CN112560994B - Time sequence-based vehicle working condition classification method and device - Google Patents
Time sequence-based vehicle working condition classification method and device Download PDFInfo
- Publication number
- CN112560994B CN112560994B CN202011569211.9A CN202011569211A CN112560994B CN 112560994 B CN112560994 B CN 112560994B CN 202011569211 A CN202011569211 A CN 202011569211A CN 112560994 B CN112560994 B CN 112560994B
- Authority
- CN
- China
- Prior art keywords
- time period
- target
- working condition
- vehicle
- vehicle working
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 47
- 239000013598 vector Substances 0.000 claims abstract description 111
- 230000006870 function Effects 0.000 claims description 56
- 238000005315 distribution function Methods 0.000 claims description 29
- 230000009467 reduction Effects 0.000 claims description 9
- 238000004590 computer program Methods 0.000 claims description 8
- 238000010276 construction Methods 0.000 claims description 6
- 238000000605 extraction Methods 0.000 claims description 4
- 238000004364 calculation method Methods 0.000 claims description 3
- 230000008569 process Effects 0.000 description 11
- 238000004422 calculation algorithm Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 230000001133 acceleration Effects 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000005520 cutting process Methods 0.000 description 2
- 238000012216 screening Methods 0.000 description 2
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000005265 energy consumption Methods 0.000 description 1
- 230000007774 longterm Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011176 pooling Methods 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000001052 transient effect Effects 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
- G06F18/232—Non-hierarchical techniques
- G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/08—Detecting or categorising vehicles
Abstract
The application discloses a time sequence-based vehicle working condition classification method and device, wherein the method comprises the following steps: acquiring a clustering result characteristic of the vehicle working condition in the ith target time period; i is an integer from 1 to n; the n target time periods are arranged in time sequence to form an object time period. And obtaining the statistical vehicle working condition characteristics of the object time period. And splicing the clustering result features of the vehicle working conditions in the n target time periods according to the time sequence to obtain the input features of the object time periods. And splicing the input characteristics of the object time period and the statistical vehicle working condition characteristics of the object time period to obtain first input data. And inputting the first input data into the target model to obtain a first to-be-classified vehicle working condition vector representation of the object time period. Clustering the first to-be-classified vehicle working condition vector representation to obtain a vehicle working condition classification result in the object time period. Based on the unsupervised target model and the clustering, the automatic classification of the vehicle working conditions in the object time period is realized.
Description
Technical Field
The application relates to the technical field of vehicles, in particular to a time sequence-based vehicle working condition classification method and device.
Background
During vehicle driving, a large number of vehicle conditions may occur. The vehicle working condition classification result obtained by classifying the vehicle working conditions can be applied to the aspects of vehicle working condition prediction, a vehicle alarm system, vehicle energy consumption analysis, user driving behavior analysis and the like.
At present, classification of vehicle working conditions is performed manually. Because the existing vehicle working condition is complex, the method for classifying the vehicle working condition by means of manual calibration consumes a great deal of labor cost and time cost, and has low accuracy. In addition, at present, the classification of the vehicle working conditions is mostly carried out aiming at the vehicle working conditions at a single moment, the vehicle working conditions in a specific time period cannot be classified, and the representation and classification of the vehicle working conditions cannot be completed based on the change rule of the vehicle working conditions in the specific time period.
Disclosure of Invention
In order to solve the technical problems, the application provides a time sequence-based vehicle working condition classification method and device, which are used for automatically completing the classification of vehicle working conditions in a time period in an unsupervised mode and improving the accuracy of vehicle working condition classification.
In order to achieve the above object, the technical solution provided by the embodiments of the present application is as follows:
The embodiment of the application provides a time sequence-based vehicle working condition classification method, which comprises the following steps:
acquiring a clustering result characteristic of the vehicle working condition in the ith target time period; i is an integer from 1 to n; the n is the number of target time periods; the n target time periods are arranged according to time sequence to form an object time period;
Acquiring statistical vehicle-like working condition characteristics of the object time period;
Splicing the clustering result features of the vehicle working conditions in the n target time periods according to the time sequence to obtain the input features of the object time periods;
The input characteristics of the object time period and the statistical vehicle working condition characteristics of the object time period are spliced to obtain first input data;
Inputting the first input data into a target model, and carrying out vector re-expression on the first input data to obtain a first vehicle working condition vector expression to be classified in the object time period;
Clustering the first to-be-classified vehicle working condition vector representation of the object time period to obtain a vehicle working condition classification result in the object time period.
Optionally, the acquiring the clustering result feature of the vehicle working condition in the ith target time period includes:
acquiring a preselected data set of a jth target moment corresponding to an ith target time period; j is an integer from 1 to m; the m target moments form an ith target time period according to the time sequence;
processing the preselected data set of the jth target moment to obtain the vehicle working condition characteristics of the jth target moment;
Splicing the vehicle working condition characteristics of m target moments according to the time sequence to obtain second input data;
Inputting the second input data into a target model, and carrying out vector re-expression on the second input data to obtain a second vehicle working condition vector expression to be classified in the ith target time period;
Clustering the second to-be-classified vehicle working condition vector representation of the ith target time period to obtain the vehicle working condition clustering result characteristic of the ith target time period.
Optionally, the clustering the first to-be-classified vehicle condition vector representation of the object time period to obtain a vehicle condition classification result in the object time period includes:
constructing a probability distribution function and an auxiliary target distribution function based on the mass center;
calculating a first error function through the centroid-based probability distribution function and an auxiliary target distribution function;
acquiring a second error function of the target model;
constructing a joint loss function through the first error function and the second error function;
Updating a clustering center by utilizing the joint loss function;
and clustering the first to-be-classified vehicle working condition vector representation of the object time period through updating of the clustering center to obtain a vehicle working condition classification result in the object time period.
Optionally, the method further comprises:
And when the clustering center is updated by utilizing the joint loss function, synchronously updating model parameters of the target model by utilizing the joint loss function.
