CN116061690A - Safety early warning method and device in electric automobile charging process - Google Patents
Safety early warning method and device in electric automobile charging process Download PDFInfo
- Publication number
- CN116061690A CN116061690A CN202310085672.6A CN202310085672A CN116061690A CN 116061690 A CN116061690 A CN 116061690A CN 202310085672 A CN202310085672 A CN 202310085672A CN 116061690 A CN116061690 A CN 116061690A
- Authority
- CN
- China
- Prior art keywords
- data
- value
- charging
- model
- early warning
- 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
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60L—PROPULSION OF ELECTRICALLY-PROPELLED VEHICLES; SUPPLYING ELECTRIC POWER FOR AUXILIARY EQUIPMENT OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRODYNAMIC BRAKE SYSTEMS FOR VEHICLES IN GENERAL; MAGNETIC SUSPENSION OR LEVITATION FOR VEHICLES; MONITORING OPERATING VARIABLES OF ELECTRICALLY-PROPELLED VEHICLES; ELECTRIC SAFETY DEVICES FOR ELECTRICALLY-PROPELLED VEHICLES
- B60L3/00—Electric devices on electrically-propelled vehicles for safety purposes; Monitoring operating variables, e.g. speed, deceleration or energy consumption
- B60L3/0023—Detecting, eliminating, remedying or compensating for drive train abnormalities, e.g. failures within the drive train
- B60L3/0046—Detecting, eliminating, remedying or compensating for drive train abnormalities, e.g. failures within the drive train relating to electric energy storage systems, e.g. batteries or capacitors
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/60—Other road transportation technologies with climate change mitigation effect
- Y02T10/70—Energy storage systems for electromobility, e.g. batteries
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Life Sciences & Earth Sciences (AREA)
- Sustainable Development (AREA)
- Sustainable Energy (AREA)
- Transportation (AREA)
- Mechanical Engineering (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a safety early warning method and device in the charging process of an electric automobile, and relates to the technical field of data processing. According to the method, based on the fault characteristics existing in the historical charging process of the electric vehicle, the actual voltage value, the current difference value, the voltage change rate, the current change rate, the temperature difference value, the number of times of triggering the emergency stop button and other multidimensional evaluation indexes are determined to be subjected to information fusion, the health grade state predicted value is constructed, the early warning grades are divided, then the deep learning is carried out by utilizing the generated countermeasure network based on the result, the running state predicted model is obtained, the charging data acquired in real time is input into the running state predicted model to be identified, and the charging running state early warning grade is obtained. The method for early warning the charging faults of the electric automobile has the advantages of high early warning precision, high recognition degree and high recognition rate, can identify the faults in time, and greatly promotes the development of the electric automobile and related industries.
Description
Technical Field
The invention relates to the technical field of data processing, in particular to a safety early warning method and device in the charging process of an electric automobile.
Background
In the global scope, the ecological environment and energy crisis problems are increasingly prominent, and compared with the traditional fuel oil automobiles, the electric automobile has great advantages in the aspects of saving petroleum resources and reducing carbon emission, and is valued by various national governments and automobile enterprises. However, in the charging process of the electric automobile, spontaneous combustion accidents may occur, which hinders the development of the electric automobile industry. It was found that over-temperature of the battery is an important cause of spontaneous combustion in charging of the electric vehicle. Therefore, a temperature early warning model of the electric vehicle charging process is built, the temperature of the electric vehicle battery is monitored in real time and the safety early warning is carried out, the charging safety can be ensured, and the sustainable development of the electric vehicle industry is facilitated.
However, most of the fault diagnoses of the existing electric vehicles in the charging process depend on expert experience, and problems cannot be found timely and damage cannot be prevented timely. In addition, the fault diagnosis system of the existing electric vehicle charging early warning device is not perfect, and the safety early warning precision of the electric vehicle is also poor.
Disclosure of Invention
The invention aims to provide a safety early warning method and a device in the charging process of an electric automobile so as to solve the problems. In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:
in a first aspect, the present application provides a safety early warning method in a charging process of an electric automobile, including:
acquiring historical data and real-time data of charging of the electric vehicle based on license plate codes of the electric vehicle respectively, and preprocessing the historical data and the real-time data respectively to obtain first data and second data, wherein the first data and the second data are physical quantities capable of representing faults in the charging process;
extracting an evaluation index based on the first data, and respectively determining a corresponding weight value and membership;
constructing a health grade state predicted value by using a weighted average principle based on the first data, the weight value and the membership;
training by generating an countermeasure network model based on the first data and the health grade state predicted value to obtain an operation state predicted model;
and inputting the second data into the running state prediction model to obtain a charging running state early warning grade.
