out by adopting internal electrochemical parameters of the power battery, the detection process is complex, the procedure is tedious, and the health state of the power battery cannot be simply and rapidly judged in the related technology are solved, the speed of judging the health state of the power battery is increased, and the accuracy is also improved.

Description

Technical Field

[0001] The present application relates to the field of battery safety, and specifically to a method, device, equipment and computer-readable storage medium for predicting the safety status of a power battery based on a multi-level model structure. Background Art

[0002] There are many reasons for accidents involving new energy vehicles. They may be sudden accidents or the result of gradually accumulated risks. However, any accident problem will be reflected at the data level, that is, the risks of new energy vehicles will definitely be reflected at the data level. Therefore, if the safety of the vehicle can be determined through data, the accident rate can be greatly reduced. During the operation of electric vehicles, the operating conditions are extremely complex, and there may be more accidental changes in voltage and more noise. Therefore, it is not easy to measure the health status of the power battery during operation. Although the internal electrochemical parameters of the power battery are currently used for safety monitoring, in the daily use of the power battery, the internal electrochemical data cannot be inspected due to safety issues, and the detection process is complicated and the procedures are cumbersome, so the health status of the power battery cannot be simply and quickly determined. Summary of the invention

[0003] The present application provides a method, device, equipment and computer-readable storage medium for predicting the safety status of a power battery based on a multi-level structural model, which can solve the technical problem in the prior art that internal electrochemical parameters of the power battery are used for safety monitoring, but the detection process is complicated, the procedures are cumbersome, and the health status of the power battery cannot be judged simply and quickly.

[0004] In a first aspect, an embodiment of the present application provides a method for predicting a power battery safety state based on a multi-level structure model. The method for predicting a power battery safety state based on a multi-level structure model includes:

[0005] According to the received prediction instruction, obtaining the data to be predicted within a first preset time period before the current moment, wherein the data to be predicted includes power battery status information and vehicle parameter information;

[0006] Based on a preset multi-level structure model, according to the power battery status information and vehicle parameter information, the safety status information of the power battery within a first preset time after the current moment is predicted, wherein the preset multi-level structure model includes a battery status model, a driving behavior evaluation model and a power battery safety prediction model.

[0007] In combination with the first aspect, in one implementation, predicting the safety status information of the power battery within a first preset time after the current moment according to the power battery status information and the vehicle parameter information includes:

[0008] Inputting the power battery status information into the battery status model, and obtaining the property information of the power battery within a first preset time period after the current moment when the battery status model outputs the power battery, wherein the property information of the power battery includes a temperature value, a voltage value, a current value, a SOC value, a SOH value, a cumulative discharge power value, a cumulative charge power value, and an insulation resistance value of the power battery;

[0009] Inputting the vehicle parameter information into the driving behavior evaluation model, obtaining the driving classification result output by the driving behavior evaluation model within a first preset time after the current moment, wherein the driving classification result includes aggressive and stable;

[0010] The property information of the power battery within a first preset time after the current moment and the driving classification result within the first preset time after the current moment are input into the power battery safety prediction model to obtain the safety status information of the power battery within the first preset time after the current moment.

[0011] In combination with the first aspect, in one embodiment, the power battery status information includes: temperature value, voltage value, current value, SOC value, SOH value, cumulative discharge power value, cumulative charge power value, insulation resistance value, charge / discharge identification value, vehicle speed, battery type, vehicle model, ambient temperature value, wind speed, weather temperature value, and road grade of the power battery within a first preset time period before the current moment.

[0012] In combination with the first aspect, in one embodiment, the vehicle parameter information includes: the maximum value, average value, median value, number of sudden accelerations, number of sudden decelerations, number of sudden turns and idling time of vehicle speed, acceleration, steering wheel angle, brake pedal opening, accelerator pedal opening within a first preset time period before the current moment.

[0013] In combination with the first aspect, in one embodiment, after predicting the safety status information of the power battery within a first preset time after the current moment according to the power battery status information and the vehicle parameter information, the method further includes:

[0014] Based on the safety status information, determining whether the power battery is in a safe state within a first preset time after the current moment;

[0015] If the power battery is not in a safe state within a first preset time period after the current moment, a safety warning message is sent.