Optionally, the inputting the first input data into the target model, and performing vector re-expression on the first input data to obtain a first to-be-classified vehicle condition vector expression of the object time period, including:
Inputting the first input data into a target model; the first input data includes a lateral feature;
extracting the transverse features through the target model to obtain first features;
performing dimension reduction on the first feature through the target model to obtain a second feature; the second feature comprises a longitudinal timing feature;
Extracting the longitudinal time sequence features through the target model; a first to-be-classified vehicle condition vector representation of the subject time period is obtained.
The embodiment of the application also provides a vehicle working condition classification device based on time sequence, which comprises:
the first acquisition unit is used for acquiring the clustering result characteristics of the vehicle working conditions in the ith target time period; i is an integer from 1 to n; the n is the number of target time periods; the n target time periods are arranged according to time sequence to form an object time period;
The second acquisition unit is used for acquiring the statistical vehicle working condition characteristics of the object time period;
The first splicing unit is used for splicing the vehicle working condition clustering result features of the n target time periods according to the time sequence to obtain the input features of the object time periods; the second splicing unit is used for splicing the input characteristics of the object time period and the statistical vehicle working condition characteristics of the object time period to obtain first input data; ;
The third acquisition unit is used for inputting the first input data into a target model, carrying out vector re-expression on the first input data, and obtaining a first vehicle working condition vector expression to be classified in the object time period;
and the clustering unit is used for clustering the first vehicle working condition vector representation to be classified in the object time period to obtain a vehicle working condition classification result in the object time period.
Optionally, the first obtaining unit includes:
A first obtaining subunit, configured to obtain a preselected data set of a jth target time corresponding to an ith target time period; j is an integer from 1 to m; the m target moments form an ith target time period according to the time sequence;
The processing subunit is used for processing the preselected data set of the jth target moment to obtain the vehicle working condition characteristics of the jth target moment;
the splicing subunit is used for splicing the vehicle working condition characteristics of the m target moments according to the time sequence to obtain second input data;
The second obtaining subunit is used for inputting the second input data into a target model, and carrying out vector re-expression on the second input data to obtain a second vehicle working condition vector expression to be classified in the ith target time period;
And the clustering subunit is used for clustering the second to-be-classified vehicle working condition vector representation of the ith target time period to obtain the vehicle working condition clustering result characteristic of the ith target time period.
Optionally, the clustering unit includes:
A first construction subunit for constructing a centroid-based probability distribution function and an auxiliary target distribution function;
a calculation subunit for calculating a first error function from the centroid-based probability distribution function and the auxiliary target distribution function;
a third obtaining subunit, configured to obtain a second error function of the target model;
A second construction subunit for constructing a joint loss function from the first error function and the second error function;
A first updating subunit, configured to update a clustering center with the joint loss function;
And the fourth acquisition subunit is used for clustering the first vehicle working condition vector representation to be classified in the object time period through updating of the clustering center to obtain a vehicle working condition classification result in the object time period.
The embodiment of the application also provides a time sequence-based vehicle working condition classification device, which comprises: the vehicle working condition classification method based on time sequence is realized when the processor executes the computer program.
The embodiment of the application also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is used for executing the time sequence-based vehicle working condition classification method.
According to the technical scheme, the application has the following beneficial effects:
The embodiment of the application provides a time sequence-based vehicle working condition classification method, which comprises the following steps: acquiring a clustering result characteristic of the vehicle working condition in the ith target time period; i is an integer from 1 to n; n is the number of target time periods. Wherein, n target time periods are arranged according to time sequence to form an object time period. And obtaining the statistical vehicle working condition characteristics of the object time period. And splicing the clustering result features of the vehicle working conditions in the n target time periods according to the time sequence to obtain the input features of the object time periods. And splicing the input characteristics of the object time period and the statistical vehicle working condition characteristics of the object time period to obtain first input data. And inputting the first input data into the target model, and carrying out vector re-expression on the first input data to obtain a first to-be-classified vehicle working condition vector expression of the object time period. Clustering the first to-be-classified vehicle working condition vector representation of the object time period to obtain a vehicle working condition classification result in the object time period. According to the application, the time sequence-based vehicle working condition characteristics in the object time period are automatically classified through the unsupervised target model and the clustering algorithm, so that the automatic classification of the vehicle working conditions in the object time period is realized, and the accuracy of the classification of the vehicle working conditions is improved.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings that are required in the embodiments or the description of the prior art will be briefly described, and it is obvious that the drawings in the following description are some embodiments of the present application, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a flow chart of a method for classifying vehicle conditions based on time sequence according to an embodiment of the present application;
FIG. 2 is a flowchart for obtaining clustering result features of vehicle conditions in a target time period according to an embodiment of the present application;
FIG. 3 is a schematic diagram of a model for obtaining clustering result characteristics of vehicle working conditions in a target time period according to an embodiment of the present application;
Fig. 4 is a schematic diagram of a vehicle condition classification device based on time sequence according to an embodiment of the present application.
Detailed Description
In order that the above-recited objects, features and advantages of the present application will become more readily apparent, a more particular description of embodiments of the application will be rendered by reference to the appended drawings and appended drawings.
Referring to fig. 1, fig. 1 is a flowchart of a method for classifying vehicle conditions based on time sequence according to an embodiment of the present application. As shown in fig. 1, the method includes S101-S106:
S101: acquiring a clustering result characteristic of the vehicle working condition in the ith target time period; i is an integer from 1 to n; n is the number of target time periods; the n target time periods are arranged in time sequence to form an object time period.
Acquiring a clustering result characteristic of the vehicle working condition in the ith target time period, wherein the value of i is an integer from 1 to n; n is the number of target time periods. The vehicle working condition clustering result features of n target time periods are obtained, wherein the vehicle working condition clustering result features are represented by vectors.
The n target time periods are arranged in time series to form the target time period. It can be appreciated that the time sequence-based vehicle condition classification method provided by the embodiment of the application can realize the purpose of automatically classifying the vehicle conditions in the object time period. Specifically, the object time period and the target time period may be selected according to actual situations, and the object time period and the target time period are not limited herein. For example, the n target time periods are one month, i.e., the object time period is one month, and the target time period is one day. On this basis, the ith target period is each day corresponding to one month.
It should also be noted that real-time vehicle operating condition data is obtained from the battery pack data set. The vehicle working condition characteristics of the ith time period can be obtained based on the real-time vehicle working condition data of the ith target time period, and the vehicle working condition clustering result characteristics of the ith target time period can be obtained through the vehicle working condition characteristics of the ith time period. The specific process of acquiring the clustering result characteristic of the vehicle working condition in the ith target time period through the vehicle working condition characteristic in the ith time period can be referred to in the following embodiments, and will not be described in detail herein.
The vehicle working condition characteristics can be obtained by simply screening and processing the vehicle working condition data in the battery pack data set. The vehicle operating mode characteristics include at least: temperature fed back by a vehicle charging pile, vehicle driving time, whether a vehicle is stopped, vehicle speed, vehicle acceleration, battery current, battery voltage, battery charging state, battery cell temperature, battery cell average temperature, highest temperature, lowest temperature, battery cell maximum voltage, battery cell minimum voltage, battery cell temperature difference and battery cell pressure difference. It should be noted that the vehicle operating condition features are specifically timing-based vehicle operating condition features.
The time sequence-based vehicle operating condition characteristics may be categorized according to power battery pack characteristics and real-time driving characteristics. Wherein the power battery pack features include at least: temperature, battery current, battery voltage, battery state of charge, cell differential pressure and cell differential temperature fed back by the vehicle charging pile. The real-time driving characteristics include at least: vehicle speed, vehicle acceleration, and whether the vehicle is stationary.
It should be noted that the vehicle operating condition features described above are non-statistical vehicle operating condition features.
S102: and obtaining the statistical vehicle working condition characteristics of the object time period.
The statistical vehicle working condition characteristics of the object time period can be obtained by simply screening and processing the vehicle working condition data in the battery pack data set. For example, when the object time period includes n target time periods, the object time period is one day, the statistical vehicle-like operating condition characteristics in one day are obtained, and the statistical vehicle-like operating condition characteristics in the n target time periods are obtained in total. It will be appreciated that the subject time period may also be one month.
The statistical vehicle working condition characteristics of the object time period at least comprise: the method comprises the steps of vehicle standing time length, vehicle running time length, charging state values before and after the vehicle standing time, temperature recording before and after the vehicle standing, and total record number of low-voltage running time length recording and sampling. It can be understood that the statistical vehicle-like working condition characteristics are mostly real-time driving characteristics of the vehicle, and the statistical vehicle-like working condition characteristics are represented by vectors.
S103: and splicing the clustering result features of the vehicle working conditions in the n target time periods according to the time sequence to obtain the input features of the object time periods.
And splicing the clustering result features of the vehicle working conditions in the n target time periods according to the time sequence to obtain the input features of the object time periods. For example, the object time period is one day, the target time period is 5min, and n is determined to be 60 according to the use time in one day. The vehicle condition clustering result features of the 60 target time periods are spliced according to time sequences, for example, the vehicle condition clustering result features of the 60 target time periods in one day are spliced according to time sequences, for example, 12: and (3) splicing the characteristics of the clustering result of the vehicle working conditions in a ratio of 05-12:10 at 12:00-12:05, 12: and the clustering result features of the vehicle working conditions are spliced in 12: and (5) after the clustering result features of the vehicle working conditions are 05-12:10, until 60 target time periods are spliced, and finally, the spliced vector is used as the input feature of the target time period. It can be understood that the clustering result features of the vehicle working conditions in the n target time periods are spliced according to the time sequence.
S104: and splicing the input characteristics of the object time period and the statistical vehicle working condition characteristics of the object time period to obtain first input data.
And splicing the input characteristics of the object time period and the statistical vehicle working condition characteristics of the object time period to obtain first input data. The first input data is represented by a vector, and the first input data is the first input data in the object time period. For example, the input features in one day and the statistical vehicle working condition features in one day are spliced to obtain first input data corresponding to one day.
In S105, the sequence length required to be input by the target model is consistent, so when the input feature of the object time period and the statistical vehicle operating condition feature of the object time period are spliced, a filling and cutting operation is performed to ensure that the feature sequences in the obtained first input data are the same length.
S105: and inputting the first input data into the target model, and carrying out vector re-expression on the first input data to obtain a first to-be-classified vehicle working condition vector expression of the object time period.
And inputting the obtained first input data in the object time period into a target model, carrying out vector re-expression on the first input data, and finally inputting the first vehicle working condition vector expression to be classified in the object time period.
The target model is used for carrying out vector re-representation on the input first input data, specifically, the target model is used for carrying out dimension reduction on the first input data in the object time period to obtain an expected vector, and the dimension of the expected vector can be selected according to actual conditions. It is understood that the expected vector is a first to-be-classified vehicle condition vector representation of the subject time period.
It should be noted that the target model may be selected according to actual needs, and the target model is not limited herein, and is a self-encoder as an example. As another example, the object model is a variant encoder.
Specifically, inputting the first input data into the target model, and performing vector re-expression on the first input data to obtain a first to-be-classified vehicle working condition vector expression of the object time period, including:
Inputting the first input data into the target model; the first input data includes a lateral feature;
Extracting transverse features through a target model to obtain first features;
Performing dimension reduction on the first feature through the target model to obtain a second feature; the second feature comprises a longitudinal timing feature;
Extracting longitudinal time sequence features through a target model; a first to-be-classified vehicle condition vector representation of the subject time period is obtained.
It can be appreciated that the first input data is represented by a vector, and the lateral quantity includes different vehicle operating condition characteristics and statistical vehicle operating condition characteristics, and the lateral quantity is a lateral characteristic; the longitudinal quantity characterizes the vehicle working condition characteristics and counts the time sequence change condition of the vehicle working condition characteristics, and the longitudinal quantity is the longitudinal time sequence characteristics.
And extracting the transverse characteristics by using the target model to obtain first characteristics. Specifically, the transverse features capture short-distance fluctuation and transverse rules among the features through convolution to obtain first features. And then reducing the dimension of the first feature to obtain a second feature. And then enabling the second feature to enter a time sequence model, extracting longitudinal time sequence features in the second feature, namely compactly representing data while retaining a time sequence structure, and obtaining a first vehicle working condition vector representation to be classified in the object time period. It can be appreciated that the vector re-representation of the vehicle conditions is achieved by the target model based on the law of variation of the vehicle conditions for a specific period of time.
