CN117907872A - Battery life prediction method and device, electronic equipment and storage medium - Google Patents
Battery life prediction method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN117907872A CN117907872A CN202410091098.XA CN202410091098A CN117907872A CN 117907872 A CN117907872 A CN 117907872A CN 202410091098 A CN202410091098 A CN 202410091098A CN 117907872 A CN117907872 A CN 117907872A
- Authority
- CN
- China
- Prior art keywords
- battery
- model
- initial
- battery life
- determining
- 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
- 238000000034 method Methods 0.000 title claims abstract description 77
- 238000012549 training Methods 0.000 claims abstract description 156
- 238000013526 transfer learning Methods 0.000 claims abstract description 25
- 238000005457 optimization Methods 0.000 claims description 31
- 238000004422 calculation algorithm Methods 0.000 claims description 23
- 238000012360 testing method Methods 0.000 claims description 17
- 238000004590 computer program Methods 0.000 claims description 16
- 230000015654 memory Effects 0.000 claims description 12
- 230000005012 migration Effects 0.000 claims description 12
- 238000013508 migration Methods 0.000 claims description 12
- 238000007781 pre-processing Methods 0.000 claims description 7
- WHXSMMKQMYFTQS-UHFFFAOYSA-N Lithium Chemical compound [Li] WHXSMMKQMYFTQS-UHFFFAOYSA-N 0.000 abstract description 5
- 229910052744 lithium Inorganic materials 0.000 abstract description 5
- 230000004927 fusion Effects 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 19
- 241000283153 Cetacea Species 0.000 description 9
- 238000004891 communication Methods 0.000 description 8
- 230000008901 benefit Effects 0.000 description 7
- 238000012545 processing Methods 0.000 description 7
- 230000032683 aging Effects 0.000 description 6
- 230000015556 catabolic process Effects 0.000 description 6
- 238000006731 degradation reaction Methods 0.000 description 6
- 230000006870 function Effects 0.000 description 6
- 239000000126 substance Substances 0.000 description 6
- HBBGRARXTFLTSG-UHFFFAOYSA-N Lithium ion Chemical compound [Li+] HBBGRARXTFLTSG-UHFFFAOYSA-N 0.000 description 4
- 230000008859 change Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 229910001416 lithium ion Inorganic materials 0.000 description 4
- 230000005540 biological transmission Effects 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 230000003287 optical effect Effects 0.000 description 3
- 238000013473 artificial intelligence Methods 0.000 description 2
- 230000003993 interaction Effects 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 238000010801 machine learning Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 210000002569 neuron Anatomy 0.000 description 2
- 238000010606 normalization Methods 0.000 description 2
- 230000027756 respiratory electron transport chain Effects 0.000 description 2
- 230000006403 short-term memory Effects 0.000 description 2
- 230000007704 transition Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000009825 accumulation Methods 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000007599 discharging Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 239000011521 glass Substances 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 238000010438 heat treatment Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 230000007786 learning performance Effects 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Landscapes
- Secondary Cells (AREA)
Abstract
The invention discloses a battery life prediction method, a device, electronic equipment and a storage medium, and relates to the technical field of battery management, wherein the method comprises the following steps: determining an initial training model; the initial training model is a training model determined through transfer learning; acquiring target battery parameters of a vehicle battery; wherein the target battery parameters include battery voltage, battery current, battery temperature, and cycle number; determining an initial battery life according to the battery voltage, the battery current, the battery temperature, the cycle number and the initial training model; and predicting the battery life according to the initial battery life and the battery life prediction model, and determining the target battery life. According to the technical scheme, based on the fusion of the initial training model, the data driving of the transfer learning and the battery life prediction model, the accurate estimation of the residual service life of the lithium battery under the condition of lacking of the training data set is realized, the accuracy of life estimation is improved, and the user experience is improved.
Description
Technical Field
The present invention relates to the field of battery management technologies, and in particular, to a method and apparatus for predicting battery life, an electronic device, and a storage medium.
Background
The lithium ion battery has the advantages of high energy density, high power density, long service life, quick charge and the like, and the advantages make the lithium ion battery one of the most ideal power batteries for new energy automobiles. In practical application, predicting the residual life of the lithium ion battery is of great significance in ensuring the normal operation of equipment and prolonging the service life of the battery. Therefore, it is very important to study how to accurately predict the remaining life of a battery.
At present, in the prior art, a data-driven prediction method and a filtering method are mainly used for a residual service life prediction method of a battery, the chemical reaction mechanism and the degradation mechanism inside a lithium ion battery are not considered in the data-driven prediction method, and the time-varying rule in the measurable state data, the environment state and the load state information of the battery is found out by directly mining the battery health state information implicit in the measurable state data, the environment state and the load data of the battery, so that the residual service life prediction is realized. The filtering method adopts probability prediction, so that noise in the data can be eliminated.
However, the existing data-driven prediction method mainly includes: random process models and artificial intelligence methods. Although the random process method has wide application range and can express the uncertainty of the prediction result, the random process method is very dependent on the initialization of super parameters, and has large calculated amount and low long-term prediction precision. Noise in the data can be eliminated by the filtering method, but the expression uncertainty is complex, the initialization process is relatively poor in timeliness, and the use experience of a user is reduced.
Disclosure of Invention
The invention provides a battery life prediction method, a device, electronic equipment and a storage medium, which are based on the fusion of an initial training model, a data driving of transfer learning and a battery life prediction model, and can accurately estimate the residual service life of a lithium battery under the condition of lacking a training data set, so that the accuracy of life estimation is improved, and the use experience of a user is improved.