In a second aspect, the application further provides a safety early warning device in electric automobile charging process, including obtaining the module, draw the module, construct module, training module and early warning module, wherein: the acquisition module is used for: the method comprises the steps of respectively obtaining historical data and real-time data of charging of the electric vehicle based on license plate codes of the electric vehicle, respectively preprocessing the historical data and the real-time data to correspondingly obtain first data and second data, wherein the first data and the second data are physical quantities capable of representing faults in the charging process;
and an extraction module: the method comprises the steps of extracting an evaluation index based on first data, and respectively determining a corresponding weight value and membership;
the construction module comprises: the health grade state prediction value is constructed by using a weighted average principle based on the first data, the weight value and the membership;
training module: the system is used for training by generating an countermeasure network model based on the first data and the health grade state predicted value to obtain an operation state predicted model;
and the early warning module is used for: and the second data is input into the running state prediction model to obtain a charging running state early warning grade.
The beneficial effects of the invention are as follows:
according to the method, based on the fault characteristics existing in the historical charging process of the electric vehicle, physical quantities which can represent faults in the charging process in historical data are determined, information fusion is carried out on multidimensional evaluation indexes such as actual voltage values, current difference values, voltage change rates, current change rates, temperature difference values, the number of times of triggering emergency stop buttons and the like after preprocessing (namely first data), health grade state predicted values are constructed, early warning grades are divided, then deep learning is carried out by utilizing a generated countermeasure network based on the results, an operation state predicted model is obtained, and second data obtained after preprocessing charging data collected in real time are input into the operation state predicted model for recognition, so that the early warning grade of the charging operation state is obtained. Compared with the method that the single evaluation index acquired in real time deviates from a preset value for early warning, the method and the device realize early warning of the charging faults of the electric automobile by information fusion through monitoring various evaluation index data of the charging of the electric automobile, not only can effectively avoid false alarms caused by wrong charging data, but also have higher precision, high identification degree and higher speed of early warning grade results, can identify the fault problems in time, and greatly promote the development of the electric automobile and related industries.
In the preprocessing method for the charging data, abnormal values are removed based on the charging data corresponding to the same type of data I D and then filled, so that the data integrity is ensured, meanwhile, the interference of data defects on a later prediction result is avoided, and the accuracy of an early warning result is improved. And then discretizing the complete data by using a time factor, wherein the obtained discrete data is more accurate than the corresponding original data, and accords with the data characteristic input criterion of the running state prediction model, and the discretized data is substituted into the running state prediction model in the later period for running, so that the running performance of the model can be improved. Then, normalization processing is carried out to accelerate gradient descent, so that convergence of an antagonistic network model can be accelerated and generated, the optimal solution time of the model is shortened, the precision of the model can be improved, and gradient explosion is prevented.
In the construction of the health grade state predicted value, the index constant weight given by the analytic hierarchy process is considered to be not changed when a certain evaluation index is deteriorated in the charging process, so that the accuracy of evaluating a large amount of charging data is affected, and the sub-item layer and the evaluation index layer constant weight are respectively corrected, so that the problem of changeable degradation degree based on the charging data is solved, and the accuracy of an early warning result is improved.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of the embodiments of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims thereof as well as the appended drawings.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention and should not be considered as limiting the scope, and that other related drawings can be obtained according to these drawings without inventive effort for a person having ordinary skill in the art.
Fig. 1 is a schematic flow chart of a safety early warning method in the charging process of an electric automobile according to an embodiment of the invention;
fig. 2 is a schematic structural diagram of a safety early warning device in the charging process of an electric vehicle according to an embodiment of the invention;
fig. 3 is a schematic structural diagram of a safety early warning method device in the charging process of an electric vehicle according to an embodiment of the present invention.
In the figure: 710 an acquisition module; 711-a tag unit; 712-modifying unit; 713-shim cells; 714-discrete units; 7141-acquisition unit; 7142 a transform unit; 7143-multidimensional cell; 715-a preprocessing unit; 720-an extraction module; 721-an evaluation unit; 722-a judging unit; 723-a calculation unit; 724-a judging unit; 725-a first statistics unit; 726-a second statistics unit; 730-building a module; 740-training module; 741-a learning unit; 750-an early warning module; 800-safety precaution method equipment; 801-a processor; an 802-memory; 803-multimedia component; 804-I/O interface; 805-a communication component.
Detailed Description
For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. The components of the embodiments of the present invention generally described and illustrated in the figures herein may be arranged and designed in a wide variety of different configurations. Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by one of ordinary skill in the art without inventive effort, are within the scope of the present invention based on the embodiments of the present invention.
It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further definition or explanation thereof is necessary in the following figures. Meanwhile, in the description of the present invention, the terms "first", "second", and the like are used only to distinguish the description, and are not to be construed as indicating or implying relative importance.
Example 1:
the embodiment provides a safety early warning method in the charging process of an electric automobile.
Referring to fig. 1, the method is shown in fig. 1 to comprise step S1, step S2, step S3 and step S4 and step S5, wherein:
step S1, historical data and real-time data of electric vehicle charging are respectively obtained based on license plate codes of the electric vehicle, first data and second data are respectively obtained through corresponding pretreatment, and the first data and the second data are physical quantities capable of representing faults in the charging process.