[0016] In combination with the first aspect, in one implementation, before acquiring the to-be-predicted data within a first preset time period before the current moment according to the received prediction instruction, the method further includes:

[0017] Collecting data to be trained, wherein the data to be trained includes power battery status information, vehicle information, and marked safety status information at each moment, the power battery status information includes power battery status information before the same time period at each moment and power battery status information after the same time period, and the vehicle information includes vehicle parameter information and driving classification results at each moment;

[0018] The first initial model is trained by the power battery attribute information before the same time period at each moment and the power battery attribute information after the same time period to generate a battery state model, wherein the first initial model includes an LSTM layer, a Dropout layer and a fully connected layer;

[0019] Training the second initial model through the vehicle parameter information and driving classification results at each moment to generate a driving behavior evaluation model, wherein the second initial model includes a K-Means clustering model and a boosted tree classifier;

[0020] The third initial model is trained by the driving classification results at each moment, the power battery attribute information after the same period of time, and the marked safety status information to generate a power battery safety prediction model, wherein the third initial model includes an LSTM layer, a Dropout layer, and a fully connected layer.

[0021] In combination with the first aspect, in one implementation, after the third initial model is trained to generate a power battery safety prediction model, the method further includes:

[0022] Testing the battery state model, the driving behavior evaluation model and the power battery safety prediction model respectively to obtain corresponding test values;

[0023] According to the test value, it is determined whether to perform secondary training on the battery state model, the driving behavior evaluation model and the power battery safety prediction model.

[0024] In a second aspect, an embodiment of the present application provides a power battery safety state prediction device based on a multi-level structure model, characterized in that the power battery safety state prediction device based on the multi-level structure model includes:

[0025] An acquisition module, configured to acquire, according to the received prediction instruction, data to be predicted within a first preset time period before the current moment, wherein the data to be predicted includes power battery status information and vehicle parameter information;

[0026] A prediction module is used to predict the safety status information of the power battery within a first preset time after the current moment based on a preset multi-level structure model and according to the power battery status information and vehicle parameter information, wherein the preset multi-level structure model includes a battery status model, a driving behavior evaluation model and a power battery safety prediction model.

[0027] In the third aspect, an embodiment of the present application provides a power battery safety status prediction device based on a multi-level structure model, wherein the power battery safety status prediction device based on the multi-level structure model comprises a processor, a memory, and a power battery safety status prediction program based on the multi-level structure model stored in the memory and executable by the processor, wherein when the power battery safety status prediction program based on the multi-level structure model is executed by the processor, the steps of the power battery safety status prediction method based on the multi-level structure model as described above are implemented.

[0028] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, on which is stored a power battery safety status prediction program based on a multi-level structure model. When the power battery safety status prediction program based on the multi-level structure model is executed by a processor, the steps of the power battery safety status prediction method based on the multi-level structure model as described above are implemented.

[0029] The beneficial effects brought by the technical solution provided by the embodiments of the present application include:

[0030] By obtaining the data to be predicted within a first preset time period before the current moment according to the received prediction instruction, wherein the data to be predicted includes power battery status information and vehicle parameter information; based on a preset multi-level structure model, according to the power battery status information and vehicle parameter information, predicting the safety status information of the power battery within a first preset time period after the current moment, wherein the preset multi-level structure model includes a battery status model, a driving behavior evaluation model and a power battery safety prediction model, the technical problem that the related technology uses internal electrochemical parameters of the power battery for safety monitoring, but the detection process is complicated, the procedure is cumbersome, and the health status of the power battery cannot be simply and quickly determined is solved, which not only improves the rate of determining the health status of the power battery, but also improves the accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a flow chart of an embodiment of a method for predicting the safety status of a power battery based on a multi-level structure model of the present application;

[0032] Figure 2 This is a functional module diagram of an embodiment of a power battery safety status prediction device based on a multi-level structure model of the present application;

[0033] Figure 3 This is a schematic diagram of the hardware structure of a power battery safety status prediction device based on a multi-level structure model involved in the embodiment of the present application. DETAILED DESCRIPTION

[0034] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution 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 part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0035] First, some technical terms in the present application are explained to facilitate those skilled in the art to understand the present application.

[0036] In order to make the objectives, technical solutions and advantages of the present application clearer, the implementation methods of the present application will be further described in detail below in conjunction with the accompanying drawings.

[0037] In a first aspect, an embodiment of the present application provides a method for predicting the safety status of a power battery based on a multi-level structure model.