According to an aspect of the present invention, there is provided a battery life prediction method including:
determining an initial training model; the initial training model is a training model determined through transfer learning;
Acquiring target battery parameters of a vehicle battery; wherein the target battery parameters include battery voltage, battery current, battery temperature, and cycle number;
Determining an initial battery life according to the battery voltage, the battery current, the battery temperature, the cycle number and the initial training model;
Predicting the battery life according to the initial battery life and the battery life prediction model, and determining the target battery life; the battery life prediction model is a prediction model for optimizing and updating the initial battery life.
Optionally, determining the initial training model includes: acquiring source domain data of a battery; the source domain data is a data set for pre-training the model; and performing model training and migration learning according to the source domain data and the basic model to determine an initial training model.
Optionally, determining the initial battery life based on the battery voltage, the battery current, the battery temperature, the number of cycles, and the initial training model includes: determining training data and test data according to the battery voltage, the battery current, the battery temperature and the cycle number; training according to the training data and the initial training model to determine a target training model; and determining the initial battery life according to the test data and the target training model.
Optionally, predicting the battery life according to the initial battery life and the battery life prediction model, and determining the target battery life includes: predicting the battery life according to the initial battery life and the battery life prediction model, and determining the predicted battery life; and determining the target battery life according to the predicted battery life and the failure threshold.
Optionally, determining the target battery life from the predicted battery life and the failure threshold includes: determining whether the predicted battery life is greater than or equal to a failure threshold; if the predicted battery life is determined to be greater than or equal to the failure threshold, performing battery replacement; if the predicted battery life is determined to be less than the failure threshold, the target battery parameters are updated and the step of determining an initial battery life based on the battery voltage, the battery current, the battery temperature, the number of cycles, and the initial training model is performed back.
Optionally, acquiring the target battery parameter of the vehicle battery includes: acquiring initial battery parameters of a vehicle battery; and determining the target battery parameters according to the initial battery parameters and the preprocessing algorithm.
Optionally, performing model training and migration learning according to the source domain data and the base model to determine an initial training model includes: after normalizing the source domain data, determining model parameters of a basic model; the model parameters comprise the number of hidden layer nodes, learning rate and iteration times; determining optimization parameters according to the number of hidden layer nodes, the learning rate, the iteration times and an optimization algorithm; and determining an initial training model according to the optimization parameters, the basic model and the transfer learning.
According to another aspect of the present invention, there is also provided a battery life prediction apparatus including:
The training model determining module is used for determining an initial training model; the initial training model is a training model determined through transfer learning;
The battery parameter acquisition module is used for acquiring target battery parameters of the vehicle battery; wherein the target battery parameters include battery voltage, battery current, battery temperature, and cycle number;
The initial life determining module is used for determining the initial life of the battery according to the battery voltage, the battery current, the battery temperature, the cycle number and the initial training model;
The target life determining module is used for predicting the life of the battery according to the initial life of the battery and the battery life prediction model, and determining the life of the target battery; the battery life prediction model is a prediction model for optimizing and updating the initial battery life.
According to another aspect of the present invention, there is also provided an electronic apparatus including: at least one processor; and a memory communicatively coupled to the at least one processor; wherein the memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the method of predicting battery life of any one of the embodiments of the present invention.
According to another aspect of the present invention, there is also provided a computer readable storage medium storing computer instructions for causing a processor to execute the method for predicting battery life according to any embodiment of the present invention.
According to the technical scheme, an initial training model is determined; the initial training model is a training model determined through transfer learning; acquiring target battery parameters of a vehicle battery; wherein the target battery parameters include battery voltage, battery current, battery temperature, and cycle number; determining an initial battery life according to the battery voltage, the battery current, the battery temperature, the cycle number and the initial training model; predicting the battery life according to the initial battery life and the battery life prediction model, and determining the target battery life; the battery life prediction model is a prediction model for optimizing and updating the initial battery life. The invention aims to solve the problem that the residual service life of the power battery is difficult to accurately predict due to the fact that the real vehicle data of the battery is less and the working environment is complex, and the data driving and model method based on transfer learning are combined, so that the residual service life of the lithium battery can be accurately estimated even in the absence of a training data set, the accuracy of battery service life estimation is improved, and the user experience is improved.
It should be understood that the description in this section is not intended to identify key or critical features of the embodiments of the invention or to delineate the scope of the invention. Other features of the present invention will become apparent from the description that follows.
Drawings
In order to more clearly illustrate the technical solutions of the present invention, the drawings that are needed in the description of the embodiments will be briefly described below, it being obvious that the drawings in the following description are only some embodiments of the present invention, and that other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flowchart of a battery life prediction method provided in the first embodiment;
FIG. 2 is a flowchart of a battery life prediction method provided in the second embodiment;
fig. 3 is a schematic structural view of a battery life prediction apparatus provided in the third embodiment;
fig. 4 is a schematic structural diagram of an electronic device provided in the fourth embodiment.
Detailed Description
In order that those skilled in the art will better understand the present invention, a more complete description of the same will be rendered by reference to the appended drawings, wherein it is to be understood that the illustrated embodiments are merely exemplary of some, but not all embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the present invention without making any inventive effort, shall fall within the scope of the present invention.
It should be noted that the terms "first," "second," and the like in the description and the claims of the present invention and the above figures are used for distinguishing between similar objects and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used may be interchanged where appropriate such that the embodiments of the invention described herein may be implemented in sequences other than those illustrated or otherwise described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
Example 1
Fig. 1 is a flowchart of a battery life prediction method provided in a first embodiment, where the present embodiment is applicable to a case of predicting remaining life of a power battery, and the method may be performed by a battery life prediction device, where the battery life prediction device may be implemented in hardware and/or software, and in a specific embodiment, the battery life prediction device may be configured in an electronic device. As shown in fig. 1, the method of this embodiment specifically includes the following steps:
S101, determining an initial training model.
The initial training model is a training model determined by a data driving method of transfer learning, namely, a model which can be quickly adapted to different data sets and has stronger prediction capability through self-learning of a large amount of data.
Specifically, existing battery capacity aging data is obtained, the battery capacity aging data is an existing battery aging data set and is generally defined as source domain data, source domain data D S＝{xi,yi},xi is an input set, y i is an output set, feature knowledge between the source domain data and the service life of a battery is obtained by inputting the input set and the output set of the source domain data into a basic model for training, and when the feature knowledge meets a preset feature threshold, the trained basic model is determined to be an initial training model. The preset feature threshold is a preset threshold for determining whether the initial training model is trained, the basic model is usually WOA-LSTM (Weighted Optimization Algorithm Long Short-term memory), which is an optimization algorithm based on the LSTM model, and the basic idea is to improve the prediction accuracy of the model through weighted optimization, for example, a whale optimization algorithm. Specifically, the WOA-LSTM assigns different weight values to the weight parameters in the LSTM model, and then adjusts the weight values through an optimization algorithm to achieve a better prediction effect.
S102, acquiring target battery parameters of a vehicle battery.
The target battery parameters of the vehicle battery are aging data sets of the vehicle battery of the real vehicle in the actual working process of the vehicle, and the target battery parameters comprise battery voltage, battery current, battery temperature and cycle times. The battery voltage refers to the operating voltage of the vehicle battery, the battery current refers to the operating current of the vehicle battery, the battery temperature refers to the phenomenon that the surface of the battery generates heat due to chemical, electrochemical change, electron transfer, substance transmission and other reasons of the internal structure of the vehicle battery in the use process, and the heat generated by the vehicle battery can be accumulated in the vehicle battery if the heat cannot be completely dissipated into the environment. The cycle number refers to the charge and discharge number of the vehicle battery, i.e., the process of completing 100% of the complete discharge/charge of the vehicle battery.
Specifically, target battery parameters of a vehicle battery of a real vehicle, such as battery voltage, battery current, battery temperature and cycle times, are obtained in the actual working process of the vehicle, and the working voltage of the vehicle battery, the working current of the vehicle battery, and the phenomena of battery surface heating generated by chemical, electrochemical change, electron migration, substance transmission and other reasons of an internal structure of the vehicle battery in the using process are mastered in real time, and if the heat generated by the vehicle battery cannot be completely dissipated into the environment, the accumulation of the heat in the vehicle battery and the charge and discharge times of the vehicle battery are caused.
The device has the advantages that the battery voltage, the battery current, the battery temperature and the cycle times of the vehicle battery are mastered in real time, and the safety of the vehicle battery and the safety of a real vehicle are improved.
S103, determining the initial battery life according to the battery voltage, the battery current, the battery temperature, the cycle number and the initial training model.
The initial battery life is obtained by predicting the life of the vehicle battery through an initial training model.
Specifically, a training data set and a test data set are determined through the battery voltage, the battery current, the battery temperature and the cycle times of the vehicle battery, then the training data set is input into an initial training model as an input set for optimization training, and finally the test data set is input into the initial training model after optimization training for life prediction to determine the initial battery life.
S104, predicting the battery life according to the initial battery life and the battery life prediction model, and determining the target battery life.
The battery life prediction model is a prediction model for optimizing and updating the initial battery life. The battery life prediction model may be, for example, a double-exponential capacity degradation model, a kalman filter, or the like, and this embodiment is not limited thereto.
Specifically, after the initial battery life is determined, the initial battery life is input into a battery life prediction model, the initial battery life is optimized and updated through the battery life prediction model, and further the battery life is estimated, so that the target battery life is determined.
In one embodiment, the training or weighting of the initial training model may be specifically to find the optimal parameters of the LSTM model, including sample number, neuron number, and loss rate, using a whale optimization algorithm; then, using the existing discharging voltage, current, temperature and cycle number data set of the vehicle battery as a source domain to train an LSTM model with optimal parameters, and obtaining a good pre-train model based on RUL prediction of WOA-LSTM; further acquiring a small amount of data sets of the real vehicle batteries as a target domain, and then utilizing the data sets of the real vehicles to carry out fine adjustment and optimization on parameters of the pre-training model, so that the model can adapt to noise distribution of engineering machinery battery data, thereby realizing transfer learning; further inputting battery data of the real vehicle into a trained prediction model to obtain the residual service life of the power battery of the real vehicle; and finally, optimizing the residual service life of the real vehicle power battery obtained in the prediction model by using an unscented Kalman filter, predicting the predicted battery life of the battery, adding the optimized predicted battery life as an online sample into a training set, retraining the WOA-LSTM model, and carrying out the next iterative prediction to determine the final target battery life.
The method has the advantages that the output of the prediction model of each period is utilized to update the training data set, and the model is retrained to conduct iterative prediction, so that the prediction accuracy and the long-term prediction capability of the model are improved.
In one embodiment, using a whale optimization algorithm to find optimal parameters for the LSTM model specifically includes: carrying out normalization processing on target battery parameters, then constructing an LSTM neural network, determining parameters needing optimizing in an LSTM model and setting corresponding ranges, wherein the optimizing objects are as follows: the number Y of hidden layer nodes, the learning rate b and the iteration number k; initializing parameters of a whale algorithm, setting random population positions X of whales, and taking 1 for iteration times, wherein the specific expression is as follows: x (I, j) = (m (I) -I (I)) ×rand (I, j) +i (I), wherein m is the upper variable limit, I is the lower variable limit, x (I, 1), x (I, 2), x (I, 3) represent key parameters of the LSTM model, namely the number Y of hidden layer nodes, the learning rate b and the iteration number k, and are substituted into the original LSTM model; further, a root mean square error (Root Mean Squared Error, RMSE) is selected as an fitness value of the LSTM model, and the RMSE is an index for measuring the prediction accuracy of the prediction model on the continuous data, and measures the root mean square difference between the prediction value and the true value, which represents the average deviation degree between the prediction value and the true value, and is one of commonly used performance evaluation indexes in regression tasks. The concrete calculation formula isWherein/>Representing the actual value,/>Representing the predicted value. Then determining a global optimal position and a local optimal position according to the fitness value of whales; updating the whale position according to a whale position updating formula, and increasing one iteration number; repeating the steps, optimizing by using whale algorithm to obtain the optimal hidden layer node number Y, the learning rate b and the iteration times k, substituting the optimized parameters into the original LSTM model to obtain an LSTM network model with optimal parameters, wherein the LSTM network model with optimal parameters is the initial training model.
The setting has the advantages that under the dynamic load of the engineering machinery, the current, the voltage and the temperature change quickly, and the acquired data have the problems of noise, abnormal value, incomplete data and the like. The WOA is utilized to optimize the number Y of neurons of the LSTM hidden layer, the learning rate b and the iteration number k, namely 3 super parameters, so that the prediction performance under dynamic load is more excellent, and the prediction accuracy is improved.
In one embodiment, the determining of the battery life prediction model specifically includes: taking the double-index capacity degradation model as an observation equation of a battery life prediction model, taking an initial battery life obtained based on a transfer learning WOA-LSTM model as an observation value, and then determining the battery life prediction model based on the observation equation and a state transition equation. State transition equations in the battery life prediction model describe the system state vector dynamics process: The observation equation, a dual exponential capacity degradation model of battery life degradation, is used to describe the relationship between the observation vector and the future state: c k＝ak·exp(bk·k)+ck·exp(dk.k) +v (k), wherein k is the number of cycles of battery charge and discharge; c k is the capacity of the kth charge-discharge cycle; a. b, c and d are parameters to be solved, and fitting can be carried out by using a tool box to obtain the parameters to be solved of the double-index capacity degradation model; w (k) is the state noise of the system; v (k) is the observation noise of the system.
According to the technical scheme, an initial training model is determined; the initial training model is a training model determined through transfer learning; acquiring target battery parameters of a vehicle battery; wherein the target battery parameters include battery voltage, battery current, battery temperature, and cycle number; determining an initial battery life according to the battery voltage, the battery current, the battery temperature, the cycle number and the initial training model; predicting the battery life according to the initial battery life and the battery life prediction model, and determining the target battery life; the battery life prediction model is a prediction model for optimizing and updating the initial battery life. On the basis of the embodiment, the method aims at the problems that the real vehicle data of the vehicle battery of the pure electric engineering machinery is less and the prediction model is difficult to sufficiently train, the existing battery data set can be used for training the initial training model by utilizing transfer learning, and then the parameters of the initial training model are finely tuned and optimized according to the vehicle battery data so as to adapt to the specific requirements of the engineering machinery. Further, the generalization capability and the estimation efficiency of the prediction model can be improved by performing migration learning on the basis of the data driving model to perform a battery life prediction model of the battery. Meanwhile, the initial training model, the generalization of transfer learning and the battery life prediction model are fused, so that the residual service life of the lithium battery can be accurately estimated under the condition of lacking a training data set, the accuracy of battery life estimation is improved, and the user experience is improved.
Example two
Fig. 2 is a flowchart of a battery life prediction method provided in the second embodiment, where the present embodiment is applicable to a case of predicting remaining life of a power battery, and the method may be performed by a battery life prediction device, and the battery life prediction device may be implemented in hardware and/or software, and in a specific embodiment, the battery life prediction device may be configured in an electronic device. On the basis of the above embodiment, for determining an initial training model, obtaining target battery parameters of a vehicle battery, determining an initial battery life according to a battery voltage, a battery current, a battery temperature, a cycle number and the initial training model, and predicting the battery life according to the initial battery life and a battery life prediction model, determining the target battery life for optimization, as shown in fig. 2, the method specifically includes the following steps:
s201, acquiring source domain data of a battery.
Wherein the source domain data is a data set of the model pre-trained.
Specifically, the existing battery capacity aging dataset is defined as source domain data, source domain data D S＝{xi,yi},xi is an input dataset, and y i is an output dataset.
S202, performing model training and migration learning according to the source domain data and the basic model to determine an initial training model.
Wherein, the basic model is usually WOA-LSTM (Weighted Optimization Algorithm Long Short-term memory), which is an LSTM model-based optimization algorithm, and the basic idea is to improve the prediction accuracy of the model through weighted optimization. The transfer learning is a machine learning method for improving learning performance, and is a method for transferring a trained model to a new model to help the new model to train. The initial training model is a training model determined by transfer learning.
Specifically, the obtained source domain data is input into a basic model for model training, and then the initial training model is determined by transfer learning of the trained model.
In a specific embodiment, optionally, performing model training and migration learning according to the source domain data and the basic model to determine an initial training model includes: after normalizing the source domain data, determining model parameters of a basic model; determining optimization parameters according to the number of hidden layer nodes, the learning rate, the iteration times and an optimization algorithm; and determining an initial training model according to the optimization parameters, the basic model and the transfer learning.
The model parameters comprise the number of hidden layer nodes, learning rate and iteration times. The number of hidden layer nodes refers to the specific number of hidden layer nodes; the learning rate, which is an important super-parameter in deep learning, determines whether and when the objective function can converge to a local minimum, and the proper learning rate enables the objective function to converge to the local minimum in a proper time. The iteration number refers to the number of iterative training of the model in the training process.
Specifically, after normalization of source domain data, determining model parameters of a basic model, such as the number of hidden layer nodes, learning rate and iteration times; model optimization is carried out according to the number of hidden layer nodes, the learning rate, the iteration times and an optimization algorithm, and optimization parameters are determined; model optimization is carried out according to the optimization parameters and the basic model, an optimized model is determined, and further migration learning is carried out on the optimized model to determine an initial training model.
The method has the advantages that the basic model is optimally trained and subjected to transfer learning, so that the initial training model is determined, the accuracy of the initial training model is improved, and the generalization capability of the initial training model is improved.
S203, acquiring initial battery parameters of the vehicle battery.
Wherein the initial battery parameters include an initial voltage, an initial current, an initial temperature, and an initial number of cycles.
Specifically, initial battery parameters such as initial voltage, initial current, initial temperature and initial cycle times of a vehicle battery of a real vehicle in an actual working process of the vehicle are obtained.
S204, determining target battery parameters according to the initial battery parameters and a preprocessing algorithm.
The target battery parameters of the vehicle battery are aging data sets of the vehicle battery of the real vehicle in the actual working process of the vehicle, and the target battery parameters comprise battery voltage, battery current, battery temperature and cycle times. The battery voltage refers to the operating voltage of the vehicle battery, the battery current refers to the operating current of the vehicle battery, the battery temperature refers to the phenomenon that the surface of the battery generates heat due to chemical, electrochemical change, electron transfer, substance transmission and other reasons of the internal structure of the vehicle battery in the use process, and the heat generated by the vehicle battery can be accumulated in the vehicle battery if the heat cannot be completely dissipated into the environment. The cycle number refers to the charge and discharge number of the vehicle battery, i.e., the process of completing 100% of the complete discharge/charge of the vehicle battery. The preprocessing algorithm is used to process the initial battery parameters.
Specifically, the initial battery parameters are processed by a preprocessing algorithm, for example, the initial battery parameters may be loaded, the missing values in the initial battery parameters are searched according to a search function, a logic matrix is generated, then the missing values are filled according to a filling function, different interpolation methods, such as moving average values or moving median values, may be selected, and finally the processed target battery parameters are determined, which is not limited in this embodiment.
S205, training data and test data are determined according to the battery voltage, the battery current, the battery temperature and the cycle number.
Specifically, after determining the target battery parameter, determining a first proportion of the training data set and a second proportion of the test data set, wherein the sum of the first proportion and the second proportion is 1, determining the training data according to the battery voltage, the battery current, the battery temperature, the cycle number, the training data set and the first proportion, and determining the test data according to the battery voltage, the battery current, the battery temperature, the cycle number, the test data set and the second proportion.
S206, training according to the training data and the initial training model to determine a target training model.
Specifically, the determined training data is input into an initial training model for training, and the trained model is determined to be a target training model.
S207, determining the initial battery life according to the test data and the target training model.
Specifically, the test data is input into a target training model, and prediction is performed according to the test data through the target training model to determine the initial battery life of the vehicle battery.
S208, predicting the battery life according to the initial battery life and the battery life prediction model, and determining the predicted battery life.
The battery life prediction model is a prediction model for optimizing and updating the initial battery life.
Specifically, the initial battery life is input into a battery life prediction model, the initial battery life is optimized through the battery life prediction model, the battery life is further predicted, and the more accurate predicted battery life is determined.
S209, determining the target battery life according to the predicted battery life and the failure threshold.
The failure threshold is a preset battery life, and is used for judging whether the predicted battery life accords with the requirement.
Specifically, if the predicted battery life is determined to be greater than or equal to the failure threshold, determining the failure threshold as the target battery life, and determining that the vehicle battery cannot work normally at this time, wherein the vehicle battery needs to be replaced; if the predicted battery life is determined to be less than the failure threshold, determining that the predicted battery life is equal to the target battery life, determining that the vehicle battery is working normally at the moment, further updating the target battery parameters, and returning to the step of determining the initial battery life according to the battery voltage, the battery current, the battery temperature, the cycle number and the initial training model.
In one embodiment, optionally, determining the target battery life based on the predicted battery life and the failure threshold comprises: determining whether the predicted battery life is greater than or equal to a failure threshold; if the predicted battery life is determined to be greater than or equal to the failure threshold, performing battery replacement; if the predicted battery life is determined to be less than the failure threshold, the target battery parameters are updated and the step of determining an initial battery life based on the battery voltage, the battery current, the battery temperature, the number of cycles, and the initial training model is performed back.
Specifically, after predicting and determining the predicted battery life, determining whether the predicted battery life is greater than or equal to an failure threshold, if the predicted battery life is greater than or equal to the failure threshold, determining that the predicted battery life does not meet the requirements, and the vehicle battery cannot be normally used and needs to be replaced; if the predicted battery life is determined to be smaller than the failure threshold value, the predicted battery life is determined to meet the requirement, the vehicle battery can be used normally, the target battery parameters are further updated, and the step of determining the initial battery life according to the battery voltage, the battery current, the battery temperature, the cycle number and the initial training model is performed.
The device has the advantages that the predicted battery life is determined, the accurate target life battery is determined, the accuracy of vehicle battery life prediction is improved, and the use safety of the vehicle battery is improved.
According to the technical scheme, source domain data of a battery are obtained; model training and migration learning are carried out according to the source domain data and the basic model to determine an initial training model; acquiring initial battery parameters of a vehicle battery; determining a target battery parameter according to the initial battery parameter and a preprocessing algorithm; determining training data and test data according to the battery voltage, the battery current, the battery temperature and the cycle number; training according to the training data and the initial training model to determine a target training model; determining an initial battery life according to the test data and the target training model; predicting the battery life according to the initial battery life and the battery life prediction model, and determining the predicted battery life; and determining the target battery life according to the predicted battery life and the failure threshold. On the basis of the embodiment, the problem that the residual service life of the power battery is difficult to accurately predict due to the fact that real vehicle data of the battery are less and the working environment is complex in the pure electric engineering machinery is solved by fusing the initial training model, the generalization of transfer learning and the battery service life prediction model, the purpose of accurately estimating the residual service life of the lithium battery under the condition of lacking a training data set is achieved, the accuracy of battery service life estimation is improved, and the user experience is improved.
Example III
Fig. 3 is a schematic structural diagram of a battery life prediction apparatus provided in a third embodiment, the apparatus including: a training model determination module 301, a battery parameter acquisition module 302, an initial lifetime determination 303, and a target lifetime determination module 303. Wherein,
A training model determination module 301, configured to determine an initial training model; the initial training model is a training model determined through transfer learning.
A battery parameter obtaining module 302, configured to obtain a target battery parameter of a vehicle battery; the target battery parameters include battery voltage, battery current, battery temperature, and cycle number, among others.
The initial life determining module 303 is configured to determine an initial battery life based on the battery voltage, the battery current, the battery temperature, the number of cycles, and the initial training model.
A target life determining module 303, configured to predict a battery life according to the initial battery life and the battery life prediction model, and determine a target battery life; the battery life prediction model is a prediction model for optimizing and updating the initial battery life.
Optionally, the training model determining module 301 is specifically configured to: acquiring source domain data of a battery; the source domain data is a data set for pre-training the model; and performing model training and migration learning according to the source domain data and the basic model to determine an initial training model.
Optionally, the initial lifetime determination module 303 is specifically configured to: determining training data and test data according to the battery voltage, the battery current, the battery temperature and the cycle number; training according to the training data and the initial training model to determine a target training model; and determining the initial battery life according to the test data and the target training model.
Optionally, the target lifetime determination module 303 is specifically configured to: predicting the battery life according to the initial battery life and the battery life prediction model, and determining the predicted battery life; and determining the target battery life according to the predicted battery life and the failure threshold.
Optionally, the target life determining module 303 is configured to determine the target battery life according to the predicted battery life and the failure threshold, and is specifically configured to: determining whether the predicted battery life is greater than or equal to a failure threshold; if the predicted battery life is determined to be greater than or equal to the failure threshold, performing battery replacement; if the predicted battery life is determined to be less than the failure threshold, the target battery parameters are updated and the step of determining an initial battery life based on the battery voltage, the battery current, the battery temperature, the number of cycles, and the initial training model is performed back.
Optionally, the battery parameter obtaining module 302 is specifically configured to: acquiring target battery parameters of a vehicle battery, comprising: acquiring initial battery parameters of a vehicle battery; and determining the target battery parameters according to the initial battery parameters and the preprocessing algorithm.
Optionally, the training model determining module 301 performs model training and migration learning according to the source domain data and the base model to determine an initial training model, which is specifically configured to: after normalizing the source domain data, determining model parameters of a basic model; the model parameters comprise the number of hidden layer nodes, learning rate and iteration times; determining optimization parameters according to the number of hidden layer nodes, the learning rate, the iteration times and an optimization algorithm; and determining an initial training model according to the optimization parameters, the basic model and the transfer learning.
The battery life prediction device provided by the embodiment can execute the battery life prediction method provided by any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.
Example IV
Fig. 4 is a schematic diagram of the structure of an electronic device provided in the fourth embodiment, which is intended to represent various forms of digital computers, such as laptops, desktops, workstations, personal digital assistants, servers, blade servers, mainframes, and other appropriate computers. Electronic equipment may also represent various forms of mobile devices, such as personal digital processing, cellular telephones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions, are meant to be exemplary only, and are not meant to limit implementations of the inventions described and/or claimed herein.
As shown in fig. 4, the electronic device 10 includes at least one processor 11, and a memory communicatively connected to the at least one processor 11, such as a Read Only Memory (ROM) 12, a random access memory (also referred to as "random access memory", random Access Memory, RAM) 13, etc., in which the memory stores a computer program executable by the at least one processor, and the processor 11 may perform various suitable actions and processes according to the computer program stored in the Read Only Memory (ROM) 12 or the computer program loaded from the storage unit 18 into the Random Access Memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 may also be stored. The processor 11, the ROM 12 and the RAM 13 are connected to each other via a bus 14. An input/output (I/O) interface 15 is also connected to bus 14.
Various components in the electronic device 10 are connected to the I/O interface 15, including: an input unit 16 such as a keyboard, a mouse, etc.; an output unit 17 such as various types of displays, speakers, and the like; a storage unit 18 such as a magnetic disk, an optical disk, or the like; and a communication unit 19 such as a network card, modem, wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information/data with other devices via a computer network, such as the internet, and/or various telecommunication networks.
The processor 11 may be a variety of general and/or special purpose processing components having processing and computing capabilities. Some examples of processor 11 include, but are not limited to, a Central Processing Unit (CPU), a Graphics Processing Unit (GPU), various specialized Artificial Intelligence (AI) computing chips, various processors running machine learning model algorithms, digital Signal Processors (DSPs), and any suitable processor, controller, microcontroller, etc. The processor 11 performs the various methods and processes described above, such as a battery life prediction method.
In some embodiments, the battery life prediction method may be implemented as a computer program tangibly embodied on a computer-readable storage medium, such as the storage unit 18. In some embodiments, part or all of the computer program may be loaded and/or installed onto the electronic device 10 via the ROM 12 and/or the communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the above-described battery life prediction method may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the battery life prediction method in any other suitable manner (e.g., by means of firmware).
Various implementations of the systems and techniques described here above can be implemented in digital electronic circuitry, integrated circuit systems, field Programmable Gate Arrays (FPGAs), application Specific Integrated Circuits (ASICs), application Specific Standard Products (ASSPs), systems On Chip (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and/or combinations thereof. These various embodiments may include: implemented in one or more computer programs, the one or more computer programs may be executed and/or interpreted on a programmable system including at least one programmable processor, which may be a special purpose or general-purpose programmable processor, that may receive data and instructions from, and transmit data and instructions to, a storage system, at least one input device, and at least one output device.
A computer program for carrying out methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general purpose computer, special purpose computer, or other programmable data processing apparatus, such that the computer programs, when executed by the processor, cause the functions/acts specified in the flowchart and/or block diagram block or blocks to be implemented. The computer program may execute entirely on the machine, partly on the machine, as a stand-alone software package, partly on the machine and partly on a remote machine or entirely on the remote machine or server.
In the context of the present invention, a computer-readable storage medium may be a tangible medium that can contain, or store a computer program for use by or in connection with an instruction execution system, apparatus, or device. The computer readable storage medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any suitable combination of the foregoing. Alternatively, the computer readable storage medium may be a machine readable signal medium. More specific examples of a machine-readable storage medium would include an electrical connection based on one or more wires, a portable computer diskette, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.
To provide for interaction with a user, the systems and techniques described here can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to a user; and a keyboard and a pointing device (e.g., a mouse or a trackball) through which a user can provide input to the electronic device. Other kinds of devices may also be used to provide for interaction with a user; for example, feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form, including acoustic input, speech input, or tactile input.
The systems and techniques described here can be implemented in a computing system that includes a background component (e.g., as a data server), or that includes a middleware component (e.g., an application server), or that includes a front-end component (e.g., a user computer having a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described here), or any combination of such background, middleware, or front-end components. The components of the system can be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: local Area Networks (LANs), wide Area Networks (WANs), blockchain networks, and the internet.
The computing system may include clients and servers. The client and server are typically remote from each other and typically interact through a communication network. The relationship of client and server arises by virtue of computer programs running on the respective computers and having a client-server relationship to each other. The server can be a cloud server, also called a cloud computing server or a cloud host, and is a host product in a cloud computing service system, so that the defects of high management difficulty and weak service expansibility in the traditional physical hosts and VPS service are overcome.
It should be appreciated that various forms of the flows shown above may be used to reorder, add, or delete steps. For example, the steps described in the present invention may be performed in parallel, sequentially, or in a different order, so long as the desired results of the technical solution of the present invention are achieved, and the present invention is not limited herein.
The above embodiments do not limit the scope of the present invention. It will be apparent to those skilled in the art that various modifications, combinations, sub-combinations and alternatives are possible, depending on design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of the present invention.
Claims (10)
1. A method for predicting battery life, the method comprising:
determining an initial training model; the initial training model is a training model determined through transfer learning;
acquiring target battery parameters of a vehicle battery; wherein the target battery parameters include battery voltage, battery current, battery temperature, and cycle number;
Determining an initial battery life from the battery voltage, the battery current, the battery temperature, the number of cycles, and the initial training model;
Predicting the battery life according to the initial battery life and the battery life prediction model, and determining the target battery life; the battery life prediction model is a prediction model for optimizing and updating the initial battery life.
2. The method of claim 1, wherein said determining an initial training model comprises:
Acquiring source domain data of a battery; the source domain data is a data set for pre-training of the model;
And performing model training and migration learning according to the source domain data and the basic model to determine the initial training model.
3. The method of claim 1, wherein said determining an initial battery life based on said battery voltage, said battery current, said battery temperature, said number of cycles, and said initial training model comprises:
Determining training data and test data based on the battery voltage, the battery current, the battery temperature, and the number of cycles;
training according to the training data and the initial training model to determine a target training model;
and determining the initial battery life according to the test data and the target training model.
4. The method according to claim 1, wherein the predicting the battery life based on the initial battery life and the battery life prediction model, determining the target battery life, comprises:
Predicting the battery life according to the initial battery life and the battery life prediction model, and determining the predicted battery life;
And determining the target battery life according to the predicted battery life and the failure threshold.
5. The method of claim 4, wherein said determining a target battery life from said predicted battery life and a failure threshold comprises:
determining whether the predicted battery life is greater than or equal to the failure threshold;
If the predicted battery life is determined to be greater than or equal to the failure threshold, performing battery replacement;
If the predicted battery life is determined to be less than the failure threshold, updating the target battery parameter, and returning to the step of performing the initial battery life determination according to the battery voltage, the battery current, the battery temperature, the number of cycles, and the initial training model.
6. The method of claim 1, wherein the obtaining the target battery parameter of the vehicle battery comprises:
Acquiring initial battery parameters of the vehicle battery;
and determining the target battery parameters according to the initial battery parameters and a preprocessing algorithm.
7. The method of claim 2, wherein said determining the initial training model from model training and migration learning based on the source domain data and a base model comprises:
after normalizing the source domain data, determining model parameters of the basic model; the model parameters comprise the number of hidden layer nodes, learning rate and iteration times;
Determining optimization parameters according to the number of hidden layer nodes, the learning rate, the iteration times and an optimization algorithm;
and determining the initial training model according to the optimization parameters, the basic model and the transfer learning.
8. A battery life prediction apparatus, comprising:
The training model determining module is used for determining an initial training model; the initial training model is a training model determined through transfer learning;
the battery parameter acquisition module is used for acquiring target battery parameters of the vehicle battery; wherein the target battery parameters include battery voltage, battery current, battery temperature, and cycle number;
An initial life determining module configured to determine an initial battery life based on the battery voltage, the battery current, the battery temperature, the number of cycles, and the initial training model;
The target life determining module is used for predicting the life of the battery according to the initial battery life and the battery life prediction model, and determining the target battery life; the battery life prediction model is a prediction model for optimizing and updating the initial battery life.
9. An electronic device, the electronic device comprising:
At least one processor; and a memory communicatively coupled to the at least one processor; wherein,
The memory stores a computer program executable by the at least one processor to enable the at least one processor to perform the battery life prediction method of any one of claims 1-7.
10. A computer readable storage medium storing computer instructions for causing a processor to perform the method of predicting battery life of any one of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410091098.XA CN117907872A (en) | 2024-01-22 | 2024-01-22 | Battery life prediction method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410091098.XA CN117907872A (en) | 2024-01-22 | 2024-01-22 | Battery life prediction method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117907872A true CN117907872A (en) | 2024-04-19 |
Family
ID=90685574
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410091098.XA Pending CN117907872A (en) | 2024-01-22 | 2024-01-22 | Battery life prediction method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117907872A (en) |
-
2024
- 2024-01-22 CN CN202410091098.XA patent/CN117907872A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112036084B (en) | Similar product life migration screening method and system | |
| Jia et al. | Improved sparrow search algorithm optimization deep extreme learning machine for lithium-ion battery state-of-health prediction | |
| CN106055775B (en) | A kind of service life of secondary cell prediction technique that particle filter is combined with mechanism model | |
| Wang et al. | A method based on improved ant lion optimization and support vector regression for remaining useful life estimation of lithium‐ion batteries | |
| CN113205233B (en) | Lithium battery life prediction method based on wolf algorithm and multi-core support vector regression | |
| CN115513951B (en) | Power load prediction method and system based on concept drift detection | |
| CN115221795A (en) | Training method, prediction method, device, equipment and medium of capacity prediction model | |
| CN116307215A (en) | Load prediction method, device, equipment and storage medium of power system | |
| CN114721345A (en) | Industrial control method, device and system based on reinforcement learning and electronic equipment | |
| CN116087787A (en) | Battery fault judging method and system based on principal component analysis method | |
| CN117407795A (en) | Battery safety prediction method and device, electronic equipment and storage medium | |
| CN117332898A (en) | New energy small time scale power time sequence rolling prediction method based on machine learning | |
| CN117332896A (en) | New energy small time scale power prediction method and system for multilayer integrated learning | |
| CN116125279A (en) | Method, device, equipment and storage medium for determining battery health state | |
| CN116859255A (en) | Method, device, equipment and medium for predicting state of health of energy storage battery | |
| CN117907872A (en) | Battery life prediction method and device, electronic equipment and storage medium | |
| CN115951236A (en) | Lithium battery state monitoring method, system, device and storage medium | |
| CN115099163A (en) | Charge state determining method and device, electronic equipment and storage medium | |
| CN117517971A (en) | Battery electric quantity prediction method and device, electronic equipment and storage medium | |
| CN118033461A (en) | Method and device for evaluating battery health state and electronic equipment | |
| CN116819342A (en) | Battery life curve determining method and device, electronic equipment and storage medium | |
| Miaomiao et al. | Indirect Prediction Method for Remaining Useful Life of Lithium-ion Battery based on Gray Wolf Optimized Extreme Learning Machine | |
| Han et al. | SOC estimation for lithium-ion batteries based on BiGRU with SE attention and Savitzky-Golay filter | |
| CN116307252A (en) | Method, device, equipment and medium for predicting operation state of service system | |
| CN117557002A (en) | Model determination method, trend prediction device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination |