CN115980586A - Power battery health state prediction method and device and computer equipment - Google Patents
Power battery health state prediction method and device and computer equipment Download PDFInfo
- Publication number
- CN115980586A CN115980586A CN202211500338.4A CN202211500338A CN115980586A CN 115980586 A CN115980586 A CN 115980586A CN 202211500338 A CN202211500338 A CN 202211500338A CN 115980586 A CN115980586 A CN 115980586A
- Authority
- CN
- China
- Prior art keywords
- data
- multivariable
- training
- power battery
- target
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Landscapes
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
Abstract
The application relates to a method and a device for predicting the health state of a power battery and computer equipment, wherein the method for predicting the health state of the power battery comprises the following steps: the method comprises the steps of obtaining target multivariable data of the power battery, processing the target multivariable data to obtain target gray-scale map data corresponding to the target multivariable data, further, analyzing the target gray-scale map data based on a prediction model with complete training to obtain a health state prediction value of the power battery.
Description
Technical Field
The application relates to the technical field of electric automobiles, in particular to a method and a device for predicting the health state of a power battery and computer equipment.
Background
With the enhancement of the practicability and diversity of electric vehicles, the market share of electric vehicles is gradually increased, wherein lithium ion power batteries are widely applied to the production of electric vehicles as power sources and energy storage devices due to the advantages of high energy density, low self-discharge rate and the like, however, the lithium ion power batteries still face the problem of battery performance degradation, and the capacity or power attenuation of the batteries can be caused after long-term use, so that the health state of the power batteries needs to be accurately predicted for the safe operation of the electric vehicles.
In the existing prediction method, relevant data of the health state of a battery is collected from a lithium battery based on data collection equipment, a plurality of relevant data sets are generated according to the health state value corresponding to each batch of relevant data, different health state prediction models are established based on each relevant data set, and the health state of the battery is predicted. Therefore, the prediction method does not consider the strong coupling relationship among a plurality of variable data of the battery, and the strong coupling relationship can interfere the prediction process, so that the accuracy of the health state prediction result of the power battery is reduced.
Aiming at the problem that the health state of the power battery cannot be accurately predicted due to the strong coupling relation among multivariable data in the related technology, no effective solution is provided at present.
Disclosure of Invention
The embodiment provides a method and a device for predicting the state of health of a power battery and computer equipment, so as to solve the problem that the state of health of the power battery cannot be accurately predicted due to a strong coupling relationship among multivariable data in the related art.
In a first aspect, a method for predicting the state of health of a power battery is provided in the present embodiment, and the method includes:
acquiring target multivariable data of the power battery;
processing the target multivariable data to obtain target gray-scale image data corresponding to the target multivariable data;
and analyzing the target gray-scale image data based on a prediction model with complete training to obtain a health state prediction value of the power battery.
In some embodiments, the processing the target multivariate data to obtain target gray-scale map data corresponding to the target multivariate data includes:
acquiring voltage data, current data and temperature data of the power battery from the target multivariable data;
and mapping the voltage data, the current data and the temperature data of the power battery to a preset numerical range, and splicing the mapped voltage data, current data and temperature data to obtain target gray-scale map data corresponding to the target multivariable data.
In some embodiments, before analyzing the target grayscale map data based on the fully trained prediction model to obtain the predicted value of the state of health of the power battery, the method further includes:
generating a confrontation network model based on a multivariable condition with complete training, performing data enhancement on first sample multivariable data, and generating a training data set corresponding to the first sample multivariable data;
and training the prediction model based on the training data set to obtain the prediction model with complete training.
In some embodiments, before the generating a confrontation network model based on the well-trained multivariate condition, and processing the first sample multivariate data to obtain the training data set corresponding to the first sample multivariate data, the method further includes:
and acquiring second sample multivariable data from the battery test database, and training the multivariable condition generation countermeasure network model through the second sample multivariable data to obtain an initial multivariable condition generation countermeasure network model.
In some embodiments, the training the prediction model based on the training data set to obtain the well-trained prediction model includes:
processing the first sample multivariable data after data enhancement in the training data set to generate a gray map data set corresponding to the training data set;
and training the prediction model based on the gray-scale image data set to obtain the prediction model with complete training.
In some embodiments, the training the multivariate conditional generation confrontation network model with the second sample multivariate data to obtain an initial multivariate conditional generation confrontation network model further comprises:
and carrying out transfer learning on the initial multivariable condition generation confrontation network model based on the first sample multivariable data to obtain the multivariable condition generation confrontation network model with complete training.
In some embodiments, before the obtaining the second sample multivariate data from the battery test database and training the multivariate conditional generation countermeasure network model with the second sample multivariate data, the method further comprises:
optimizing a condition generating countermeasure network model through a gated cycle cell network, generating the multivariate condition generating countermeasure network model.