[0038] In one embodiment, referring to Figure 1 , Figure 1 This is a flow chart of the first embodiment of the power battery safety status prediction method based on the multi-level structure model of this application. Figure 1 As shown, the power battery safety status prediction method based on the multi-level structure model includes:

[0039] Step S10: acquiring the data to be predicted within a first preset time period before the current moment according to the received prediction instruction, wherein the data to be predicted includes power battery status information and vehicle parameter information;

[0040] Exemplarily, when a prediction instruction is received, the data to be predicted within the first preset time before the current moment is obtained. For example, when the safety status information of the power battery at 10 o'clock is received and the current moment is 9 o'clock, the data to be predicted within one hour between 8 o'clock and 9 o'clock is obtained, and the data to be predicted includes the power battery status information and vehicle parameter information between 8 o'clock and 9 o'clock. The power battery status information includes temperature value, voltage value, current value, SOC value, SOH value, cumulative discharge power value, cumulative charge power value and / or insulation resistance value; the vehicle parameter information includes vehicle speed, acceleration, steering wheel angle, brake pedal opening, accelerator pedal opening, maximum value, average value, median, number of sudden accelerations, number of sudden decelerations, number of sudden turns and / or idling time.

[0041] Step S20: Based on a preset multi-level structure model, according to the power battery status information and vehicle parameter information, predict the safety status information of the power battery within a first preset time period after the current moment, wherein the preset multi-level structure model includes a battery status model, a driving behavior evaluation model and a power battery safety prediction model.

[0042] Exemplarily, the preset multi-level structure model includes a battery state model, a driving behavior evaluation model and a power battery safety prediction model. By inputting the temperature value, voltage value, current value, SOC value, SOH value, cumulative discharge power value, cumulative charge power value and / or insulation resistance value into the battery state model, the attribute information of the power battery output by the battery state model is obtained; the maximum value, average value, median value, number of sudden accelerations, number of sudden decelerations, number of sudden turns and / or idling time of vehicle speed, acceleration, steering wheel angle, brake pedal opening, accelerator pedal opening, is input into the driving behavior evaluation model to obtain the driving classification result output by the driving behavior evaluation model. Then, the attribute information of the power battery and the driving classification result are input into the power battery safety prediction model to obtain the safety state information output by the power battery safety prediction model, thereby obtaining the safety state information of the power battery from 9 o'clock to 10 o'clock.

[0043] Specifically, predicting the safety status information of the power battery within a first preset time after the current moment according to the power battery status information and the vehicle parameter information includes:

[0044] The power battery status information is input into the battery status model, and the attribute information of the power battery within the first preset time after the current moment output by the battery status model is obtained, wherein the attribute information of the power battery includes the temperature value, voltage value, current value, SOC value, SOH value, cumulative discharge power value, cumulative charge power value and insulation resistance value of the power battery; the vehicle parameter information is input into the driving behavior evaluation model, and the driving classification result within the first preset time after the current moment output by the driving behavior evaluation model is obtained, wherein the driving classification result includes aggressive type and stability; the attribute information of the power battery within the first preset time after the current moment and the driving classification result within the first preset time after the current moment are input into the power battery safety prediction model, and the safety status information of the power battery within the first preset time after the current moment is obtained.

[0045] Exemplarily, the temperature value, voltage value, current value, SOC value, SOH value, cumulative discharge power value, cumulative charge power value, insulation resistance value, charge / discharge identification value, vehicle speed, battery type, vehicle model, ambient temperature value, wind speed, weather temperature value and road grade of the power battery within the first preset time before the current moment are input into the battery state model to obtain the property information of the power battery output by the battery state model, and the power battery property information includes temperature value, voltage value, current value, SOC value, SOH value, cumulative discharge power value, cumulative charge power value and insulation resistance value. The maximum value, average value, median value, number of sudden accelerations, number of sudden decelerations, number of sudden turns and idling time of the vehicle speed, acceleration, steering wheel angle, brake pedal opening, accelerator pedal opening within the first preset time before the current moment are input into the driving behavior evaluation model to obtain the driving classification result output by the driving behavior evaluation model, and the driving classification result includes aggressive type and stability. Then, the temperature value, voltage value, current value, SOC value, SOH value, cumulative discharge power value, cumulative charge power value and insulation resistance value of the power battery within the first preset time after the current moment, as well as the aggressiveness or stability within the first preset time after the previous moment are input into the power battery safety prediction model to predict the safety status information of the power battery within the first preset time after the current moment.

[0046] After obtaining the safety status information of the power battery within the first preset time after the current moment, determine whether the power battery within the first preset time after the current moment is in a safe state; if the power battery within the first preset time after the current moment is not in a safe state, send a safety warning message. For example, the safety status information of the power battery within the first preset time after the current moment is a safety status value, and the safety status value is compared with the preset safety status value. If the safety status value is less than the preset safety status value, it is determined that the power battery within the first preset time after the current moment is in an unsafe state, and a safety warning message is sent; if the safety status value is greater than or equal to the preset safety status value, it is determined that the power battery within the first preset time after the current moment is in a safe state.

[0047] Specifically, before obtaining the data to be predicted within the first preset time period before the current moment according to the received prediction instruction, the method further includes: collecting data to be trained, wherein the data to be trained includes power battery status information, vehicle information and marked safety status information at each moment, the power battery status information includes power battery status information before the same time period at each moment and power battery status information after the same time period, and the vehicle information includes vehicle parameter information and driving classification results at each moment; training the first initial model through the power battery attribute information before the same time period at each moment and the power battery attribute information after the same time period, generating A battery status model is generated, wherein the first initial model includes an LSTM layer, a Dropout layer and a fully connected layer; the second initial model is trained through the vehicle parameter information and the driving classification results at each moment to generate a driving behavior evaluation model, wherein the second initial model includes a K-Means clustering model and a boosted tree classifier; the third initial model is trained through the driving classification results at each moment, the power battery attribute information after the same length of time and the marked safety status information to generate a power battery safety prediction model, wherein the third initial model includes an LSTM layer, a Dropout layer and a fully connected layer.

[0048] Exemplarily, the data to be trained is collected, wherein the data to be trained includes the power battery status information, vehicle information and marked safety status information at each moment, the power battery status information includes the power battery status information before the same time length at each moment and the power battery status information after the same time length, and the vehicle information includes the vehicle parameter information and driving classification results at each moment. The power battery status information before the same time length at each moment and the power battery status information after the same time length are X. X includes the fields: battery temperature, battery voltage, current, SOC, SOH, cumulative discharge power, cumulative charge power, insulation resistance, charge / discharge identification, vehicle speed, battery type, vehicle model, ambient temperature, wind speed, weather, road grade. Determine the output Y of the data to be trained. Y includes the fields: battery temperature, battery voltage, current, SOC, SOH, cumulative discharge power, cumulative charge power, insulation resistance. Clean or fill in abnormal values ​​and missing values ​​for [X,Y]. When only one frame has an abnormal value or missing value, and both the previous and next frames of the frame have data, the linear difference method is used to fill the middle value; when there are abnormal values ​​or missing values ​​in two or more consecutive frames, the signals at these time points are deleted. Preferably, the data is normalized and one-hot encoded after data cleaning.

[0049] Standardize X. Assume that the time interval between each two frames of the acquired signal is T, and divide X into windows according to the set sequence length S. For example, S is set to 10, which means that each window consists of 10 frames of continuous data. S indicates how long the state is traced back to predict the state at the next moment in the future. After dividing the window, the dimension of the data in each window is (m, p), where m is the length of the set sequence and p is the number of state quantities contained in X. It should be noted that the signal sequence in each window must be sorted from early to late in time and the time interval between the two frames must be consistent. The dimension of X after standardization is (b, m, p), where b represents the number of divided windows, m is the length of the set sequence, and p is the number of state quantities contained in X.

[0050] Standardize Y. Perform first-order differences on all fields in Y, such as dv=vi+1-vi, where vi+1 is the voltage value at time i+1 and vi is the voltage value at time i. In (4), X is divided into windows with a sequence length of S. Let the last frame of a window Wi divided by X correspond to time t, then the training output yi corresponding to the window W should be the corresponding state quantity change within a period of time n after t, so as to configure the corresponding target value for each window of X. The dimension of Y after standardization is (b,n,q), where b represents the number of divided windows, n is the corresponding predicted time series length, which indicates how long the state is expected to be predicted, and q is the number of state quantities contained in Y.

[0051] The first initial model is trained by using the standardized power battery status information before the same time period and the power battery status information after the same time period to generate a battery status model, wherein the first initial model includes an LSTM layer, a Dropout layer and a fully connected layer.

[0052] The second initial model is trained by using the vehicle parameter information and driving classification results at each moment to generate a driving behavior evaluation model, wherein the second initial model includes a K-Means clustering model and a boosted tree classifier.

[0053] The third initial model is trained through the driving classification results at each moment, the power battery attribute information after the same period of time, and the marked safety status information to generate a power battery safety prediction model, wherein the third initial model includes an LSTM layer, a Dropout layer, and a fully connected layer, and the power battery attribute information after the same period of time includes the following fields: battery temperature, battery voltage, current, SOC, SOH, cumulative discharge power, cumulative charge power, insulation resistance, charge / discharge identification, vehicle speed, battery type, vehicle model, and recent driving behavior classification results calculated by M2.

[0054] Determine the safety status information Y of the output mark of the training data. Clean or fill in the outliers and missing values ​​of the power battery attribute information after the same time length. When there are outliers or missing values ​​in only one frame, and there are data in both the front and back needles of the frame data, the linear difference method is used to fill in the middle value; when there are outliers or missing values ​​in two or more consecutive frames, delete the signals at these time points. Preferably, the data is normalized and one-hot encoded after data cleaning.

[0055] Standardize the power battery attribute information after the same duration. Assume that the time interval between each two frames of the collected signal is T, and divide the power battery attribute information after the same duration into windows according to the set sequence length S. For example, S is set to 10, indicating that each window consists of 10 frames of continuous data. S indicates how long the state is traced back to predict the state at the next moment in the future. After dividing the window, the dimension of the data in each window is (m, p), where m is the length of the set sequence, and n is the number of state quantities contained in the power battery attribute information after the same duration. It should be noted that the signal sequence in each window must be sorted from early to late in time and the time interval between the two frames must be consistent. The dimension of the power battery attribute information after the same duration after standardization is (b, m, p), where b represents the number of divided windows, m is the length of the set sequence, n is the corresponding predicted time series length, which indicates how long the state is expected to be predicted, and p is the number of state quantities contained in the power battery attribute information after the same duration. Standardize Y. The training label yi corresponding to each window of power battery attribute information after the same duration is the battery safety status within a period of time after the last frame of the power battery attribute information window after the same duration. The dimension of Y after standardization is (b, 1, 1) where b represents the number of divided windows.

[0056] The battery status model, driving behavior evaluation model and power battery safety prediction model are tested respectively to obtain corresponding test values; according to the test values, it is determined whether to perform secondary training on the battery status model, driving behavior evaluation model and power battery safety prediction model.

[0057] In this embodiment, by obtaining the data to be predicted within a first preset time period before the current moment according to the received prediction instruction, wherein the data to be predicted includes power battery status information and vehicle parameter information; based on a preset multi-level structure model, according to the power battery status information and vehicle parameter information, the safety status information of the power battery within a first preset time period after the current moment is predicted, wherein the preset multi-level structure model includes a battery status model, a driving behavior evaluation model and a power battery safety prediction model, which solves the technical problem in the related art that the internal electrochemical parameters of the power battery are used for safety monitoring, but the detection process is complicated, the procedures are cumbersome, and the health status of the power battery cannot be simply and quickly determined. Not only the rate of determining the health status of the power battery is improved, but also the accuracy is improved.

[0058] In a second aspect, an embodiment of the present application also provides a power battery safety status prediction device based on a multi-level structure model.

[0059] In one embodiment, referring to Figure 2 , Figure 2 This is a functional module diagram of an embodiment of a power battery safety status prediction device based on a multi-level structure model of this application. Figure 2 As shown, the power battery safety state prediction device based on the multi-level structure model includes:

[0060] The acquisition module 10 is used to acquire the data to be predicted within a first preset time period before the current moment according to the received prediction instruction, wherein the data to be predicted includes power battery status information and vehicle parameter information;

[0061] The prediction module 20 is used to predict the safety status information of the power battery within a first preset time period after the current moment based on a preset multi-level structure model and according to the power battery status information and vehicle parameter information, wherein the preset multi-level structure model includes a battery status model, a driving behavior evaluation model and a power battery safety prediction model.

[0062] Furthermore, in one embodiment, the prediction module 20 is used to:

[0063] Inputting the power battery status information into the battery status model, and obtaining the property information of the power battery within a first preset time period after the current moment when the battery status model outputs the power battery, wherein the property information of the power battery includes a temperature value, a voltage value, a current value, a SOC value, a SOH value, a cumulative discharge power value, a cumulative charge power value, and an insulation resistance value of the power battery;

[0064] Inputting the vehicle parameter information into the driving behavior evaluation model, obtaining the driving classification result output by the driving behavior evaluation model within a first preset time after the current moment, wherein the driving classification result includes aggressive and stable;

[0065] The property information of the power battery within a first preset time after the current moment and the driving classification result within the first preset time after the current moment are input into the power battery safety prediction model to obtain the safety status information of the power battery within the first preset time after the current moment.

[0066] Furthermore, in one embodiment, the power battery safety status prediction device based on the multi-level structure model also includes a new module for: the power battery status information includes: the temperature value, voltage value, current value, SOC value, SOH value, cumulative discharge power value, cumulative charging power value, insulation resistance value, charge / discharge identification value, vehicle speed, battery type, vehicle model, ambient temperature value, wind speed, weather temperature value, and road grade of the power battery within a first preset time period before the current moment.

[0067] Furthermore, in one embodiment, the power battery safety status prediction device based on the multi-level structure model also includes a new module for: the vehicle parameter information includes: the maximum value, average value, median value, number of sudden accelerations, number of sudden decelerations, number of sudden turns and idling time of vehicle speed, acceleration, steering wheel angle, brake pedal opening, accelerator pedal opening within a first preset time period before the current moment.

[0068] Furthermore, in one embodiment, the power battery safety state prediction device based on the multi-level structure model further includes a new module for: determining whether the power battery is in a safe state within a first preset time after the current moment based on the safety state information;

[0069] If the power battery is not in a safe state within a first preset time period after the current moment, a safety warning message is sent.

[0070] Furthermore, in one embodiment, the power battery safety state prediction device based on the multi-level structure model also includes a new module for:

[0071] Collecting data to be trained, wherein the data to be trained includes power battery status information, vehicle information, and marked safety status information at each moment, the power battery status information includes power battery status information before the same time period at each moment and power battery status information after the same time period, and the vehicle information includes vehicle parameter information and driving classification results at each moment;

[0072] The first initial model is trained by the power battery attribute information before the same time period at each moment and the power battery attribute information after the same time period to generate a battery state model, wherein the first initial model includes an LSTM layer, a Dropout layer and a fully connected layer;

[0073] Training the second initial model through the vehicle parameter information and driving classification results at each moment to generate a driving behavior evaluation model, wherein the second initial model includes a K-Means clustering model and a boosted tree classifier;

[0074] The third initial model is trained by the driving classification results at each moment, the power battery attribute information after the same period of time, and the marked safety status information to generate a power battery safety prediction model, wherein the third initial model includes an LSTM layer, a Dropout layer, and a fully connected layer.

[0075] Furthermore, in one embodiment, the power battery safety state prediction device based on the multi-level structure model also includes a new module for:

[0076] Testing the battery state model, the driving behavior evaluation model and the power battery safety prediction model respectively to obtain corresponding test values;

[0077] According to the test value, it is determined whether to perform secondary training on the battery state model, the driving behavior evaluation model and the power battery safety prediction model.

[0078] Among them, the functional implementation of each module in the above-mentioned power battery safety status prediction device based on the multi-level structure model corresponds to the various steps in the above-mentioned power battery safety status prediction method embodiment based on the multi-level structure model, and its functions and implementation processes will not be repeated here one by one.

[0079] In a third aspect, an embodiment of the present application provides a power battery safety status prediction device based on a multi-level structure model. The power battery safety status prediction device based on the multi-level structure model can be a personal computer (PC), a laptop computer, a server, or other device with data processing capabilities.

[0080] Reference Figure 3 , Figure 3 The hardware structure diagram of the power battery safety state prediction device based on the multi-level structure model involved in the embodiment of the present application is shown in FIG. In the embodiment of the present application, the power battery safety state prediction device based on the multi-level structure model may include a processor, a memory, a communication interface and a communication bus.

[0081] The communication bus may be of any type and is used to interconnect the processor, the memory, and the communication interface.

[0082] The communication interface includes an input / output (I / O) interface, a physical interface, and a logical interface, etc., which are used to realize the interconnection of devices inside the power battery safety status prediction device based on the multi-level structure model, and an interface for realizing the interconnection between the power battery safety status prediction device based on the multi-level structure model and other devices (such as other computing devices or user devices). The physical interface can be an Ethernet interface, an optical fiber interface, an ATM interface, etc.; the user device can be a display (Display), a keyboard (Keyboard), etc.

[0083] The memory can be various types of storage media, such as random access memory (RAM), read-only memory (ROM), non-volatile RAM (NVRAM), flash memory, optical storage, hard disk, programmable ROM (PROM), erasable PROM (EPROM), electrically erasable PROM (EEPROM), etc.

[0084] The processor may be a general-purpose processor, which may call a power battery safety state prediction program based on a multi-level structure model stored in a memory, and execute a power battery safety state prediction method based on a multi-level structure model provided in an embodiment of the present application. For example, the general-purpose processor may be a central processing unit (CPU). Among them, the method executed when the power battery safety state prediction program based on a multi-level structure model is called may refer to the various embodiments of the power battery safety state prediction method based on a multi-level structure model of the present application, which will not be repeated here.

[0085] Those skilled in the art will understand that Figure 3 The hardware structure shown in the figure does not constitute a limitation on the present application, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0086] In a fourth aspect, an embodiment of the present application also provides a computer-readable storage medium.

[0087] A power battery safety status prediction program based on a multi-level structure model is stored on a computer-readable storage medium of the present application. When the power battery safety status prediction program based on the multi-level structure model is executed by a processor, the steps of the power battery safety status prediction method based on the multi-level structure model as described above are implemented.

[0088] Among them, the method implemented when the power battery safety status prediction program based on the multi-level structure model is executed can refer to the various embodiments of the power battery safety status prediction method based on the multi-level structure model of the present application, and will not be repeated here.

[0089] It should be noted that the serial numbers of the above-mentioned embodiments of the present application are only for description and do not represent the advantages or disadvantages of the embodiments.

[0090] The terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices. The terms "first", "second" and "third" are used to distinguish different objects, etc., and do not represent a sequence, nor do they limit "first", "second" and "third" to different types.

[0091] In the description of the embodiments of the present application, "exemplary", "for example" or "for example" are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary", "for example" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary", "for example" or "for example" is intended to present related concepts in a specific way.

[0092] In the description of the embodiments of the present application, unless otherwise specified, “ / ” means or, for example, A / B can mean A or B; the “and / or” in the text is merely a description of the association relationship of associated objects, indicating that three relationships may exist, for example, A and / or B can mean: A exists alone, A and B exist at the same time, and B exists alone. In addition, in the description of the embodiments of the present application, “multiple” refers to two or more than two.

[0093] In some processes described in the embodiments of the present application, multiple operations or steps that appear in a specific order are included, but it should be understood that these operations or steps may not be executed in the order in which they appear in the embodiments of the present application or in parallel, and the sequence number of the operation is only used to distinguish the different operations, and the sequence number itself does not represent any execution order. In addition, these processes may include more or fewer operations, and these operations or steps may be executed in sequence or in parallel, and these operations or steps may be combined.

[0094] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, disk, CD) as described above, and includes a number of instructions for a terminal device to execute the methods described in each embodiment of the present application.

[0095] The above are only preferred embodiments of the present application, and are not intended to limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A method for predicting the safety status of a power battery based on a multi-level structure model, characterized in that: The power battery safety state prediction method based on the multi-level structure model includes: According to the received prediction instruction, obtaining the data to be predicted within a first preset time period before the current moment, wherein the data to be predicted includes power battery status information and vehicle parameter information; Based on a preset multi-level structure model, according to the power battery status information and vehicle parameter information, the safety status information of the power battery within a first preset time after the current moment is predicted, wherein the preset multi-level structure model includes a battery status model, a driving behavior evaluation model and a power battery safety prediction model.

2. The method for predicting the safety status of a power battery based on a multi-level structure model according to claim 1, characterized in that: The predicting, based on the power battery status information and the vehicle parameter information, the safety status information of the power battery within a first preset time after the current moment, includes: Inputting the power battery status information into the battery status model, and obtaining the property information of the power battery within a first preset time period after the current moment when the battery status model outputs the power battery, wherein the property information of the power battery includes a temperature value, a voltage value, a current value, a SOC value, a SOH value, a cumulative discharge power value, a cumulative charge power value, and an insulation resistance value of the power battery; Inputting the vehicle parameter information into the driving behavior evaluation model, obtaining the driving classification result output by the driving behavior evaluation model within a first preset time after the current moment, wherein the driving classification result includes aggressive and stable; The property information of the power battery within a first preset time after the current moment and the driving classification result within the first preset time after the current moment are input into the power battery safety prediction model to obtain the safety status information of the power battery within the first preset time after the current moment.

3. The method for predicting the safety status of a power battery based on a multi-level structure model according to claim 1, characterized in that: The power battery status information includes: temperature value, voltage value, current value, SOC value, SOH value, cumulative discharge power value, cumulative charge power value, insulation resistance value, charge / discharge identification value, vehicle speed, battery type, vehicle model, ambient temperature value, wind speed, weather temperature value and road grade of the power battery within a first preset time period before the current moment.

4. The method for predicting the safety status of a power battery based on a multi-level structure model according to claim 1, characterized in that: The vehicle parameter information includes: the maximum value, average value, median value, number of sudden accelerations, number of sudden decelerations, number of sudden turns and idling time of vehicle speed, acceleration, steering wheel angle, brake pedal opening, accelerator pedal opening within a first preset time period before the current moment.

5. The method for predicting the safety status of a power battery based on a multi-level structure model according to claim 1, characterized in that: After predicting the safety status information of the power battery within a first preset time after the current moment according to the power battery status information and the vehicle parameter information, the method further includes: Based on the safety status information, determining whether the power battery is in a safe state within a first preset time after the current moment; If the power battery is not in a safe state within a first preset time period after the current moment, a safety warning message is sent.

6. The method for predicting the safety status of a power battery based on a multi-level structure model according to claim 1, characterized in that: Before obtaining the to-be-predicted data within a first preset time period before the current moment according to the received prediction instruction, the method further includes: Collecting data to be trained, wherein the data to be trained includes power battery status information, vehicle information, and marked safety status information at each moment, the power battery status information includes power battery status information before the same time period at each moment and power battery status information after the same time period, and the vehicle information includes vehicle parameter information and driving classification results at each moment; The first initial model is trained by the power battery attribute information before the same time period at each moment and the power battery attribute information after the same time period to generate a battery state model, wherein the first initial model includes an LSTM layer, a Dropout layer and a fully connected layer; Training the second initial model through the vehicle parameter information and driving classification results at each moment to generate a driving behavior evaluation model, wherein the second initial model includes a K-Means clustering model and a boosted tree classifier; The third initial model is trained by the driving classification results at each moment, the power battery attribute information after the same period of time, and the marked safety status information to generate a power battery safety prediction model, wherein the third initial model includes an LSTM layer, a Dropout layer, and a fully connected layer.

7. The method for predicting the safety status of a power battery based on a multi-level structure model according to claim 6, characterized in that: After the third initial model is trained to generate a power battery safety prediction model, the method further includes: Testing the battery state model, the driving behavior evaluation model and the power battery safety prediction model respectively to obtain corresponding test values; According to the test value, it is determined whether to perform secondary training on the battery state model, the driving behavior evaluation model and the power battery safety prediction model.

8. A power battery safety status prediction device based on a multi-level structure model, characterized in that: The power battery safety state prediction device based on the multi-level structure model includes: An acquisition module, configured to acquire, according to the received prediction instruction, data to be predicted within a first preset time period before the current moment, wherein the data to be predicted includes power battery status information and vehicle parameter information; A prediction module is used to predict the safety status information of the power battery within a first preset time after the current moment based on a preset multi-level structure model and according to the power battery status information and vehicle parameter information, wherein the preset multi-level structure model includes a battery status model, a driving behavior evaluation model and a power battery safety prediction model.

9. A power battery safety status prediction device based on a multi-level structure model, characterized in that: The power battery safety status prediction device based on a multi-level structure model comprises a processor, a memory, and a power battery safety status prediction program based on a multi-level structure model stored in the memory and executable by the processor, wherein when the power battery safety status prediction program based on a multi-level structure model is executed by the processor, the steps of the power battery safety status prediction method based on a multi-level structure model as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a power battery safety status prediction program based on a multi-level structure model, wherein when the power battery safety status prediction program based on the multi-level structure model is executed by a processor, the steps of the power battery safety status prediction method based on a multi-level structure model as described in any one of claims 1 to 7 are implemented.

Images