It should be noted that, taking the self-encoder as an example, the first vehicle condition vector to be classified of the object time period output by the target model is expressed as the output of the encoder in the self-encoder, which is the intermediate output of the self-encoder, that is, the step does not involve the decoder in the self-encoder.
It can be understood that the first input data is input into the target model, the expected vector is obtained by extracting the network structure in the target model, and the output expected vector is the result of fusing the time sequence feature, the vehicle working condition feature and the statistical vehicle working condition feature.
S106: clustering the first to-be-classified vehicle working condition vector representation of the object time period to obtain a vehicle working condition classification result in the object time period.
After the first to-be-classified vehicle working condition vector representation of the object time period is obtained, clustering the first to-be-classified vehicle working condition vector representation of the object time period to obtain a vehicle working condition classification result in the object time period.
When the experimental object is a vehicle, the vehicle working condition data in the object time period is real-time vehicle working condition data in the object time period of the vehicle, and the vehicle working condition classification result of the vehicle is obtained through the target model and the clustering. When the experimental object is a preset number of vehicles, the real-time vehicle working condition data in the object time period is the vehicle working condition data in the object time period of the preset number of vehicles, and the vehicle working condition classification result of the preset number of vehicles in the object time period is obtained through the target model and clustering.
In specific implementation, clustering the first to-be-classified vehicle working condition vector representation of the object time period to obtain a vehicle working condition classification result in the object time period, including:
constructing a probability distribution function and an auxiliary target distribution function based on the mass center;
calculating a first error function by a centroid-based probability distribution function and an auxiliary target distribution function;
Acquiring a second error function of the target model;
constructing a joint loss function through the first error function and the second error function;
Updating the clustering center by using the joint loss function;
Clustering the first vehicle working condition vector representation to be classified in the object time period through updating of the clustering center, and obtaining a vehicle working condition classification result in the object time period.
The first working condition vector representation of the vehicle to be classified is represented by a plurality of working condition vectors, a probability distribution function based on mass centers is used for calculating the similarity between each working condition vector and a clustering center, and the similarity is represented by probability distribution and is not a fixed value. In particular, the centroid-based probability distribution function may use t-distribution, i.e., the similarity between the operating condition vector and the cluster center is calculated using t-distribution. It should be noted that, the initial cluster center is selected randomly, and the cluster center is the centroid.
It can be appreciated that the auxiliary target distribution function has two limiting conditions, one of which is high in reliability; and secondly, the joint loss function obtained by using the auxiliary target distribution function can be successfully used for updating the model parameters of the clustering center and the target model. And selecting an auxiliary target distribution function in the clustering process according to the limiting condition.
It should be noted that, a first error function based on the probability distribution of the centroid and the auxiliary target distribution is calculated, and a second error function is calculated based on the first to-be-classified vehicle condition vector representation output by the target model and the first input data. The joint loss function is obtained by weighted summation of a first error function and a second error function of the target model in a clustering algorithm. In some embodiments, the first error function is KL divergence and the second error function is a mean square error of the target model. It will be appreciated that the use of a joint loss function may control the convergence between the centroid-based probability distribution and the auxiliary target distribution. When the cluster center is updated with the joint loss function, the joint loss function is also used for synchronously updating model parameters of the target model.
It should be further noted that, in the clustering process, the number of the clustering categories is selected according to the actual situation, for example, the clustering categories are selected as 10 categories or 20 categories. Clustering the first vehicle working condition vector representation to be classified in the object time period through updating of the clustering center, and obtaining a vehicle working condition classification result in the object time period. For example, the clustering class is selected to be 10 classes, after the updating of the clustering center is finished, the probability that a certain working condition vector of a certain vehicle corresponds to 10 clustering classes is selected, and the class with the largest probability value is used as the final classification result of the working condition vector. Based on this, all the vehicle condition classification results within the object time period can be obtained.
It can be understood that the target model in the application is an unsupervised model, and the clustering process is an unsupervised process. The vehicle working condition classification method based on time sequence provided by the embodiment of the application is an unsupervised classification method.
After the vehicle working condition classification result is obtained in the object time period, the vehicle working condition classification result can be visually represented, the vehicle working condition classification result is given an interpretability, and the vehicle working condition classification result has practical significance. If can be used for distinguishing whether the vehicle is running at high speed most of the time, judge whether the vehicle is the commercial vehicle running long distance, can also check whether other vehicles have hidden danger. Visual representation of the clustering results, for example: the blue color is in a category 1 and represents a daily on-off vehicle (the corresponding vehicle working condition characteristics are that the use frequency is average, the travel time is fixed, the vehicle running speed is in an average value, the vehicle is often slowly charged at night, and the like). Yellow is category 2, indicating that vehicle habits are often transient and fast charging (the corresponding vehicle operating condition is characterized by a voltage curve that varies with a large slope over a period of time and is a long-term phenomenon). The red color is category 3, and indicates that the vehicle is often kept still for a long time, and gradually experiences pressure difference when in use (the vehicle is easy to cause larger faults, and is warned in actual use). It can be understood that the category corresponding to the final cluster includes time series characteristics, power battery pack characteristics and the working conditions of the summary expression after partial characteristics of the real-time driving characteristics.
The method provided by the embodiment of the application comprises the following steps: acquiring a clustering result characteristic of the vehicle working condition in the ith target time period; i is an integer from 1 to n; n is the number of target time periods. And obtaining the statistical vehicle working condition characteristics of the object time period. And splicing the clustering result features of the vehicle working conditions in the n target time periods according to the time sequence to obtain the input features of the object time periods. And splicing the input characteristics of the object time period and the statistical vehicle working condition characteristics of the object time period to obtain first input data. And inputting the first input data into the target model, and carrying out vector re-expression on the first input data to obtain a first to-be-classified vehicle working condition vector expression of the object time period. The object time periods are composed of n target time periods according to time sequence arrangement. Clustering the first to-be-classified vehicle working condition vector representation of the object time period to obtain a vehicle working condition classification result in the object time period. According to the embodiment of the application, the time sequence-based vehicle working condition characteristics in the object time period are automatically classified through the unsupervised target model and the clustering algorithm, so that the efficient automatic classification of the vehicle working conditions in the object time period is realized, and the accuracy of the classification of the vehicle working conditions is improved.
Referring to fig. 2, fig. 2 is a flowchart of obtaining a clustering result feature of vehicle working conditions in a target period according to an embodiment of the present application. As shown in fig. 2, obtaining a clustering result feature of the vehicle working condition in the ith target time period includes:
S201: acquiring a preselected data set of a jth target moment corresponding to an ith target time period; j is an integer from 1 to m; the m target moments form an ith target time period according to the time sequence.
Acquiring a preselected data set of a jth target moment corresponding to an ith target time period; j is an integer from 1 to m; the m target moments form an ith target time period according to the time sequence. For example, the i-th target period is 12:00-12:15, wherein the selected time dimension is 1s, then m target moments are 300 1s, and 300 1s form an ith target time period 12:00-12:15, wherein m can be selected according to practical conditions.
It will be appreciated that the preselected data set is real-time vehicle operating condition data acquired in the battery pack data set. The preselected data sets for the m target moments constitute real-time vehicle operating condition data for the ith target time period.
S202: and processing the preselected data set of the jth target moment to obtain the vehicle working condition characteristics of the jth target moment.
And processing the preselected data set of the jth target moment to obtain the vehicle working condition characteristics of the jth target moment. The value of j is an integer from 1 to m, and the vehicle working condition characteristics of m target moments form the vehicle working condition characteristics of the ith target time period according to time sequence.
Referring to fig. 3, fig. 3 is a schematic diagram of a model for obtaining clustering result characteristics of vehicle working conditions in a target time period according to an embodiment of the present application. Those skilled in the art will appreciate that the schematic diagram shown in fig. 3 is merely one example in which embodiments of the present application may be implemented, and that the scope of applicability of embodiments of the application is not limited in any way by the framework. As shown in fig. 3, as an example, the target time is 1s, and the vehicle condition characteristic in each second at the target time represents the vehicle condition characteristic in 1s, that is, the vehicle condition characteristic at any one target time, for example, may represent the vehicle condition characteristic at the jth target time.
S203: and splicing the vehicle working condition characteristics at m target moments according to the time sequence to obtain second input data.
After the preselected data set of the jth target moment is obtained, the value of j is an integer from 1 to m, and the vehicle working condition characteristics of m target moments can be obtained. And splicing the vehicle working condition characteristics at m target moments according to the time sequence to obtain second input data.
Because the target model in S105 requires that the input sequence length is consistent, when the vehicle condition features at m target moments are spliced according to the time sequence, the input features are subjected to the filling cutting operation, and the vehicle condition features are cut off or filled to the same length.
S204: and inputting the second input data into the target model, and carrying out vector re-expression on the second input data to obtain a second vehicle working condition vector expression to be classified in the ith target time period.
After the second input data is obtained, the second input data is input into the target model, the second input data is subjected to vector re-expression, and the second vehicle working condition vector expression to be classified in the ith target time period is obtained.
The target model is used for carrying out vector re-representation on the input second input data, and specifically, the target model is used for carrying out dimension reduction on the second input data in the ith target time period to obtain an expected vector, and the dimension of the expected vector can be selected according to actual conditions. It is understood that the expected vector is a second to-be-classified vehicle condition vector representation of the ith target period.
The detailed process of inputting the second input data into the target model, and performing vector re-expression on the second input data to obtain the second to-be-classified vehicle condition vector expression of the ith target time period may refer to the previous embodiment, and will not be described in detail herein. As an example, referring to fig. 3, with the encoder in the self-encoder, first, the first feature is obtained by capturing the transverse law between the short-distance fluctuation and the feature for the convolution that first passes through the convolutional neural network, and then the second feature is obtained by dimension reduction. And then enabling the second feature to enter two long-short-period memory networks, extracting longitudinal time sequence features in the second feature, namely, compactly representing data while retaining a time sequence structure, and obtaining a second vehicle working condition vector representation to be classified in the ith target time period. The second vehicle working condition vector to be classified in the ith target time period is composed of vehicle working condition characteristics to be classified in m target moments. As an example, the dimension reduction is max pooling.
It should be noted that, referring to fig. 3, taking the self-encoder as an example, the second vehicle condition vector to be classified of the ith target period output by the target model is expressed as the output of the encoder in the self-encoder, and is the intermediate output of the self-encoder, and this step does not involve the decoder in the self-encoder.
It can be understood that the second input data is input into the target model, and the expected vector is obtained by extracting the network structure in the target model, and the output expected vector is a result of fusing the time sequence feature and the vehicle working condition feature, that is, the output expected vector is a result of fusing the time sequence feature, the power battery pack feature, the real-time driving feature and the like.
S205: clustering is carried out based on the second to-be-classified vehicle working condition vector representation of the ith target time period, and vehicle working condition clustering result characteristics of the ith target time period are obtained.
After the second vehicle working condition vector representation to be classified in the ith target time period is obtained, clustering is carried out based on the second vehicle working condition vector representation to be classified in the ith target time period, and vehicle working condition clustering result characteristics in the ith target time period are obtained.
It should be noted that, the detailed feature of the clustering result of the vehicle condition for obtaining the ith target period may refer to the previous embodiment, which is not described in detail herein. In the clustering process, referring to fig. 3, as an example, model parameters of the target model are parameters of a convolutional neural network and a two-way long-short-term memory network in a model structure.
According to the process for acquiring the clustering result characteristics of the vehicle working conditions in the target time periods, which is provided by the embodiment of the application, the clustering result characteristics of the vehicle working conditions in each target time period are acquired by dividing each target time period into a plurality of target moments and utilizing a target model and a clustering algorithm. Based on the method, the vehicle condition clustering result features of the target time periods, namely the vehicle condition clustering result features of the object time periods and the statistical vehicle condition features of the object time periods, are utilized, and the unsupervised target model and the clustering algorithm are combined to obtain the vehicle condition classification result of the object time periods. By the method, the automatic classification of the vehicle working conditions in the object time period is realized, and the accuracy of the classification of the vehicle working conditions is improved.