It can be understood that in this step, the historical charging data of the electric vehicle in the database is utilized based on the unique code of the license plate code of the electric vehicle, and meanwhile, the real-time charging data of the electric vehicle charging is collected through various sensors in the charging pile and the information collecting equipment. And then, respectively preprocessing missing data, interference information and the like in the historical data and the real-time data, so that the data integrity is ensured, the interference of data defects on a later prediction result can be avoided, and the accuracy of an early warning result is improved. It should be noted that, in the historical data and the real-time data in this embodiment, when a fault occurs in the charging process, parameters such as an actual voltage value, a current difference value, a voltage change rate, a current change rate, a temperature difference value, and the number of times of touching the scram button may change obviously, so that the data types are collected from the historical data and the real-time data for preprocessing.
The pretreatment method includes step S11, step S12, step S13, step S14 and step S15.
And step S11, marking according to data types based on the historical data or the real-time data respectively to obtain type data IDs.
It will be appreciated that in this step, the type data ID is obtained by marking with different ID labels, respectively, according to the type of data based on the history data or the real-time data. For example, the actual voltage value, the current difference value, the voltage change rate, the current change rate, the temperature difference value, and the type data ID corresponding to the number of times of touching the scram button are A, B, C, D, E and F, respectively.
And step S12, modifying based on the time sequence and the type data ID to obtain the characteristic sequences under the same dimension and different moments.
It can be understood that in this step, the ID of each acquired information is modified for each type of data based on the time sequence, respectively, so as to obtain the feature sequences at different moments in the same dimension. For example, the type data IDs of the actual voltage values acquired at the nth, n+1th, n+2th, n+3th, n+4th, and n+5th seconds are a+1, a+2, a+3, a+4, and a+5, respectively, and the characteristic sequences are composed of a+1, a+2, a+3, a+4, and a+5.
And step S13, eliminating abnormal values based on all the characteristic sequences respectively, and filling missing values to obtain complete data.
It will be appreciated that in this step, the rejection is performed according to the range of outliers preset for each different type of data. And filling empty values appearing during data acquisition and missing data after abnormal values are removed by adopting a multi-order Lagrange interpolation method respectively to obtain complete data so as to ensure the integrity of an electric vehicle charging data set, providing training data for training a later generation countermeasure network model, and improving the overall performance of an operation state prediction model.
And step S14, performing discrete processing based on the complete data to obtain discrete data.
It can be understood that in this step, continuous complete data is converted into a section of discretized interval, so as to obtain more accurate discrete data than corresponding original data, so that the discrete data is more in accordance with the data feature input criterion of the running state prediction model, and the discretized data is substituted into the later running state prediction model to run, so that the running performance of the model can be improved. The discretization processing method comprises an equidistant discretization method, an equal-frequency discretization method, a K-means model discretization method and the like.
Further, the above discrete method includes step S141, step S142, and step S143.
And step S141, collecting the complete data by using a standard time interval to form a time sequence vector.
It can be understood that, in this step, parameters corresponding to the interval points are acquired in the complete data based on the same time interval, and a time series vector is constructed based on the time sequence corresponding to the complete data at each interval point, and the time series vector corresponding to the actual voltage value is x= { x m1 ,x m2 …,x mn -wherein x is a time series vector; x is x mn Is the charging operation state data at the nth time interval of the mth type.
And step S142, changing through a time factor based on the time sequence vector to obtain different standard time sequence vectors.
It can be understood that in this step, the time scale factor is introduced to discretize the time sequence vector according to the formula (1), so as to achieve the goal of freely selecting the standard time sequence vector under different time lengths and improve the accuracy of discrete data acquisition.
Wherein:
is a standard time series vector; τ is a time factor; m is the total number of time interval points; n is the number of time-separated points in the time sequence; k is charging running state data of different data types; x is x mk And the discrete charge operation state data corresponding to m time intervals corresponding to the kth charge operation state data.
And S143, constructing a composite vector based on the different standard time sequence vectors to obtain discrete data.
It will be appreciated that in this step, discrete data is constructed according to equation (2):
wherein: composite vector
Is a vector of order 1×h; τ is a time factor; m is a data type, namely different dimensions of the data; d, d j Is a standard time series vector; h is a m Is a dimension vector; η (eta) m Is a delay time vector. In order to ensure that the nonlinear relation of each running state data is not changed, a dimension vector h is embedded in the formula (2) m The delay time vector eta is introduced for compressing the running state data and not losing the data m Discrete data based on multi-dimensional information fusion and delayed transmission are obtained, so that the integrity of data transmission can be guaranteed, and the accuracy of an early warning result can be improved.
And step S15, carrying out normalization processing based on the discrete data, and correspondingly obtaining first data or second data.
It can be appreciated that in this step, there may be a large difference in the range and the variation characteristics of different types of charging data, and if training is performed to generate the countermeasure network directly using the original data, large data noise may be introduced, so that a large deviation occurs in the final obtained early warning result. After normalization processing, discrete data can be calculated in the same dimension, so that different kinds of charging data can have the same importance degree in the operation process of generating the reactance network. After the charging data features are normalized, gradient descent can be accelerated, convergence of an antagonistic network model can be accelerated to be generated, optimal solution time of the model is shortened, accuracy of the model can be improved, and gradient explosion is prevented. The normalization method includes 0-1 normalization and Z-score normalization.
And S2, extracting an evaluation index based on the first data, and respectively determining a corresponding weight value and membership degree.
It can be understood that in this step, the weight value is determined according to the importance degree of each evaluation index in the charge health state evaluation by using algorithms such as expert scoring, PCA principal component analysis, pearson correlation coefficient, and the like. And simultaneously, determining the probability of the membership of the corresponding set of each evaluation index by adopting a fuzzy membership analysis method.
In detail, step S2 includes step S21, step S22, step S23, step S24, step S25, and step S26.
And S21, constructing an operation state evaluation system by using a analytic hierarchy process based on the first data.
It can be understood that in this step, a three-layered structure model of a target layer, a criterion layer, and a solution layer formed sequentially from top to bottom is constructed by using a hierarchical analysis method based on the preprocessed historical data, wherein the target layer is an operation state evaluation system (respectively corresponding to health, sub-health, abnormal, serious, and other operation states), the criterion layer is an electrical evaluation index, a temperature evaluation index, and the like, and the solution layer is an actual voltage value, a current change rate, a voltage change rate, and the like corresponding to the electrical evaluation index, and a temperature difference value and the like corresponding to the temperature evaluation index.
And S22, constructing an index judgment matrix based on the running state evaluation system.
It can be understood that in this step, the evaluation indexes are compared in pairs based on the three-layer structural model layer by layer to obtain a relation of relative importance, and each evaluation index is scored by using a nine-score scale method to obtain a discrimination matrix, where the discrimination matrix is shown in formula (3):
A＝(a ij ) n×n (3)
wherein: a is a discrimination matrix; a, a ij Dividing the evaluation index i and the evaluation index j of the current level of the judgment matrix into importance degrees of the last level; i and j are respectively different kinds of evaluation indexes; n is the dimension of the hierarchical model.
And S23, respectively calculating the eigenvector and the maximum eigenvalue based on the index judgment matrix.
And S24, calculating a consistency ratio based on the maximum characteristic value, determining a weight value according to the characteristic vector if the consistency ratio is smaller than 0.1, and otherwise, reallocating the weight value.
It will be appreciated that in this step, the consistency ratio of the metric judgment matrix from the consistency is obtained based on the maximum eigenvalue and equation (4):
wherein: r is the consistency ratio; lambda (lambda) max The maximum eigenvalue of the matrix is judged; n is the order of the discrimination matrix; e is an average random uniformity index. When R is<And when 0.1, the judgment matrix meets the consistency index, which indicates that the weight distribution is reasonable. However, the Chang Quan heavy value corresponding to the feature vector of the charging data cannot be adjusted due to the change of each feature index of the charging data, so that the accuracy of evaluating a large amount of charging data is affected, the constant weight value is corrected according to the formula (5), the problem that the degradation degree generated based on the charging data is changeable is solved, and the accuracy of the early warning result is improved:
wherein: a, a ij And
dividing into discrimination matrix typesThe evaluation index i and the evaluation index j of the current level pair the Chang Quan weight value and the variable weight value of the previous level; i and j are respectively different kinds of evaluation indexes; n is the dimension of the hierarchical structure model; m is the number of data types under the ith evaluation index; alpha is a weight adjustment rate index; lambda (lambda) ik Is the degree of degradation. If R is more than or equal to 0.1, the consistency ratio is not satisfied, and the weight value is required to be reassigned.
Step S25, determining a sample domain based on the first data, and determining an operation state set based on the sample domain; the running state set is a running data range corresponding to different health levels.
It can be understood that in this step, a sample field corresponding to the first data after the history charging data is preprocessed is set as U, and any charging data sample U is selected 0 E U, then obtaining a variable charging pile set A according to a movable boundary of the charging data sample field U * 。
Step S26, respectively counting based on the sample domain and the running state set to obtain membership degree, wherein the membership degree is a limit value of the occurrence frequency of the running state set in the sample domain.
It will be appreciated that in this step, the degree of membership is calculated according to equation (6):
wherein: μ is the membership; u (U) 0 For any charge data sample; a is that * Is a variable set of charging posts.
And S3, constructing a health grade state predicted value based on the first data, the weight value, the membership degree and a weighted average principle.
It will be appreciated that in this step, the health index is calculated based on equation (7), and the health classification based on the health index is as shown in table 1:
wherein: HI is a health index matrix; d is the dimension of charging data; omega i First data for an ith dimension; a, a ij A weight value corresponding to each evaluation index weight; mu (mu) ij And (5) evaluating the membership degree corresponding to the index weight for each. Let the health class be classified into four operating state classes of health, sub-health, abnormal and severe:
TABLE 1 health class division table
And step S4, training by generating an countermeasure network model based on the first data and the health grade state predicted value to obtain an operation state predicted model.
It can be understood that in this step, deep learning is used to learn each evaluation index and health grade state in the charging process, and training is performed on the generated countermeasure network model based on the first data corresponding to the time sequence, so as to obtain an operation state prediction model, and implement real-time prediction of charging data corresponding to the next time, so as to improve the recognition rate of the health grade.
In detail, the training method of the operation state prediction model includes step S41: based on the first data and the health grade state predicted value, respectively carrying out alternate training on a generator model and a discriminator model for generating an countermeasure network model, wherein one training cycle process of the alternate training is as follows: taking the first data as an input value of the generator model to obtain generated data; and then taking the generated data and the health grade state predicted value as input values of the discriminator model, carrying out weight updating by utilizing a loss value and a residual value fed back by the discriminator model, minimizing the residual value and the loss value, and obtaining an operation state predicted model, wherein the residual value is a Wasserstein distance between the health grade state predicted value and the generated data. The Wasserstein distance is used as the equivalent optimized distance measurement to solve the problems that the gradient vanishing, the gradient instability, the instability of the training model and the like of the generated countermeasure network model are easy to occur due to unreasonable equivalent optimized distance measurement mode.
And S5, inputting the second data into the running state prediction model to obtain a charging running state early warning grade.
It can be understood that in this step, the second data is input into the running state prediction model to predict the running state of the health level corresponding to the next moment in real time, and the corresponding early warning level of the charging running state is obtained according to the early warning level classification condition in table 1.
Example 2:
as shown in fig. 2, the embodiment provides a safety early warning device in the charging process of an electric automobile, which includes an acquisition module 710, an extraction module 720, a construction module 730, a training module 740 and an early warning module 750, wherein:
acquisition module 710: the method is used for respectively acquiring historical data and real-time data of charging of the electric vehicle based on license plate codes of the electric vehicle, respectively preprocessing the historical data and the real-time data and correspondingly acquiring first data and second data, wherein the first data and the second data are physical quantities capable of representing faults in the charging process.
Optionally, the acquiring module 710 includes a marking unit 711, a modifying unit 712, a padding unit 713, a discrete unit 714, and a preprocessing unit 715, where:
the marking unit 711: the method comprises the steps of marking according to data types based on historical data or real-time data respectively to obtain type data IDs;
modification unit 712: the method is used for modifying based on the time sequence and the type data ID to obtain a characteristic sequence under the same dimension at different moments;
the shim cell 713: the method comprises the steps of removing abnormal values based on all the characteristic sequences respectively, filling missing values, and obtaining complete data;
discrete unit 714: and the discrete processing is used for carrying out discrete processing based on the complete data to obtain discrete data.
Further, the discrete unit 714 includes an acquisition unit 7141, a transformation unit 7142, and a multi-dimensional unit 7143, wherein:
acquisition unit 7141: the system is used for acquiring the complete data by using standard time intervals to form a time sequence vector;
a transform unit 7142: the time sequence vector is used for obtaining different standard time sequence vectors based on the time sequence vector through the change of the time factor;
multidimensional unit 7143: and the method is used for constructing a composite vector based on the different standard time sequence vectors to obtain discrete data.
The preprocessing unit 715: and the normalization processing is performed based on the discrete data, so that first data or second data are correspondingly obtained.
Extraction module 720: and the evaluation index extraction module is used for extracting the evaluation index based on the first data and respectively determining the corresponding weight value and membership degree.
Optionally, the extracting module 720 includes an evaluating unit 721, a judging unit 722, a calculating unit 723, a judging unit 724, a first statistics unit 725 and a second statistics unit 726, wherein:
an evaluation unit 721: the method comprises the steps of constructing an operation state evaluation system based on the first data by using a analytic hierarchy process;
the evaluation unit 722: the method comprises the steps of constructing an index judgment matrix based on the running state evaluation system;
a calculation unit 723: the method comprises the steps of respectively calculating a feature vector and a maximum feature value based on the index judgment matrix;
determination unit 724: the method comprises the steps of calculating a consistency ratio based on the maximum eigenvalue, determining a weight value according to the eigenvector if the consistency ratio is smaller than 0.1, otherwise, reassigning the weight value;
the first statistics unit 725: for determining a sample domain based on the first data and a set of operating states based on the sample domain; the running state set is a running data range corresponding to different health levels;
a second statistics unit 726: and the system is used for respectively carrying out statistics based on the sample domain and the running state set to obtain membership degree, wherein the membership degree is a limit value of the occurrence frequency of the running state set in the sample domain.
The construction module 730: the health grade state prediction value is constructed by using a weighted average principle based on the first data, the weight value and the membership;
training module 740: and training by generating an countermeasure network model based on the first data and the health grade state predicted value to obtain an operation state predicted model.
In detail, the training module 740 includes a learning unit 741, wherein:
learning unit 741: and the training cycle process for alternately training the generator model and the discriminator model for generating the countermeasure network model based on the first data and the health grade state predicted value is as follows: taking the first data as an input value of the generator model to obtain generated data; and then taking the generated data and the health grade state predicted value as input values of the discriminator model, carrying out weight updating by utilizing a loss value and a residual value fed back by the discriminator model, minimizing the residual value and the loss value, and obtaining an operation state predicted model, wherein the residual value is a Wasserstein distance between the health grade state predicted value and the generated data.
Early warning module 750: and the second data is input into the running state prediction model to obtain a charging running state early warning grade.
It should be noted that, regarding the apparatus in the above embodiments, the specific manner in which the respective modules perform the operations has been described in detail in the embodiments regarding the method, and will not be described in detail herein.
Example 3:
corresponding to the above method embodiment, the present embodiment further provides a safety pre-warning method device 800 in an electric vehicle charging process, where the safety pre-warning method device 800 in an electric vehicle charging process described below and the safety pre-warning method in an electric vehicle charging process described above may be referred to correspondingly.
Fig. 3 is a block diagram illustrating a safety precaution method apparatus 800 during charging of an electric vehicle, according to an exemplary embodiment. As shown in fig. 3, the safety precaution method device 800 in the electric automobile charging process may include: a processor 801, a memory 802. The safety precaution method device 800 in the electric automobile charging process may further include one or more of a multimedia component 803, an i/O interface 804, and a communication component 805.
The processor 801 is configured to control overall operation of the safety early warning method device 800 in the electric vehicle charging process, so as to complete all or part of the steps in the safety early warning method in the electric vehicle charging process. The memory 802 is used to store various types of data to support the operation of the safety precaution method device 800 during the charging of the electric vehicle, which may include, for example, instructions for any application or method operating on the safety precaution method device 800 during the charging of the electric vehicle, as well as application-related data, such as contact data, messages, pictures, audio, video, and the like. The Memory 802 may be implemented by any type or combination of volatile or non-volatile Memory devices, such as static random access Memory (Static Random Access Memory, SRAM for short), electrically erasable programmable Read-Only Memory (Electrically Erasable Programmable Read-Only Memory, EEPROM for short), erasable programmable Read-Only Memory (Erasable Programmable Read-Only Memory, EPROM for short), programmable Read-Only Memory (Programmable Read-Only Memory, PROM for short), read-Only Memory (ROM for short), magnetic Memory, flash Memory, magnetic disk, or optical disk. The multimedia component 803 may include a screen and an audio component. Wherein the screen may be, for example, a touch screen, the audio component being for outputting and/or inputting audio signals. For example, the audio component may include a microphone for receiving external audio signals. The received audio signals may be further stored in the memory 802 or transmitted through the communication component 805. The audio assembly further comprises at least one speaker for outputting audio signals. The I/O interface 804 provides an interface between the processor 801 and other interface modules, which may be a keyboard, mouse, buttons, etc. These buttons may be virtual buttons or physical buttons. The communication component 805 is configured to perform wired or wireless communication between the safety precaution method device 800 and other devices in the charging process of the electric automobile. Wireless communication, such as Wi-Fi, bluetooth, near field communication (Near FieldCommunication, NFC for short), 2G, 3G or 4G, or a combination of one or more thereof, the respective communication component 805 may thus comprise: wi-Fi module, bluetooth module, NFC module.
In an exemplary embodiment, the apparatus 800 for safety precaution method during charging of an electric vehicle may be implemented by one or more application specific integrated circuits (Application Specific Integrated Circuit, abbreviated as ASIC), digital signal processor (DigitalSignal Processor, abbreviated as DSP), digital signal processing apparatus (Digital Signal Processing Device, abbreviated as DSPD), programmable logic device (Programmable Logic Device, abbreviated as PLD), field programmable gate array (Field Programmable Gate Array, abbreviated as FPGA), controller, microcontroller, microprocessor or other electronic component for executing the safety precaution method during charging of an electric vehicle.
In another exemplary embodiment, a computer storage medium is also provided that includes program instructions that, when executed by a processor, implement the steps of the safety precaution method described above in the charging of an electric vehicle. For example, the computer storage medium may be the memory 802 including the program instructions described above, which may be executed by the processor 801 of the safety precaution method device 800 during charging of an electric vehicle to perform the safety precaution method during charging of an electric vehicle described above.
Example 4:
corresponding to the above method embodiment, a storage medium is further provided in this embodiment, and a storage medium described below and a safety pre-warning method described above in an electric vehicle charging process may be referred to correspondingly.
A storage medium, on which a computer program is stored, which when executed by a processor implements the steps of the safety precaution method in the electric vehicle charging process of the method embodiment described above.
The storage medium may be a usb disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (Random Access Memory, RAM), a magnetic disk, or an optical disk, etc. that can store various program codes.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, and various modifications and variations of the present invention will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
The foregoing is merely illustrative of the present invention, and the present invention is not limited thereto, and any person skilled in the art can easily think about variations or substitutions within the technical scope of the present invention, and the scope of the present invention is intended to be covered by the present invention. Therefore, the protection scope of the invention is subject to the protection scope of the claims.
Claims (10)
1. The safety early warning method in the charging process of the electric automobile is characterized by comprising the following steps of:
acquiring historical data and real-time data of charging of the electric vehicle based on license plate codes of the electric vehicle respectively, and preprocessing the historical data and the real-time data respectively to obtain first data and second data, wherein the first data and the second data are physical quantities capable of representing faults in the charging process;
extracting an evaluation index based on the first data, and respectively determining a corresponding weight value and membership;
constructing a health grade state predicted value by using a weighted average principle based on the first data, the weight value and the membership;
training by generating an countermeasure network model based on the first data and the health grade state predicted value to obtain an operation state predicted model;
and inputting the second data into the running state prediction model to obtain a charging running state early warning grade.
2. The method for pre-warning safety during charging of an electric vehicle according to claim 1, wherein the method for pre-processing comprises:
marking according to data types based on the historical data or the real-time data respectively to obtain type data IDs;
modifying based on the time sequence and the type data ID to obtain a characteristic sequence under the same dimension at different moments;
removing abnormal values based on all the characteristic sequences respectively, and filling missing values to obtain complete data;
performing discrete processing based on the complete data to obtain discrete data;
and carrying out normalization processing based on the discrete data to correspondingly obtain first data or second data.
3. The method of claim 2, wherein performing discrete processing based on the complete data to obtain discrete data comprises:
acquiring the complete data by using a standard time interval to form a time sequence vector;
based on the time sequence vector, changing through a time factor to obtain different standard time sequence vectors;
and constructing a composite vector based on the different standard time sequence vectors to obtain discrete data.
4. The method of claim 1, wherein the extracting the evaluation index based on the first data and determining the corresponding weight value and membership degree respectively comprises:
constructing an operation state evaluation system by using a analytic hierarchy process based on the first data;
constructing an index judgment matrix based on the running state evaluation system;
respectively calculating a feature vector and a maximum feature value based on the index judgment matrix;
calculating a consistency ratio based on the maximum eigenvalue, determining a weight value according to the eigenvector if the consistency ratio is smaller than 0.1, otherwise, reassigning the weight value;
determining a sample domain based on the first data and determining a set of operating states based on the sample domain; the running state set is a running data range corresponding to different health levels;
and respectively counting based on the sample domain and the running state set to obtain membership degree, wherein the membership degree is a limit value of the occurrence frequency of the running state set in the sample domain.
5. The method of claim 1, wherein training with generating an countermeasure network model based on the first data and the health level state prediction value to obtain an operational state prediction model comprises:
based on the first data and the health grade state predicted value, respectively carrying out alternate training on a generator model and a discriminator model for generating an countermeasure network model, wherein one training cycle process of the alternate training is as follows: taking the first data as an input value of the generator model to obtain generated data; and then taking the generated data and the health grade state predicted value as input values of the discriminator model, carrying out weight updating by utilizing a loss value and a residual value fed back by the discriminator model, minimizing the residual value and the loss value, and obtaining an operation state predicted model, wherein the residual value is a Wasserstein distance between the health grade state predicted value and the generated data.
6. Safety precaution device in electric automobile charging process, its characterized in that includes:
the acquisition module is used for: the method comprises the steps of respectively obtaining historical data and real-time data of charging of the electric vehicle based on license plate codes of the electric vehicle, respectively preprocessing the historical data and the real-time data to correspondingly obtain first data and second data, wherein the first data and the second data are physical quantities capable of representing faults in the charging process;
and an extraction module: the method comprises the steps of extracting an evaluation index based on first data, and respectively determining a corresponding weight value and membership;
the construction module comprises: the health grade state prediction value is constructed by using a weighted average principle based on the first data, the weight value and the membership;
training module: the system is used for training by generating an countermeasure network model based on the first data and the health grade state predicted value to obtain an operation state predicted model;
and the early warning module is used for: and the second data is input into the running state prediction model to obtain a charging running state early warning grade.
7. The device for early warning of in-process charging of an electric vehicle of claim 6, wherein the acquisition module comprises:
a marking unit: the method comprises the steps of marking according to data types based on the historical data or the real-time data respectively to obtain type data IDs;
a modification unit: the method is used for modifying based on the time sequence and the type data ID to obtain a characteristic sequence under the same dimension at different moments;
a shim cell: the method comprises the steps of removing abnormal values based on all the characteristic sequences respectively, filling missing values, and obtaining complete data;
discrete unit: the discrete processing is used for carrying out discrete processing based on the complete data to obtain discrete data;
pretreatment unit: and the normalization processing is performed based on the discrete data, so that first data or second data are correspondingly obtained.
8. The in-process safety precaution device of claim 7, wherein the discrete unit comprises:
the acquisition unit: the system is used for acquiring the complete data by using standard time intervals to form a time sequence vector;
a conversion unit: the time sequence vector is used for obtaining different standard time sequence vectors based on the time sequence vector through the change of the time factor;
multidimensional unit: and the method is used for constructing a composite vector based on the different standard time sequence vectors to obtain discrete data.
9. The in-process safety precaution device of claim 6, wherein the extraction module comprises:
an evaluation unit: the method comprises the steps of constructing an operation state evaluation system based on the first data by using a analytic hierarchy process;
a judging unit: the method comprises the steps of constructing an index judgment matrix based on the running state evaluation system;
a calculation unit: the method comprises the steps of respectively calculating a feature vector and a maximum feature value based on the index judgment matrix;
a judging unit: the method comprises the steps of calculating a consistency ratio based on the maximum eigenvalue, determining a weight value according to the eigenvector if the consistency ratio is smaller than 0.1, otherwise, reassigning the weight value;
a first statistical unit: for determining a sample domain based on the first data and a set of operating states based on the sample domain; the running state set is a running data range corresponding to different health levels;
a second statistical unit: and the system is used for respectively carrying out statistics based on the sample domain and the running state set to obtain membership degree, wherein the membership degree is a limit value of the occurrence frequency of the running state set in the sample domain.
10. The in-process safety precaution device of claim 9, wherein the training module comprises:
a learning unit: and the training cycle process for alternately training the generator model and the discriminator model for generating the countermeasure network model based on the first data and the health grade state predicted value is as follows: taking the first data as an input value of the generator model to obtain generated data; and then taking the generated data and the health grade state predicted value as input values of the discriminator model, carrying out weight updating by utilizing a loss value and a residual value fed back by the discriminator model, minimizing the residual value and the loss value, and obtaining an operation state predicted model, wherein the residual value is a Wasserstein distance between the health grade state predicted value and the generated data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310085672.6A CN116061690A (en) | 2023-02-02 | 2023-02-02 | Safety early warning method and device in electric automobile charging process |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310085672.6A CN116061690A (en) | 2023-02-02 | 2023-02-02 | Safety early warning method and device in electric automobile charging process |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116061690A true CN116061690A (en) | 2023-05-05 |
Family
ID=86171244
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310085672.6A Pending CN116061690A (en) | 2023-02-02 | 2023-02-02 | Safety early warning method and device in electric automobile charging process |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116061690A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116882981A (en) * | 2023-09-07 | 2023-10-13 | 深圳市海雷新能源有限公司 | Intelligent battery management system based on data analysis |
-
2023
- 2023-02-02 CN CN202310085672.6A patent/CN116061690A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116882981A (en) * | 2023-09-07 | 2023-10-13 | 深圳市海雷新能源有限公司 | Intelligent battery management system based on data analysis |
| CN116882981B (en) * | 2023-09-07 | 2023-11-21 | 深圳市海雷新能源有限公司 | Intelligent battery management system based on data analysis |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107797067B (en) | Lithium ion battery life migration prediction method based on deep learning | |
| Qi et al. | A hybrid method for improved stability prediction in construction projects: A case study of stope hangingwall stability | |
| CN102185735B (en) | Network security situation prediction method | |
| CN113343633A (en) | Thermal runaway fault classification and risk prediction method and system for power lithium battery | |
| CN109604186A (en) | Power battery performance flexibility assesses method for separating | |
| CN113807570B (en) | XGBoost-based reservoir dam risk level assessment method and system | |
| CN113762329A (en) | Method and system for constructing state prediction model of large rolling mill | |
| CN113156917A (en) | Power grid equipment fault diagnosis method and system based on artificial intelligence | |
| CN115469227B (en) | Aggregation variation self-encoder and dynamic regular lithium battery abnormality detection method | |
| CN116737510B (en) | Data analysis-based intelligent keyboard monitoring method and system | |
| CN112434390A (en) | PCA-LSTM bearing residual life prediction method based on multi-layer grid search | |
| CN115508711B (en) | Electric vehicle power lithium battery residual life prediction method based on deep network model | |
| CN113760670A (en) | Cable joint abnormity early warning method and device, electronic equipment and storage medium | |
| CN116061690A (en) | Safety early warning method and device in electric automobile charging process | |
| CN104217296A (en) | Listed company performance comprehensive evaluation method | |
| CN114460481A (en) | Energy storage battery thermal runaway early warning method based on Bi-LSTM and attention mechanism | |
| CN112836967A (en) | New energy automobile battery safety risk assessment system | |
| CN116901707A (en) | Power battery pack fault early warning method, system and vehicle | |
| CN114943281A (en) | Intelligent decision-making method and system for heat pipe cooling reactor | |
| CN114398347A (en) | Fuel cell data analysis system and electronic device based on data correlation | |
| CN111221704A (en) | Method and system for determining operation state of office management application system | |
| Jiang et al. | Prediction on the remaining useful life of rolling bearings using ensemble DLSTM | |
| CN115616437B (en) | Intelligent debugging control method and system for EPC special power supply | |
| CN112485676B (en) | Battery energy storage system state estimation early warning method under digital mirror image | |
| Grindheim et al. | Comparative Study of Machine Learning Methods for State of Health Estimation of Maritime Battery Systems |
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 |