In a second aspect, a power battery state of health prediction device is provided in the present embodiment, the device includes:
the acquisition module is used for acquiring target multivariable data of the power battery;
the processing module is used for processing the target multivariable data to obtain target gray-scale image data corresponding to the target multivariable data;
and the prediction module analyzes the target gray-scale image data based on a prediction model with complete training to obtain a health state prediction value of the power battery.
In a third aspect, in the present embodiment, there is provided a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the processor implements the method for predicting the health status of the power battery according to the first aspect.
In a fourth aspect, in the present embodiment, a storage medium is provided, on which a computer program is stored, which when executed by a processor implements the power battery state of health prediction method of the first aspect.
Compared with the related art, the method, the device and the computer equipment for predicting the health state of the power battery provided in the embodiment are used for obtaining the target multivariable data of the power battery, processing the target multivariable data to obtain the target gray-scale map data corresponding to the target multivariable data, further, analyzing the target gray-scale map data based on a well-trained prediction model to obtain the predicted value of the health state of the power battery, solving the problem that the health state of the power battery cannot be accurately predicted due to the strong coupling relation among the multivariable data, and realizing the improvement of the accuracy of the prediction result of the health state of the power battery.
The details of one or more embodiments of the application are set forth in the accompanying drawings and the description below to provide a more thorough understanding of the application.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:
fig. 1 is a block diagram of a hardware structure of a terminal device of a method for predicting a state of health of a power battery according to an embodiment of the present disclosure;
FIG. 2 is a flow chart of a method for predicting the state of health of a power battery according to an embodiment of the present disclosure;
FIG. 3 is a model framework diagram of a method for predicting the state of health of a power battery according to an embodiment of the present disclosure;
fig. 4 is a schematic flow chart of a method for predicting a state of health of a power battery according to an embodiment of the present disclosure;
FIG. 5 is a preferred flow chart of a method for predicting the state of health of a power battery according to an embodiment of the present application;
fig. 6 is a block diagram of a power battery state of health prediction apparatus according to an embodiment of the present application.
In the figure: 102. a processor; 104. a memory; 106. a transmission device; 108. an input-output device; 10. an acquisition module; 20. a processing module; 30. and a prediction module.
Detailed Description
For a clearer understanding of the objects, aspects and advantages of the present application, reference is made to the following description and accompanying drawings.
Unless defined otherwise, technical or scientific terms referred to herein shall have the same general meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The use of the terms "a" and "an" and "the" and similar referents in the context of this application do not denote a limitation of quantity, either in the singular or the plural. The terms "comprises," "comprising," "has," "having," and any variations thereof, as referred to in this application, are intended to cover non-exclusive inclusions; for example, a process, method, and system, article, or apparatus that comprises a list of steps or modules (elements) is not limited to the listed steps or modules, but may include other steps or modules (elements) not listed or inherent to such process, method, article, or apparatus. Reference throughout this application to "connected," "coupled," and the like is not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. Reference to "a plurality" in this application means two or more. "and/or" describes an association relationship of associated objects, meaning that three relationships may exist, for example, "A and/or B" may mean: a exists alone, A and B exist simultaneously, and B exists alone. In general, the character "/" indicates a relationship in which the objects associated before and after are an "or". The terms "first," "second," "third," and the like in this application are used for distinguishing between similar items and not necessarily for describing a particular sequential or chronological order.
The method embodiments provided in the present embodiment may be executed in a terminal, a computer, or a similar computing device. For example, the method is executed on a terminal, and fig. 1 is a block diagram of a hardware structure of the terminal of the method for predicting the state of health of a power battery according to the embodiment. As shown in fig. 1, the terminal may include one or more processors 102 (only one shown in fig. 1) and a memory 104 for storing data, wherein the processor 102 may include, but is not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA. The terminal may also include a transmission device 106 for communication functions and an input-output device 108. It will be understood by those of ordinary skill in the art that the structure shown in fig. 1 is merely an illustration and is not intended to limit the structure of the terminal described above. For example, the terminal may also include more or fewer components than shown in FIG. 1, or have a different configuration than shown in FIG. 1.
The memory 104 may be used to store a computer program, for example, a software program and a module of an application software, such as a computer program corresponding to the power battery state of health prediction method in the present embodiment, and the processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, so as to implement the method described above. The memory 104 may include high speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory 104 can further include memory located remotely from the processor 102, which can be connected to the terminal over a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The transmission device 106 is used to receive or transmit data via a network. The network described above includes a wireless network provided by a communication provider of the terminal. In one example, the transmission device 106 includes a Network adapter (NIC) that can be connected to other Network devices via a base station to communicate with the internet. In one example, the transmission device 106 may be a Radio Frequency (RF) module, which is used to communicate with the internet in a wireless manner.
In the present embodiment, a method for predicting the state of health of a power battery is provided, and fig. 2 is a flowchart of the method for predicting the state of health of a power battery in the present embodiment, as shown in fig. 2, the flowchart includes the following steps:
and step S210, acquiring target multivariable data of the power battery.
Step S220 is performed to process the target multivariate data to obtain target gray-scale map data corresponding to the target multivariate data.
It should be noted that the target multivariate data in this embodiment includes characteristic data such as voltage data, current data and temperature data of the power battery, different target multivariate data can be dynamically selected as training data and corresponding input values of the prediction model according to actual conditions, and the width and length of the generated gray scale map are dynamically adjusted according to the selected target multivariate data scale in the conversion process of the target multivariate data.
And step S230, analyzing the target gray-scale image data based on the prediction model with complete training to obtain the predicted value of the health state of the power battery.
It should be noted that the prediction model adopted in this embodiment is a two-dimensional convolutional neural network model, the multivariate data in the training data set is converted to generate a corresponding gray-scale map data set, and the two-dimensional convolutional neural network model is trained based on the gray-scale map data set, so that the two-dimensional convolutional neural network model with complete training can be applied to predicting the State of Health (State of Health, abbreviated as SOH) of the power battery.
In the current prediction method, relevant data of the battery health state are collected from a lithium battery based on data collection equipment, a plurality of relevant data sets are generated according to the health state value corresponding to each batch of relevant data, different health state prediction models are established based on each relevant data set, and the health state of the battery is predicted. Therefore, the prediction method does not consider the strong coupling relationship among a plurality of variable data of the battery, and the strong coupling relationship can interfere the prediction process, so that the accuracy of the health state prediction result of the power battery is reduced. On the basis of the prior art, the target multivariable data are converted into corresponding target gray-scale map data, the target gray-scale map data are analyzed through the prediction model, so that the coupling relation among the multivariable data is fully considered in the prediction process, specifically, the target multivariable data of the power battery are obtained, the target multivariable data are processed, the target gray-scale map data corresponding to the target multivariable data are obtained, further, the target gray-scale map data are analyzed based on the prediction model which is completely trained, the health state prediction value of the power battery is obtained, the problem that the health state of the power battery cannot be accurately predicted due to the strong coupling relation among the multivariable data is solved, and the accuracy of the health state prediction result of the power battery is improved.
In some embodiments, the processing the target multivariate data to obtain the target gray-scale map data corresponding to the target multivariate data includes the following steps:
step S221, acquiring voltage data, current data and temperature data of the power battery from the target multivariable data;
step S222, mapping the voltage data, the current data and the temperature data of the power battery to a preset numerical range, and splicing the mapped voltage data, current data and temperature data to obtain target gray-scale map data corresponding to the target multivariable data.
Specifically, voltage data, current data and temperature data of the power battery are mapped to 0-255, and the mapped voltage data, current data and temperature data are spliced to generate target gray-scale map data corresponding to the target multivariable data.
It should be noted that the gray scale value range of the gray scale map is 0 to 255, where the gray scale value refers to the brightness of each pixel point, and each pixel point of the gray scale map only contains one sampling color, so in this embodiment, the voltage data, the current data, and the temperature data of the power battery are mapped to 0 to 255, so that each target multivariate data is converted into a single pixel point in the target gray scale map, and then the target gray scale map can be analyzed through gray scale value distribution, and feature information such as the coupling relationship between the target multivariate data can be further obtained based on the gray scale map.
According to the embodiment, the voltage data, the current data and the temperature data of the power battery are obtained from the target multivariable data, the voltage data, the current data and the temperature data of the power battery are mapped to the preset numerical range, the mapped voltage data, current data and temperature data are spliced, and the target gray-scale map data corresponding to the target multivariable data are obtained, so that the health state of the power battery can be analyzed based on the multivariable data in the gray-scale map form, and the accuracy of the health state prediction result is improved.
In some embodiments, before analyzing the target grayscale map data based on a well-trained prediction model to obtain a predicted value of the state of health of the power battery, the method further includes the following steps:
generating a confrontation network model based on a multivariable condition with complete training, performing data enhancement on first sample multivariable data, and generating a training data set corresponding to the first sample multivariable data;
and training the prediction model based on the training data set to obtain the prediction model with complete training.
Specifically, in a data cloud platform of the power battery, a market vehicle with the battery aging degree reaching an upper limit is screened out, and corresponding battery aging data is extracted as first sample multivariate data, wherein the first sample multivariate data comprises voltage data, current data, temperature data and SOH values in the historical charging process of the market vehicle.
Further, based on the specified SOH value, generating a corresponding battery aging data through a complete training multivariable condition generation confrontation network model (MV-CGAN model for short), so as to realize data expansion, meanwhile, when the aging data of the market vehicle is lost, generating the corresponding battery aging data through the complete training multivariable condition generation confrontation network model based on the specified SOH value, selecting data of a corresponding time period to perform data completion on the aging data of the market vehicle, and further forming a training data set based on the battery aging data after the data expansion and the data completion.
According to the embodiment, the confrontation network model is generated based on the multivariate condition with complete training, the first sample multivariate data is subjected to data enhancement, the training data set corresponding to the first sample multivariate data is generated, the prediction model is trained based on the training data set, and the prediction model with complete training is obtained, so that a large amount of high-quality training data can be obtained through data enhancement, and the prediction model is more reliable.
In some embodiments, before generating the confrontation network model based on the well-trained multivariate condition and processing the first sample multivariate data to obtain the training data set corresponding to the first sample multivariate data, the method further includes the following steps:
and acquiring second sample multivariable data from the battery test database, and training the multivariable condition generation countermeasure network model through the second sample multivariable data to obtain an initial multivariable condition generation countermeasure network model.
Specifically, cycle life test data of the power battery, including voltage data, current data, temperature data and corresponding SOH values in each test charging process, are obtained from a battery test database and are used as second sample multivariable data.
Further, training the multivariable condition generation confrontation network model based on the second sample multivariable data until the model reaches the optimization termination condition, and obtaining the initial multivariable condition generation confrontation network model.
According to the embodiment, the multivariate data of the second sample is obtained from the battery test database, the multivariate conditional generation confrontation network model is trained through the multivariate data of the second sample, the initial multivariate conditional generation confrontation network model is obtained, and therefore preliminary training of the multivariate conditional generation confrontation network model is achieved.
In some embodiments, training the prediction model based on the training data set to obtain a fully-trained prediction model, includes the following steps:
step S231, processing the first sample multivariable data after data enhancement in the training data set to generate a gray-scale map data set corresponding to the training data set;
step S232, training the prediction model based on the gray-scale image data set to obtain a prediction model with complete training.
Specifically, in the training data set, the data-enhanced first sample multivariate data comprises voltage data, current data and temperature data of the power battery, the voltage data, the current data and the temperature data of the power battery are mapped to 0-255, and the mapped voltage data, current data and temperature data are spliced to generate a corresponding gray map data set, for example, preset voltage data x 1 Current data x 2 And temperature data x 3 Is T, is x by the following formula 1 The jth voltage data element x in (b) 1,j And (3) conversion is carried out:
wherein x is 1,max Denotes x 1 The voltage data element with the largest median value, x 1,min Denotes x 1 The voltage data element having the smallest value in the above equation, and the right side of the above equation indicates that the voltage data element x is normalized by the normalization method 1,j Mapping to 0-1, multiplying normalization result with 255 to obtain mapped voltage data, and applying the above formula to voltage data x 1 Current data x 2 And temperature data x 3 And the converted data elements are spliced to generate a gray-scale image data set with the width of 3 and the length of T.
Further, training the prediction model based on the gray-scale image data set until the mean square error loss of the prediction model is smaller than a preset threshold value to obtain the prediction model with complete training, so that the prediction model fully combines the coupling relation among the multivariable data when analyzing the target multivariable data.
It should be noted that the preset threshold is 0.01, and the mean square error loss of the prediction model is defined as follows:
loss cnn ＝avg(||SOH pre -SOH real || 2 )；
therein, loss cnn For predicting mean square error loss, SOH, of the model pre As predicted values of state of health, and SOH real The true value of the health state.
According to the embodiment, the first sample multivariable data after data enhancement in the training data set is processed to generate the gray-scale map data set corresponding to the training data set, the prediction model is trained on the basis of the gray-scale map data set to obtain the well-trained prediction model, and therefore in the prediction process, the health state of the power battery is analyzed by combining the coupling relation among the voltage data, the current data and the temperature data, and the prediction accuracy is improved.
In some embodiments, the method further comprises the following steps after training the multivariate conditional generation countermeasure network model by using the second sample multivariate data to obtain an initial multivariate conditional generation countermeasure network model:
and based on the first sample multivariable data, carrying out transfer learning on the initial multivariable condition generation confrontation network model to obtain a completely-trained multivariable condition generation confrontation network model.
Specifically, the first sample multivariate data in this embodiment is battery aging data of market vehicles, and based on the first sample multivariate data, migration learning is performed on the initial multivariate condition generation countermeasure network model, that is, parameter fine tuning is performed on the initial multivariate condition generation countermeasure network model until the optimization training of the model is completed, where only training configuration items, such as learning rate, batch size, and the like, are adjusted.
It should be noted that the migration learning refers to using a model obtained by pre-training based on task a as an initial point of developing a model for task B, and further performing fine tuning on the model to adapt to task B.
According to the embodiment, based on the first sample multivariate data, the initial multivariate condition generation confrontation network model is subjected to transfer learning, and the well-trained multivariate condition generation confrontation network model is obtained, so that the reliability and the practicability of the generated data can be improved in the data enhancement process.
In some embodiments, the obtaining the second sample multivariate data from the battery test database, and before training the multivariate conditional generation countermeasure network model with the second sample multivariate data, further comprises the following steps:
and optimizing the condition generation countermeasure network model through the gated circulation unit network to generate a multivariable condition generation countermeasure network model.
It should be noted that, as shown in fig. 3, in the multivariable condition generation countermeasure network model constructed based on a Gated current Unit (GRU) network, the generator is composed of three groups of GRU networks, and the discriminator is composed of three groups of bidirectional GRU networks, and the cycle life test data refers to the voltage data, current data, temperature data and corresponding SOH value during each test charging process, and the data generated during each charging process is defined as one sample x = [ x ] value 1 ,x 2 ,x 3 ]Wherein x is 1 、x 2 And x 3 Respectively representing voltage data, current data and temperature data during charging, in the present embodiment, a multivariate condition generation countermeasure network model may be constructed based on an arbitrary deep learning model in addition to the GRU network.
Specifically, in the generator of the model, three sets of noise data z are randomly generated 1 、z 2 And z 3 Inputting each group of noise and corresponding SOH value into corresponding GRU network, then three groups of networks respectively generate voltage data G (z) 1 ) Current data G (z) 2 ) And temperature data G (z) 3 ) In the model arbiter, the bidirectional GRU network 1 is taken as an example, and the real voltage data x is used 1 And the generated voltage data G (z) 1 ) And inputting the corresponding GRU networks with the corresponding SOH values respectively to obtain the probability that the input data is real, so as to judge whether the input data is real or generated.
Further, in the training process of the model, for the discriminator of the model, the label corresponding to the real data is set to be 1, that is, the data is true, and the label corresponding to the generated data is set to be 0, that is, the data is false, so as to obtain the sum of three groups of network losses, and the GRU network parameters in the discriminator are updated through an Adaptive Moment Estimation (Adam) optimizer, and at this time, the generator parameters are not updated, and any real sample voltage data x is used 1 For example, the calculation process in a bidirectional GRU network is as follows:
wherein x is 1, Is x 1 The jth voltage data element in (b),
and &>
Respectively representing a forward GRU network element and a reverse GRU network element in a bidirectional GRU network, with the difference that the reverse GRU network element is a backward-forward input of data elements, z j Represents the jth element of the output vector,then inputting the output vector of the bidirectional GRU network into a full connection layer to obtain the probability that the sample voltage data is true; for a generator of the model, the generated data is input into a discriminator, a label corresponding to the generated data is set to be 1, namely the data is true, further, based on an output result of the discriminator and the corresponding label, a corresponding loss value is obtained through calculation, parameters of three groups of GRU networks in the generator are updated through an Adam optimizer, and at the moment, the parameters of the discriminator are not updated; and repeatedly and alternately executing the training process of the discriminator and the generator until the loss values corresponding to the discriminator and the generator reach an equilibrium state, so that the authenticity of the generated data is enhanced.
According to the embodiment, the multivariable condition generation confrontation network model is constructed on the basis of the gated cycle unit network, so that the confrontation network model can be generated through the multivariable condition, data enhancement is performed on sample multivariable data, and a complete training data set is obtained.
Fig. 4 is a schematic flow chart of the method for predicting the state of health of the power battery according to the embodiment, and as shown in fig. 4, the specific process of the method for predicting the state of health of the power battery is as follows:
acquiring cycle life test data from a battery test database S410, training a multivariable condition generation countermeasure network model through the cycle life test data to obtain an initial multivariable condition generation countermeasure network model S420, acquiring battery aging data of market vehicles from a data cloud platform of the power battery S430, and performing transfer learning on the initial multivariable condition generation countermeasure network model based on the battery aging data of the market vehicles, namely performing parameter fine tuning on the initial multivariable condition generation countermeasure network model S440; generating an antagonistic network model based on a multivariable condition with complete training, performing data enhancement on battery aging data of market vehicles, and generating a corresponding training data set S450; and processing the battery aging data after data enhancement in the training data set to generate a corresponding gray-scale image data set S460, training the convolutional neural network model based on the gray-scale image data set to obtain a completely trained convolutional neural network model S470, and predicting the health state of the power battery through the completely trained convolutional neural network model to obtain the SOH value S480 of the power battery.
The present embodiment is described and illustrated below by means of preferred embodiments.
Fig. 5 is a preferred flowchart of the method for predicting the state of health of the power battery according to the embodiment, and as shown in fig. 5, the method for predicting the state of health of the power battery includes the following steps:
step S510, optimizing the condition generation confrontation network model through the gate control circulation unit network, and generating a multivariable condition generation confrontation network model;
step S520, obtaining second sample multivariable data from the battery test database, training the multivariable condition generation countermeasure network model through the second sample multivariable data, and obtaining an initial multivariable condition generation countermeasure network model;
step S530, based on the first sample multivariable data, carrying out transfer learning on the initial multivariable condition generation countermeasure network model to obtain a multivariable condition generation countermeasure network model with complete training;
step S540, generating a confrontation network model based on a multivariable condition with complete training, performing data enhancement on first sample multivariable data, and generating a training data set corresponding to the first sample multivariable data;
step S550, processing the first sample multivariable data after data enhancement in the training data set to generate a gray map data set corresponding to the training data set;
step S560, training the prediction model based on the gray-scale image data set to obtain a prediction model with complete training;
step S570, acquiring target multivariable data of the power battery;
step S580, processing the target multivariable data to obtain target gray-scale image data corresponding to the target multivariable data;
and step S590, analyzing the target gray-scale image data based on the prediction model with complete training to obtain the health state prediction value of the power battery.
According to the embodiment, the condition generation countermeasure network model is optimized through the gate control cycle unit network, the multivariable condition generation countermeasure network model is generated, the multivariable condition generation countermeasure network model is optimized and trained on the basis of first sample multivariable data and second sample multivariable data, the multivariable condition generation countermeasure network model with complete training is obtained, the countermeasure network model with complete training is generated on the basis of the multivariable condition with complete training, data enhancement is performed on the first sample multivariable data, a training data set corresponding to the first sample multivariable data is generated, the first sample multivariable data with enhanced data in the training data set is processed, a gray map data set corresponding to the training data set is generated, the prediction model is trained on the basis of the gray map data set, a prediction model with complete training is obtained, the target gray map data corresponding to the target multivariable data is obtained, further, the target gray map data is analyzed on the basis of the prediction model with complete training, the health state prediction value of the power battery is obtained, the problem that the health state of the power battery cannot be accurately predicted due to the strong coupling relationship between the multivariable data is solved, and the accurate prediction result of the health state of the power battery is improved is achieved.
It should be noted that the steps illustrated in the above-described flow diagrams or in the flow diagrams of the figures may be performed in a computer system, such as a set of computer-executable instructions, and that, although a logical order is illustrated in the flow diagrams, in some cases, the steps illustrated or described may be performed in an order different than presented herein.
In this embodiment, a power battery health status prediction apparatus is further provided, and the apparatus is used to implement the foregoing embodiments and preferred embodiments, and the description of which has been already made is omitted. The terms "module," "unit," "sub-unit," and the like as used below may implement a combination of software and/or hardware of predetermined functions. Although the means described in the embodiments below are preferably implemented in software, an implementation in hardware, or a combination of software and hardware is also possible and contemplated.
Fig. 6 is a block diagram of the structure of the power battery state of health prediction device of the present embodiment, and as shown in fig. 6, the device includes: the system comprises an acquisition module 10, a processing module 20 and a prediction module 30;
the acquisition module 10 is used for acquiring target multivariable data of the power battery;
the processing module 20 is used for processing the target multivariable data to obtain target gray-scale image data corresponding to the target multivariable data;
and the prediction module 30 analyzes the target gray-scale image data based on the well-trained prediction model to obtain the health state prediction value of the power battery.
By the device provided by the embodiment, the target multivariable data of the power battery are obtained, the target multivariable data are processed, the target gray-scale map data corresponding to the target multivariable data are obtained, further, the target gray-scale map data are analyzed based on a prediction model with complete training, and the health state prediction value of the power battery is obtained.
In some embodiments, on the basis of fig. 6, the apparatus further comprises a conversion module for obtaining voltage data, current data and temperature data of the power battery from the target multivariate data; and mapping the voltage data, the current data and the temperature data of the power battery to a preset numerical range, and splicing the mapped voltage data, current data and temperature data to obtain target gray-scale map data corresponding to the target multivariable data.
In some embodiments, on the basis of fig. 6, the apparatus further includes a first training module, configured to generate a confrontation network model based on a well-trained multivariate condition, perform data enhancement on the first sample multivariate data, and generate a training data set corresponding to the first sample multivariate data; and training the prediction model based on the training data set to obtain the prediction model with complete training.