Referring to fig. 4, fig. 4 is a schematic diagram of a time-sequence-based vehicle condition classification device according to an embodiment of the present application, where the device 400 includes:
a first obtaining unit 401, configured to obtain a clustering result feature of the vehicle condition in the ith target period; i is an integer from 1 to n; the n is the number of target time periods; the n target time periods are arranged according to time sequence to form an object time period;
A second obtaining unit 402, configured to obtain a statistical vehicle-like operating condition characteristic of the object time period;
A first splicing unit 403, configured to splice the vehicle condition clustering result features of the n target time periods according to a time sequence, so as to obtain input features of the object time periods; a second stitching unit 404, configured to stitch the input feature of the object time period and the statistical vehicle operating mode feature of the object time period to obtain first input data;
A third obtaining unit 405, configured to input the first input data into a target model, and perform vector re-representation on the first input data, so as to obtain a first vehicle working condition vector representation to be classified in the object time period;
And the clustering unit 406 is configured to cluster the first to-be-classified vehicle condition vector representation of the object time period, and obtain a vehicle condition classification result in the object time period.
Optionally, in some implementations of the embodiments of the present application, the first obtaining unit 401 includes:
A first obtaining subunit, configured to obtain a preselected data set of a jth target time corresponding to an ith target time period; j is an integer from 1 to m; the m target moments form an ith target time period according to the time sequence;
The processing subunit is used for processing the preselected data set of the jth target moment to obtain the vehicle working condition characteristics of the jth target moment;
the splicing subunit is used for splicing the vehicle working condition characteristics of the m target moments according to the time sequence to obtain second input data;
The second obtaining subunit is used for inputting the second input data into a target model, and carrying out vector re-expression on the second input data to obtain a second vehicle working condition vector expression to be classified in the ith target time period;
And the clustering subunit is used for clustering the second to-be-classified vehicle working condition vector representation of the ith target time period to obtain the vehicle working condition clustering result characteristic of the ith target time period.
Optionally, in some implementations of the embodiments of the present application, the clustering unit 406 includes:
A first construction subunit for constructing a centroid-based probability distribution function and an auxiliary target distribution function;
a calculation subunit for calculating a first error function from the centroid-based probability distribution function and the auxiliary target distribution function;
a third obtaining subunit, configured to obtain a second error function of the target model;
A second construction subunit for constructing a joint loss function from the first error function and the second error function;
A first updating subunit, configured to update a clustering center with the joint loss function;
And the fourth acquisition subunit is used for clustering the first vehicle working condition vector representation to be classified in the object time period through updating of the clustering center to obtain a vehicle working condition classification result in the object time period.
Optionally, in some implementations of the embodiments of the present application, the clustering subunit includes:
a second updating subunit, configured to update a clustering center using the joint loss function;
And a third updating subunit, configured to synchronously update model parameters of the target model with the joint loss function when the cluster center is updated with the joint loss function.
Optionally, in some implementations of the embodiments of the present application, the third obtaining unit 405 includes:
an input subunit for inputting the first input data into a target model; the first input data includes a lateral feature;
The first extraction subunit is used for extracting the transverse characteristics through the target model to obtain first characteristics;
the dimension reduction subunit is used for reducing the dimension of the first feature through the target model to obtain a second feature; the second feature comprises a longitudinal timing feature;
a second extraction subunit, configured to extract the longitudinal time sequence feature through the target model; a first to-be-classified vehicle condition vector representation of the subject time period is obtained.
The embodiment of the application provides a time sequence-based vehicle working condition classification device, which is used for acquiring a vehicle working condition clustering result characteristic of an ith target time period; i is an integer from 1 to n; n is the number of target time periods. And obtaining the statistical vehicle working condition characteristics of the object time period. And splicing the clustering result features of the vehicle working conditions in the n target time periods according to the time sequence to obtain the input features of the object time periods. And splicing the input characteristics of the object time period and the statistical vehicle working condition characteristics of the object time period to obtain first input data. And inputting the first input data into the target model, and carrying out vector re-expression on the first input data to obtain a first to-be-classified vehicle working condition vector expression of the object time period. The object time periods are composed of n target time periods according to time sequence arrangement. Clustering the first to-be-classified vehicle working condition vector representation of the object time period to obtain a vehicle working condition classification result in the object time period. The embodiment of the application provides a time sequence-based vehicle working condition classification device, which classifies time sequence-based vehicle working condition characteristics in an object time period through an unsupervised target model and a clustering algorithm, realizes automatic classification of vehicle working conditions in the object time period, and improves accuracy of vehicle working condition classification.
The embodiment of the application also provides a time sequence-based vehicle working condition classification device, which comprises: the vehicle condition classification method based on time sequence according to the embodiment is realized when the processor executes the computer program.
The embodiment of the application also provides a computer readable storage medium, wherein the computer readable storage medium stores a computer program, and the computer program is used for executing the time sequence-based vehicle working condition classification method according to the embodiment.
It should be noted that, in the present description, each embodiment is described in a progressive manner, and each embodiment is mainly described in a different manner from other embodiments, and identical and similar parts between the embodiments are all enough to refer to each other. For the method disclosed in the embodiment, since it corresponds to the system disclosed in the embodiment, the description is relatively simple, and the relevant points refer to the system part.
It should also be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (9)
1. A method for classifying vehicle conditions based on time sequence, the method comprising:
acquiring a clustering result characteristic of the vehicle working condition in the ith target time period; i is an integer from 1 to n; the n is the number of target time periods; the n target time periods are arranged according to time sequence to form an object time period;
Acquiring statistical vehicle-like working condition characteristics of the object time period;
Splicing the clustering result features of the vehicle working conditions in the n target time periods according to the time sequence to obtain the input features of the object time periods;
The input characteristics of the object time period and the statistical vehicle working condition characteristics of the object time period are spliced to obtain first input data;
Inputting the first input data into a target model, and carrying out vector re-expression on the first input data to obtain a first vehicle working condition vector expression to be classified in the object time period;
Clustering the first to-be-classified vehicle working condition vector representation of the object time period to obtain a vehicle working condition classification result in the object time period;
inputting the first input data into a target model, and performing vector re-expression on the first input data to obtain a first vehicle working condition vector expression to be classified of an object time period, wherein the method comprises the following steps:
Inputting the first input data into a target model; the first input data includes a lateral feature;
extracting the transverse features through the target model to obtain first features;
performing dimension reduction on the first feature through the target model to obtain a second feature; the second feature comprises a longitudinal timing feature;
Extracting the longitudinal time sequence features through the target model; a first to-be-classified vehicle condition vector representation of the subject time period is obtained.
2. The method of claim 1, wherein the obtaining the vehicle condition cluster result feature for the ith target time period comprises:
acquiring a preselected data set of a jth target moment corresponding to an ith target time period; j is an integer from 1 to m; the m target moments form an ith target time period according to the time sequence;
processing the preselected data set of the jth target moment to obtain the vehicle working condition characteristics of the jth target moment;
Splicing the vehicle working condition characteristics of m target moments according to the time sequence to obtain second input data;
Inputting the second input data into a target model, and carrying out vector re-expression on the second input data to obtain a second vehicle working condition vector expression to be classified in the ith target time period;
And clustering based on the second to-be-classified vehicle working condition vector representation of the ith target time period to obtain the vehicle working condition clustering result characteristic of the ith target time period.
3. The method of claim 1, wherein clustering the first to-be-classified vehicle condition vector representation of the object time period to obtain a vehicle condition classification result within the object time period comprises:
constructing a probability distribution function and an auxiliary target distribution function based on the mass center;
calculating a first error function through the centroid-based probability distribution function and an auxiliary target distribution function;
acquiring a second error function of the target model;
constructing a joint loss function through the first error function and the second error function;
Updating a clustering center by utilizing the joint loss function;
and clustering the first to-be-classified vehicle working condition vector representation of the object time period through updating of the clustering center to obtain a vehicle working condition classification result in the object time period.
4. A method according to claim 3, wherein said updating a cluster center with said joint loss function comprises:
Updating a clustering center by utilizing the joint loss function;
And when the clustering center is updated by utilizing the joint loss function, synchronously updating model parameters of the target model by utilizing the joint loss function.
5. A time sequence-based vehicle condition classification device, the device comprising:
the first acquisition unit is used for acquiring the clustering result characteristics of the vehicle working conditions in the ith target time period; i is an integer from 1 to n; the n is the number of target time periods; the n target time periods are arranged according to time sequence to form an object time period;
The second acquisition unit is used for acquiring the statistical vehicle working condition characteristics of the object time period;
The first splicing unit is used for splicing the vehicle working condition clustering result features of the n target time periods according to the time sequence to obtain the input features of the object time periods;
The second splicing unit is used for splicing the input characteristics of the object time period and the statistical vehicle working condition characteristics of the object time period to obtain first input data;
The third acquisition unit is used for inputting the first input data into a target model, carrying out vector re-expression on the first input data, and obtaining a first vehicle working condition vector expression to be classified in the object time period;
the clustering unit is used for clustering the first vehicle working condition vector representation to be classified in the object time period to obtain a vehicle working condition classification result in the object time period;
the third acquisition unit includes:
an input subunit for inputting the first input data into a target model; the first input data includes a lateral feature;
The first extraction subunit is used for extracting the transverse characteristics through the target model to obtain first characteristics;
the dimension reduction subunit is used for reducing the dimension of the first feature through the target model to obtain a second feature; the second feature comprises a longitudinal timing feature;
a second extraction subunit, configured to extract the longitudinal time sequence feature through the target model; a first to-be-classified vehicle condition vector representation of the subject time period is obtained.
6. The apparatus of claim 5, wherein the first acquisition unit comprises:
A first obtaining subunit, configured to obtain a preselected data set of a jth target time corresponding to an ith target time period; j is an integer from 1 to m; the m target moments form an ith target time period according to the time sequence;
The processing subunit is used for processing the preselected data set of the jth target moment to obtain the vehicle working condition characteristics of the jth target moment;
the splicing subunit is used for splicing the vehicle working condition characteristics of the m target moments according to the time sequence to obtain second input data;
The second obtaining subunit is used for inputting the second input data into a target model, and carrying out vector re-expression on the second input data to obtain a second vehicle working condition vector expression to be classified in the ith target time period;
And the clustering subunit is used for clustering the second to-be-classified vehicle working condition vector representation of the ith target time period to obtain the vehicle working condition clustering result characteristic of the ith target time period.
7. The apparatus of claim 5, wherein the clustering unit comprises:
A first construction subunit for constructing a centroid-based probability distribution function and an auxiliary target distribution function;
a calculation subunit for calculating a first error function from the centroid-based probability distribution function and the auxiliary target distribution function;
a third obtaining subunit, configured to obtain a second error function of the target model;
A second construction subunit for constructing a joint loss function from the first error function and the second error function;
A first updating subunit, configured to update a clustering center with the joint loss function;
And the fourth acquisition subunit is used for clustering the first vehicle working condition vector representation to be classified in the object time period through updating of the clustering center to obtain a vehicle working condition classification result in the object time period.
8. A time-series-based vehicle condition classification apparatus, comprising: a memory, a processor, and a computer program stored on the memory and executable on the processor, which when executed, implements the time-series-based vehicle condition classification method of any one of claims 1-4.
9. A computer-readable storage medium, wherein a computer program for executing the time-series-based vehicle condition classification method according to any one of claims 1 to 4 is stored in the computer-readable storage medium.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011569211.9A CN112560994B (en) | 2020-12-26 | 2020-12-26 | Time sequence-based vehicle working condition classification method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011569211.9A CN112560994B (en) | 2020-12-26 | 2020-12-26 | Time sequence-based vehicle working condition classification method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112560994A CN112560994A (en) | 2021-03-26 |
| CN112560994B true CN112560994B (en) | 2024-05-03 |
Family
ID=75033261
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011569211.9A Active CN112560994B (en) | 2020-12-26 | 2020-12-26 | Time sequence-based vehicle working condition classification method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112560994B (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108569297A (en) * | 2017-03-14 | 2018-09-25 | 厦门雅迅网络股份有限公司 | A kind of recognition methods of vehicle driving-cycle and system |
| CN110020623A (en) * | 2019-04-04 | 2019-07-16 | 中山大学 | Physical activity identifying system and method based on condition variation self-encoding encoder |
| CN110348526A (en) * | 2019-07-15 | 2019-10-18 | 武汉绿色网络信息服务有限责任公司 | A kind of device type recognition methods and device based on semi-supervised clustering algorithm |
| CN110562096A (en) * | 2019-08-22 | 2019-12-13 | 东软睿驰汽车技术（沈阳）有限公司 | Remaining mileage prediction method and device |
| CN111476822A (en) * | 2020-04-08 | 2020-07-31 | 浙江大学 | Laser radar target detection and motion tracking method based on scene flow |
| CN111931868A (en) * | 2020-09-24 | 2020-11-13 | 常州微亿智造科技有限公司 | Time series data abnormity detection method and device |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5301310B2 (en) * | 2009-02-17 | 2013-09-25 | 株式会社日立製作所 | Anomaly detection method and anomaly detection system |
| EP3404593A1 (en) * | 2017-05-15 | 2018-11-21 | Tata Consultancy Services Limited | Method and system for data based optimization of performance indicators in process and manufacturing industries |
-
2020
- 2020-12-26 CN CN202011569211.9A patent/CN112560994B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108569297A (en) * | 2017-03-14 | 2018-09-25 | 厦门雅迅网络股份有限公司 | A kind of recognition methods of vehicle driving-cycle and system |
| CN110020623A (en) * | 2019-04-04 | 2019-07-16 | 中山大学 | Physical activity identifying system and method based on condition variation self-encoding encoder |
| CN110348526A (en) * | 2019-07-15 | 2019-10-18 | 武汉绿色网络信息服务有限责任公司 | A kind of device type recognition methods and device based on semi-supervised clustering algorithm |
| CN110562096A (en) * | 2019-08-22 | 2019-12-13 | 东软睿驰汽车技术（沈阳）有限公司 | Remaining mileage prediction method and device |
| CN111476822A (en) * | 2020-04-08 | 2020-07-31 | 浙江大学 | Laser radar target detection and motion tracking method based on scene flow |
| CN111931868A (en) * | 2020-09-24 | 2020-11-13 | 常州微亿智造科技有限公司 | Time series data abnormity detection method and device |
Non-Patent Citations (5)
| Title |
|---|
| Data-Driven Identification of Characteristic Real-Driving Cycles Based on k-Means Clustering and Mixed-Integer Optimization;Daniel Forster等;《IEEE transactions on Vehicular Technology》;20191231;第69卷(第03期);2398-2410 * |
| Deep representation clustering-based fault diagnosis method with unsupervised data applied to rotating machinery;Xiang Li等;《Mechanical Systems and Signal Processing》;20200930;第143卷;1-18 * |
| 基于K-均值聚类算法的行驶工况构建方法;秦大同等;《 吉林大学学报(工学版)》;20150310;第46卷(第02期);383-389 * |
| 基于变分自编码器的带属性网络表示学习与深度嵌入聚类;余平刚;《中国优秀硕士学位论文全文数据库 (信息科技辑)》;20190815(第2019（08）期);I139-17 * |
| 基于大数据技术的货车行驶工况分析;朱庭男;《中国优秀硕士学位论文全文数据库 (信息科技辑)》;20190115(第2019（01）期);I138-2021 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112560994A (en) | 2021-03-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110568359B (en) | Lithium battery residual life prediction method | |
| CN110068774A (en) | Estimation method, device and the storage medium of lithium battery health status | |
| WO2020224724A1 (en) | Server-side characterisation of rechargeable batteries | |
| EP3719708A1 (en) | Model test method and device | |
| CN112858919A (en) | Battery system online fault diagnosis method and system based on cluster analysis | |
| CN110275119A (en) | A kind of cell health state assessment models construction method, appraisal procedure and device | |
| CN112016237B (en) | Deep learning method, device and system for lithium battery life prediction | |
| CN105406532A (en) | Method and device for detecting available energy of electrical vehicle and battery of electrical vehicle | |
| CN112986830A (en) | Lithium battery capacity estimation method based on convolution time memory neural network | |
| CN109768340B (en) | Method and device for estimating voltage inconsistency in battery discharge process | |
| CN116826933B (en) | Knowledge-graph-based hybrid energy storage battery power supply backstepping control method and system | |
| CN109800446B (en) | Method and device for estimating voltage inconsistency in discharging process of lithium ion battery | |
| CN109143074A (en) | A kind of power battery model parameter identification method and system | |
| CN113064089A (en) | Internal resistance detection method, device, medium and system of power battery | |
| CN113608140A (en) | Battery fault diagnosis method and system | |
| EP4031886A1 (en) | Condition value for rechargeable batteries | |
| CN115754726A (en) | Battery life prediction and maintenance method, electronic equipment and storage medium | |
| CN117031310A (en) | Method for predicting residual service life of power battery of electric automobile | |
| CN112560994B (en) | Time sequence-based vehicle working condition classification method and device | |
| CN116774089A (en) | Convolutional neural network battery state of health estimation method and system based on feature fusion | |
| CN114062948B (en) | Lithium ion battery SOC estimation method based on 3DCNN | |
| CN116298944A (en) | Charging remaining time display control method, device, equipment, vehicle and medium | |
| CN114239388A (en) | Method and system for predicting remaining life of battery by fusing test and Internet of vehicles | |
| CN115169630A (en) | Electric vehicle charging load prediction method and device | |
| CN114236415A (en) | Power battery health state detection method, device, equipment and storage medium |
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 | ||
| GR01 | Patent grant |