In some embodiments, based on fig. 6, the apparatus further includes a second training module, configured to obtain second sample multivariate data from the battery test database, train the multivariate conditional generation countermeasure network model through the second sample multivariate data, and obtain an initial multivariate conditional generation countermeasure network model.
In some embodiments, based on fig. 6, the apparatus further includes a third training module, configured to process the first sample multivariate data after data enhancement in the training data set, and generate a gray-scale map data set corresponding to the training data set; and training the prediction model based on the gray-scale image data set to obtain a prediction model with complete training.
In some embodiments, on the basis of fig. 6, the apparatus further includes an optimization module, configured to perform migration learning on the initial multivariate condition-generated confrontation network model based on the first sample multivariate data, so as to obtain a well-trained multivariate condition-generated confrontation network model.
In some embodiments, the apparatus further comprises a generation module configured to optimize the conditional generation confrontation network model by gating the cycle unit network to generate a multivariate conditional generation confrontation network model, based on fig. 6.
The above modules may be functional modules or program modules, and may be implemented by software or hardware. For a module implemented by hardware, the modules may be located in the same processor; or the modules can be respectively positioned in different processors in any combination.
There is also provided in this embodiment a computer device comprising a memory having a computer program stored therein and a processor arranged to run the computer program to perform the steps of any of the method embodiments described above.
Optionally, the computer device may further include a transmission device and an input/output device, wherein the transmission device is connected to the processor, and the input/output device is connected to the processor.
It should be noted that, for specific examples in this embodiment, reference may be made to the examples described in the foregoing embodiments and optional implementations, and details are not described again in this embodiment.
In addition, in combination with the method for predicting the state of health of the power battery provided in the above embodiment, a storage medium may also be provided to implement in this embodiment. The storage medium having stored thereon a computer program; the computer program, when executed by a processor, implements any one of the power battery state of health prediction methods in the above embodiments.
It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to be limiting. All other embodiments, which can be derived by a person skilled in the art from the examples provided herein without any inventive step, shall fall within the scope of protection of the present application.
It is obvious that the drawings are only examples or embodiments of the present application, and it is obvious to those skilled in the art that the present application can be applied to other similar cases according to the drawings without creative efforts. Moreover, it should be appreciated that in the development of any such actual implementation, as in any engineering or design project, numerous implementation-specific decisions must be made to achieve the developers' specific goals, such as compliance with system-related and business-related constraints, which may vary from one implementation to another.
The term "embodiment" is used herein to mean that a particular feature, structure, or characteristic described in connection with the embodiment can be included in at least one embodiment of the present application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. It is to be expressly or implicitly understood by one of ordinary skill in the art that the embodiments described in this application may be combined with other embodiments without conflict.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the patent protection. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, and these are all within the scope of protection of the present application. Therefore, the protection scope of the present application shall be subject to the appended claims.
Claims (10)
1. A method for predicting the state of health of a power battery is characterized by comprising the following steps:
acquiring target multivariable data of the power battery;
processing the target multivariable data to obtain target gray-scale image data corresponding to the target multivariable data;
and analyzing the target gray-scale image data based on a prediction model with complete training to obtain a health state prediction value of the power battery.
2. The method for predicting the state of health of a power battery according to claim 1, wherein the step of processing the target multivariate data to obtain target gray-scale map data corresponding to the target multivariate data comprises:
acquiring voltage data, current data and temperature data of the power battery from the target multivariable data;
and mapping the voltage data, the current data and the temperature data of the power battery to a preset numerical range, and splicing the mapped voltage data, current data and temperature data to obtain target gray-scale map data corresponding to the target multivariable data.
3. The method for predicting the state of health of a power battery according to claim 1, wherein before analyzing the target grayscale map data based on the well-trained prediction model to obtain the predicted value of the state of health of the power battery, the method further comprises:
generating an confrontation network model based on a multivariate condition with complete training, performing data enhancement on first sample multivariate data, and generating a training data set corresponding to the first sample multivariate data;
and training the prediction model based on the training data set to obtain the prediction model with complete training.
4. The method for predicting the state of health of a power battery according to claim 3, wherein the generating a confrontation network model based on the multivariate condition with complete training, processing the first sample multivariate data, and before obtaining the training data set corresponding to the first sample multivariate data, further comprises:
and acquiring second sample multivariable data from the battery test database, and training the multivariable condition generation countermeasure network model through the second sample multivariable data to obtain an initial multivariable condition generation countermeasure network model.
5. The method for predicting the state of health of a power battery according to claim 3, wherein the training the prediction model based on the training data set to obtain the well-trained prediction model comprises:
processing the first sample multivariable data after data enhancement in the training data set to generate a gray-scale map data set corresponding to the training data set;
and training the prediction model based on the gray-scale image data set to obtain the prediction model with complete training.
6. The method for predicting the state of health of a power battery according to claim 4, wherein the training of the multivariate conditionally-generated confrontational network model through the second sample multivariate data to obtain the initial multivariate conditionally-generated confrontational network model further comprises:
and carrying out transfer learning on the initial multivariable condition generation confrontation network model based on the first sample multivariable data to obtain the multivariable condition generation confrontation network model with complete training.
7. The method according to claim 4, wherein the step of obtaining second sample multivariate data from the battery test database, and before training the multivariate condition generation countermeasure network model using the second sample multivariate data, further comprises:
and optimizing the condition generation countermeasure network model through the gated circulation unit network, and generating the multivariate condition generation countermeasure network model.
8. A power battery state of health prediction apparatus, the apparatus comprising:
the acquisition module is used for acquiring target multivariable data of the power battery;
the processing module is used for processing the target multivariable data to obtain target gray-scale image data corresponding to the target multivariable data;
and the prediction module analyzes the target gray-scale image data based on a prediction model with complete training to obtain a health state prediction value of the power battery.
9. A computer arrangement comprising a memory and a processor, characterized in that the memory has stored therein a computer program, the processor being arranged to run the computer program to perform the steps of the power cell state of health prediction method according to any one of claims 1 to 7.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the power cell state of health prediction method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211500338.4A CN115980586A (en) | 2022-11-28 | 2022-11-28 | Power battery health state prediction method and device and computer equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211500338.4A CN115980586A (en) | 2022-11-28 | 2022-11-28 | Power battery health state prediction method and device and computer equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115980586A true CN115980586A (en) | 2023-04-18 |
Family
ID=85960158
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211500338.4A Pending CN115980586A (en) | 2022-11-28 | 2022-11-28 | Power battery health state prediction method and device and computer equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115980586A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116613865A (en) * | 2023-07-18 | 2023-08-18 | 广东电网有限责任公司东莞供电局 | Battery quick-charging method, battery energy storage system and energy storage power station |
-
2022
- 2022-11-28 CN CN202211500338.4A patent/CN115980586A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116613865A (en) * | 2023-07-18 | 2023-08-18 | 广东电网有限责任公司东莞供电局 | Battery quick-charging method, battery energy storage system and energy storage power station |
| CN116613865B (en) * | 2023-07-18 | 2024-03-08 | 广东电网有限责任公司东莞供电局 | Battery quick-charging method, battery energy storage system and energy storage power station |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112181666A (en) | Method, system, equipment and readable storage medium for equipment evaluation and federal learning importance aggregation based on edge intelligence | |
| CN109753751B (en) | MEC random task migration method based on machine learning | |
| CN112364981B (en) | Differentiable searching method and device for mixed precision neural network | |
| US20220217792A1 (en) | Industrial 5g dynamic multi-priority multi-access method based on deep reinforcement learning | |
| CN110309874B (en) | Negative sample screening model training method, data screening method and data matching method | |
| CN110926782A (en) | Circuit breaker fault type judgment method and device, electronic equipment and storage medium | |
| Yang et al. | Deep reinforcement learning based wireless network optimization: A comparative study | |
| CN115980586A (en) | Power battery health state prediction method and device and computer equipment | |
| CN114154646A (en) | Efficiency optimization method for federal learning in mobile edge network | |
| CN109375999A (en) | A kind of MEC Random Task moving method based on Bayesian network | |
| CN116363452A (en) | Task model training method and device | |
| CN115759434A (en) | Method, system, computer device and storage medium for predicting battery capacity in real time | |
| WO2022144542A1 (en) | Method for predicting condition parameter degradation of a cell | |
| CN117289167A (en) | Battery remaining life prediction method, device and medium based on multiple neural network | |
| CN116796821A (en) | Efficient neural network architecture searching method and device for 3D target detection algorithm | |
| CN116609677A (en) | Battery state estimation method | |
| CN115001937B (en) | Smart city Internet of things-oriented fault prediction method and device | |
| Wu et al. | Least-squares support vector machine-based learning and decision making in cognitive radios | |
| CN112765892B (en) | Intelligent switching judgment method in heterogeneous Internet of vehicles | |
| CN115587616A (en) | Network model training method and device, storage medium and computer equipment | |
| US11797372B2 (en) | Method and apparatus for generating time series data based on multi-condition constraints, and medium | |
| CN114238658A (en) | Link prediction method and device of time sequence knowledge graph and electronic equipment | |
| Wang et al. | Adaptive Compute Offloading Algorithm for Metasystem Based on Deep Reinforcement Learning | |
| CN113298255A (en) | Deep reinforcement learning robust training method and device based on neuron coverage rate | |
| Kousias et al. | Long short term memory networks for bandwidth forecasting in mobile broadband networks under mobility |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |