CN109214424B - Method for predicting charging time of new energy automobile by using regression analysis and clustering method - Google Patents
Method for predicting charging time of new energy automobile by using regression analysis and clustering method Download PDFInfo
- Publication number
- CN109214424B CN109214424B CN201810874257.8A CN201810874257A CN109214424B CN 109214424 B CN109214424 B CN 109214424B CN 201810874257 A CN201810874257 A CN 201810874257A CN 109214424 B CN109214424 B CN 109214424B
- Authority
- CN
- China
- Prior art keywords
- charging
- clustering
- curve
- initial
- time
- 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
Images
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
- G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/3644—Constructional arrangements
- G01R31/3648—Constructional arrangements comprising digital calculation means, e.g. for performing an algorithm
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
Abstract
The invention provides a method for predicting charging time of a new energy automobile by using a regression analysis and clustering method, and belongs to the technical field of computer application. The method comprises the steps of fitting historical data by using a regression analysis method, clustering historical charging curves by using a k-means clustering algorithm, and predicting the charging time of the new energy automobile through the historical charging curves. With the development of new energy vehicles, the prediction of the battery charging time is very important, and the method can be used for predicting the battery charging time by mining the charging pile data, so that the user experience of the new energy vehicles can be greatly improved.
Description
Technical Field
The invention belongs to the field of machine learning, and relates to a method for predicting charging time of a new energy automobile by using a regression analysis and clustering method.
Background
In modern life, people can not leave a convenient vehicle, namely an automobile, so that a safe and comfortable travel environment is provided for people, and the life of people is greatly improved. However, the new energy automobile brings many problems while providing convenience, as the number of automobiles increases, spacious roads become crowded, traffic pressure becomes higher and higher, and meanwhile, the automobiles continuously discharge harmful gases such as oxynitride, carbon monoxide and hydrocarbon, so that environmental pollution is further aggravated.
The new energy automobile is rapidly developing as a new industry, and the new energy automobile gradually becomes a new direction for automobile industry innovation due to the characteristics of energy conservation, environmental protection, carbon reduction and the like. In recent years, new energy automobile technology in China is in the forefront of the world and has entered a rapid development stage, so that research in the field of new energy automobiles becomes a popular direction. In the process of popularization, there are some problems hindering the development of new energy automobiles, such as: fill electric pile problem, car battery problem etc. The method mainly solves the problem of charging time of the new energy automobile, and after a user starts to charge the automobile, the time of charging the new energy automobile to different electric quantities is predicted through a regression analysis and clustering method, so that better service is provided for the user, and better development of the field of the new energy automobile is promoted.
Fig. 1 is a charging curve for one of the vehicles with SOC (battery charge percentage) on the abscissa and time (unit: seconds) on the ordinate. The reason why the curve is stepped is that the whole number is obtained when the SOC data are obtained, so the same SOC may appear at different moments, and the curve is stepped. By analyzing the charging curve, the SOC of a vehicle is basically linear along with the change of the time when the vehicle starts to be charged, and when a certain percentage is charged, the time-SOC curve bends due to the protection mechanism of battery management software, and the charging power is reduced.
Disclosure of Invention
The invention provides a method for predicting the charging time of a new energy automobile by using a regression analysis and clustering method through analyzing the characteristics of curves, and aims to provide time prediction for a user when the electric quantity of the new energy automobile reaches a certain percentage so as to meet different requirements of the user.
The technical scheme of the invention is as follows:
a method for predicting charging time of a new energy automobile by using a regression analysis and clustering method comprises the following steps:
step one, learning historical data
(1) Complementing historical data
(1.1) collecting historical charging data of the same type of charging pile, requiring that the initial charging percentage in the historical charging data is between 0% and 80%, and drawing a charging curve for the historical charging data; and carrying out correlation analysis on the charging time and the charging percentage by using a Pearson correlation coefficient, wherein the Pearson correlation coefficient is defined as:
wherein X is a charging time, Y is a charging amount, Cov (X, Y) is a covariance of X and Y, Var (X) is a variance of X, and Var (Y) is a variance of Y; the more the absolute value of the pearson correlation coefficient r (X, Y) is close to 1, the more linear the charging time and the charging percentage are; when the charging amount is below 80%, the charging time and the charging percentage are in a linear relation by analyzing the Pearson correlation coefficient;
and (1.2) reversely complementing the part of each drawn charging curve below 80% by utilizing linear regression analysis to obtain all complete historical charging curves.
(2) Clustering complete historical charging curves
For the same type of charging pile, the complete historical charging curves show different initial slopes, the complete historical charging curves close to the initial slopes have clustering phenomena, all the complete historical charging curves are clustered by using a k-means clustering algorithm, and the initial slopes refer to the slopes of the complete historical charging curves in the beginning straight line part; the k-means clustering algorithm is as follows:
(2.1) extracting the initial slope of each complete historical charging curve;
(2.2) randomly extracting initial slopes of K complete historical charging curves, and taking the initial slopes as initial centroids of each cluster, wherein the K initial centroids are total;
(2.3) respectively calculating the distance between the initial slope of each complete historical charging curve and the initial slope of the extracted K complete historical charging curves;
(2.4) clustering each complete historical charging curve to a class with the minimum distance to form K clusters;
(2.5) recalculating centroids of the K clusters, and using the centroids as new clustering standards; respectively calculating the distance between the initial slope of each complete historical charging curve and the centroids of the K clusters, clustering each complete historical charging curve to the class with the minimum distance, calculating the centroids of the K clusters again, calculating the distances, clustering again, repeating the steps until the centroids calculated in the previous and subsequent steps are not changed, stopping calculation, and finishing clustering;
(2.6) giving n (n > ═ 2) different values to K, and repeating the steps (2.2) to (2.5) to obtain n different clustering results, wherein each clustering result comprises K clusters;
(3) and (3) calculating the error square sum of the n clustering results obtained in the step (2.6) by using the error square sum as an objective function:
wherein K represents K clusters in a certain clustering result, ciAnd representing the centroid of the ith cluster in the clustering result, dist represents the distance, and x is the initial slope of any complete historical charging curve in the ith cluster in the clustering result.
And taking the different values of K as an abscissa and the SSE obtained by calculation as an ordinate to draw a curve. As the value of K increases, the calculated SSE becomes smaller and smaller, so that the initial part of the obtained curve is a monotonously decreasing curve. However, as the value of K continues to increase, the improvement in SSE does not continue to decrease significantly, i.e., the rate of decrease in SSE slows. According to the elbow rule, taking the value of K at the corresponding point of the elbow of the curve as the final clustering number of the complete historical charging curve, and carrying out final clustering on the complete historical charging curve according to the clustering number;
(4) fitting the curved part of the complete historical charging curve obtained in the step (1.2) by using a quadratic polynomial regression analysis method, wherein the specific process is as follows:
(4.1) converting the unary quadratic polynomial into a binary first order polynomial, wherein the unary quadratic polynomial is as follows:
Y=aX2+bX+c
wherein X is the charging time and Y is the charging amount, converting the unary quadratic polynomial into a binary quadratic polynomial, i.e. making X2=X1,X=X2Obtaining:
Y=aX1+bX2+c
then, coefficients a, b and c are determined by using a least square method.
(4.2) obtaining a fitting curve expression of each complete historical charging curve by using the principle of a least square method and after completion and polynomial regression operation;
(5) and (4) taking m points with SOC between 0% and 100% for each obtained fitting curve, calculating time at the m points, and then calculating the average value of all curves in each cluster at the same SOC according to the final clustering result in the step (3) to obtain K charging curves.
Step two, prediction of actual charging time
(1) Collecting data points
And in the process of charging one vehicle, reading the battery charging percentage SOC of the new energy vehicle at intervals, and recording the moment.
(2) Calculating slope, determining grouping
In the process of charging the new energy automobile, after a new data point is obtained each time, the new data point and the previous data points are used for calculating the slope, and when the slope calculated at the present time and later does not change any more, the classification of the charging curve of the new energy automobile which is being charged is determined according to the slope at the present time.
(3) Determining time of different charge percentages
And (5) predicting the charging time required by the new energy automobile to reach the required SOC by using the historical charging curve obtained in the step one (5) according to the classification of the new energy automobile under charging.
The invention has the beneficial effects that: the method is different from the conventional method for predicting the charging time of the battery, the charging time of the new energy automobile is predicted by using the historical data of the charging pile, the internal mechanism of the battery does not need to be deeply researched, the data such as the current, the voltage, the ohmic internal resistance and the like of the battery are also measured, and the data is easy to obtain and high in accuracy.
Drawings
Fig. 1 corresponds to a charging curve corresponding to one of the vehicles;
FIG. 2 is a flow chart of a specific implementation process;
FIG. 3 corresponds to a flowchart of historical data learning;
fig. 4 corresponds to a flowchart of the charging time prediction.
Detailed description of the invention
The specific method can be divided into two parts of historical data learning and charging time prediction, such as fig. 2.
Step one, learning historical data, as shown in FIG. 3
(1) Complementing historical data
(1.1) collecting historical charging data of the same type of charging pile, requiring that the initial charging percentage in the historical charging data is between 0% and 80%, and drawing a charging curve for the historical charging data; performing correlation analysis on the charging time and the charging percentage by using a Pearson correlation coefficient, and obtaining that the charging time and the charging percentage are in a linear relation when the charging amount is below 80% by analyzing the Pearson correlation coefficient;
and (1.2) reversely complementing the part of each drawn charging curve below 80% by utilizing linear regression analysis to obtain all complete historical charging curves.
(2) Clustering complete historical charging curves
For the same type of charging pile, the complete historical charging curves show different initial slopes, the complete historical charging curves close to the initial slopes have clustering phenomena, all the complete historical charging curves are clustered by using a k-means clustering algorithm, and the initial slopes refer to the slopes of the complete historical charging curves in the beginning straight line part; the k-means clustering algorithm is as follows:
(2.1) extracting the initial slope of each complete historical charging curve;
(2.2) randomly extracting the initial slope of 5 complete historical charging curves, and taking the initial slope as the initial centroid of each cluster, wherein the total number of the initial centroids is 5;
(2.3) respectively calculating the distance between the initial slope of each complete historical charging curve and the initial slope of the extracted 5 complete historical charging curves;
(2.4) clustering each complete historical charging curve to a class with the minimum distance to form 5 clusters;
(2.5) recalculating centroids of the 5 clusters and using the centroids as new clustering criteria; respectively calculating the distance between the initial slope of each complete historical charging curve and the centroids of the 5 clusters, clustering each complete historical charging curve to the class with the minimum distance, calculating the centroids of the 5 clusters again, calculating the distances, clustering again, repeating the steps until the centroids calculated in the previous and subsequent steps are not changed, stopping calculation, and finishing clustering;
(2.6) assigning K to n (where n is 5,6, …,25) different values, and repeating the steps (2.2) to (2.5) to obtain n different clustering results, wherein each clustering result contains K clusters;
(3) and (3) calculating the error square sum of the n (n is 5,6, …,25) kinds of clustering results obtained in the step (2.6) by using the error square sum as an objective function, and drawing a curve by taking the different values of K as an abscissa and the SSE obtained by calculation as an ordinate. As the value of K increases, the calculated SSE becomes smaller and smaller, so that the initial part of the obtained curve is a monotonously decreasing curve. However, as the value of K continues to increase, the improvement in SSE does not continue to decrease significantly, i.e., the rate of decrease in SSE slows. According to the elbow rule, taking the value of K at the corresponding point of the elbow of the curve as the final clustering number of the complete historical charging curve, and carrying out final clustering on the complete historical charging curve according to the clustering number;
(4) and (3) fitting the bent part of the complete historical charging curve obtained in the step (1.2) by utilizing a quadratic polynomial regression analysis method, and determining coefficients a, b and c.
(4.2) obtaining a fitting curve expression of each complete historical charging curve by using the principle of a least square method and after completion and polynomial regression operation;
(5) and (4) taking 101 points with SOC between 0% and 100% for each obtained fitted curve, calculating the time at the 101 points, and then calculating the average value of all curves in each cluster at the same SOC according to the final clustering result in the step (3) to obtain K charging curves.
Step two, prediction of actual charging time, as shown in FIG. 4
(1) Collecting data points
And in the process of charging one vehicle, reading the battery charging percentage SOC of the new energy vehicle at intervals, and recording the moment.
(2) Calculating slope, determining grouping
In the process of charging the new energy automobile, after a new data point is obtained each time, the new data point and the previous data points are used for calculating the slope, and when the slope calculated at the present time and later does not change any more, the classification of the charging curve of the new energy automobile which is being charged is determined according to the slope at the present time.
(3) Determining time of different charge percentages
And (5) predicting the charging time required by the new energy automobile to reach the required SOC by using the historical charging curve obtained in the step one (5) according to the classification of the new energy automobile under charging.
In the process of charging a new energy automobile, the charging percentage of a battery at the moment is extracted at intervals, the time at the moment is recorded, regression analysis is carried out according to historical data at the moment, and the slope of the battery in a straight line part is calculated until the value of the slope is basically unchanged. The calculated slope value is used to judge which group the curve can be divided into, and then the learned curve according to the group of historical data can be used to predict the electric quantity of the automobile at which moment. After each vehicle is charged, the charging condition of the vehicle can be learned.
Claims (1)
1. A method for predicting the charging time of a new energy automobile by using a regression analysis and clustering method is characterized by comprising the following steps:
step one, learning historical data
(1) Complementing historical data
(1.1) collecting historical charging data of the same type of charging pile, requiring that the initial charging percentage in the historical charging data is between 0% and 80%, and drawing a charging curve for the historical charging data; and carrying out correlation analysis on the charging time and the charging percentage by using a Pearson correlation coefficient, wherein the Pearson correlation coefficient is defined as:
wherein X is the charge time, Y is the charge percentage, Cov (X, Y) is the covariance of X and Y, Var (X) is the variance of X, Var (Y) is the variance of Y; the more the absolute value of the pearson correlation coefficient r (X, Y) is close to 1, the more linear the charging time and the charging percentage are; when the charging percentage is below 80%, the charging time and the charging percentage are in a linear relation by analyzing the Pearson correlation coefficient;
(1.2) performing reverse completion on the part, below 80 percent of charging percentage, of each drawn charging curve by utilizing linear regression analysis to obtain all complete historical charging curves;
(2) clustering complete historical charging curves
For the same type of charging pile, the complete historical charging curves show different initial slopes, the complete historical charging curves close to the initial slopes have clustering phenomena, all the complete historical charging curves are clustered by using a k-means clustering algorithm, and the initial slopes refer to the slopes of the complete historical charging curves in the beginning straight line part; the k-means clustering algorithm is as follows:
(2.1) extracting the initial slope of each complete historical charging curve;
(2.2) randomly extracting initial slopes of K complete historical charging curves, and taking the initial slopes as initial centroids of each cluster, wherein the K initial centroids are total;
(2.3) respectively calculating the distance between the initial slope of each complete historical charging curve and the initial slope of the extracted K complete historical charging curves;
(2.4) clustering each complete historical charging curve to a class with the minimum distance to form K clusters;
(2.5) recalculating centroids of the K clusters, and using the centroids as new clustering standards; respectively calculating the distance between the initial slope of each complete historical charging curve and the centroids of the K clusters, clustering each complete historical charging curve to the class with the minimum distance, calculating the centroids of the K clusters again, calculating the distances, clustering again, repeating the steps until the centroids calculated in the two previous and next times are not changed, stopping calculation, and finishing clustering;
(2.6) endowing K with n different values, wherein n is more than or equal to 2, and repeating the steps (2.2) to (2.5) to obtain n different clustering results, wherein each clustering result comprises K clusters;
(3) and (3) calculating the error square sum of the n clustering results obtained in the step (2.6) by using the error square sum as an objective function:
wherein K represents K clusters in a certain clustering result, ciRepresenting the centroid of the ith cluster in the clustering result, dist represents the distance, and x is the initial slope of any complete historical charging curve in the ith cluster in the clustering result;
taking different values of K as abscissa and SSE obtained by calculation as ordinate, and drawing a curve; the SSE obtained by calculation is smaller and smaller with the increase of the K value, so that the initial part of the obtained curve is a monotonously decreasing curve; however, as the value of K continues to increase, the improvement in SSE does not continue to decrease significantly, i.e., the rate of decrease in SSE slows; according to the elbow rule, taking the value of K at the corresponding point of the elbow of the curve as the final clustering number of the complete historical charging curve, and carrying out final clustering on the complete historical charging curve according to the clustering number;
(4) fitting the curved part of the complete historical charging curve obtained in the step (1.2) by using a quadratic polynomial regression analysis method, wherein the specific process is as follows:
(4.1) converting the unary quadratic polynomial into a binary first order polynomial, wherein the unary quadratic polynomial is as follows:
Y=aX2+bX+c
wherein X is the charging time and Y is the charging percentage, converting a unary quadratic polynomial into a binary quadratic polynomial, i.e. let X be2=X1,X=X2Obtaining:
Y=aX1+bX2+c
then determining coefficients a, b and c by using a least square method;
(4.2) obtaining a fitting curve expression of each complete historical charging curve by using the principle of a least square method and after completion and polynomial regression operation;
(5) taking m points with SOC between 0% and 100% for each obtained fitting curve, calculating time at the m points, and then calculating the average value of all curves in each cluster at the same SOC according to the final clustering result in the step (3) to obtain K historical charging curves;
step two, prediction of actual charging time
(1) Collecting data points
In the process of charging one vehicle, reading the battery charging percentage SOC of the new energy vehicle at intervals, and recording the moment;
(2) calculating slope, determining grouping
In the process of charging the new energy automobile, after a new data point is obtained every time, calculating a slope by using the new data point and the previous data points, and determining the classification of a charging curve of the new energy automobile which is being charged according to the slope when the slope calculated at the present time and the slope calculated at the later time are not changed;
(3) determining time of different charge percentages
And (5) predicting the charging time required by the new energy automobile to reach the required SOC by using the historical charging curve obtained in the step one (5) according to the classification of the new energy automobile under charging.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810874257.8A CN109214424B (en) | 2018-08-03 | 2018-08-03 | Method for predicting charging time of new energy automobile by using regression analysis and clustering method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810874257.8A CN109214424B (en) | 2018-08-03 | 2018-08-03 | Method for predicting charging time of new energy automobile by using regression analysis and clustering method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109214424A CN109214424A (en) | 2019-01-15 |
CN109214424B true CN109214424B (en) | 2020-02-18 |
Family
ID=64988518
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810874257.8A Active CN109214424B (en) | 2018-08-03 | 2018-08-03 | Method for predicting charging time of new energy automobile by using regression analysis and clustering method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109214424B (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110674835B (en) * | 2019-03-22 | 2021-03-16 | 集美大学 | Terahertz imaging method and system and nondestructive testing method and system |
CN112131606A (en) * | 2020-09-24 | 2020-12-25 | 合肥城市云数据中心股份有限公司 | Dynamic data difference privacy histogram publishing method based on K-means + + combined elbow method autonomous clustering technology |
CN113942415B (en) * | 2021-12-06 | 2023-05-05 | 张志刚 | Big data-based electric vehicle charging pile power resource distribution method and system |
CN116662629A (en) * | 2023-08-02 | 2023-08-29 | 杭州宇谷科技股份有限公司 | Charging curve retrieval method, system, device and medium based on time sequence clustering |
CN117406007B (en) * | 2023-12-14 | 2024-02-13 | 山东佰运科技发展有限公司 | Charging pile charging data detection method and system |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105353312A (en) * | 2015-09-28 | 2016-02-24 | 华晨汽车集团控股有限公司 | Prediction method of power battery SOC |
CN105631483A (en) * | 2016-03-08 | 2016-06-01 | 国家电网公司 | Method and device for predicting short-term power load |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20170286838A1 (en) * | 2016-03-29 | 2017-10-05 | International Business Machines Corporation | Predicting solar power generation using semi-supervised learning |
CN106156895A (en) * | 2016-07-29 | 2016-11-23 | 国网山东省电力公司经济技术研究院 | A kind of charging electric vehicle load forecasting method based on fuzzy C-means clustering with substep grid search support vector regression |
CN107145986B (en) * | 2017-05-24 | 2021-02-05 | 北京中电普华信息技术有限公司 | Charging electric quantity prediction method and device |
-
2018
- 2018-08-03 CN CN201810874257.8A patent/CN109214424B/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN105353312A (en) * | 2015-09-28 | 2016-02-24 | 华晨汽车集团控股有限公司 | Prediction method of power battery SOC |
CN105631483A (en) * | 2016-03-08 | 2016-06-01 | 国家电网公司 | Method and device for predicting short-term power load |
Also Published As
Publication number | Publication date |
---|---|
CN109214424A (en) | 2019-01-15 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109214424B (en) | Method for predicting charging time of new energy automobile by using regression analysis and clustering method | |
CN107909179B (en) | Method for constructing prediction model of running condition of plug-in hybrid vehicle and vehicle energy management method | |
CN109693576B (en) | Electric vehicle charging scheduling optimization method based on simulated annealing algorithm | |
CN111459995B (en) | Multi-mode vehicle speed prediction method based on driving data | |
CN112327168A (en) | XGboost-based electric vehicle battery consumption prediction method | |
CN111301426A (en) | Method for predicting energy consumption in future driving process based on GRU network model | |
CN115759462A (en) | Charging behavior prediction method and device for electric vehicle user and electronic equipment | |
Wei et al. | Online estimation of driving range for battery electric vehicles based on SOC-segmented actual driving cycle | |
CN111783943A (en) | Driver braking strength prediction method based on LSTM neural network | |
CN114954022A (en) | Electric automobile cloud cooperative control device and method | |
CN111967194B (en) | Battery classification method based on cloud historical data | |
CN111325402A (en) | Method for predicting charging behavior of electric vehicle user based on BP neural network | |
CN112287980A (en) | Power battery screening method based on typical feature vector | |
CN112305441A (en) | Power battery health state assessment method under integrated clustering | |
CN115936205A (en) | Vehicle charging planning method for parking lot | |
CN112036602A (en) | 5G electric vehicle charging prediction method and system integrating human-computer intelligence | |
Lin et al. | Deep Q-learning network based trip pattern adaptive battery longevity-conscious strategy of plug-in fuel cell hybrid electric vehicle | |
CN109408955B (en) | Energy consumption analysis method and system for electric automobile | |
CN113376540B (en) | LSTM battery health state estimation method based on evolutionary attention mechanism | |
CN114646888A (en) | Assessment method and system for capacity attenuation of power battery | |
CN113183827A (en) | New forms of energy electric automobile intelligent cloud central control management platform that traveles based on artificial intelligence | |
CN111126449A (en) | Battery fault classification diagnosis method based on cluster analysis | |
CN112070300A (en) | Electric vehicle charging platform selection method based on multi-objective optimization | |
CN111976543A (en) | Charging and discharging control method for battery pack of electric automobile | |
Gao et al. | Control strategy for PHEB based on actual driving cycle with driving style characteristic |
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 | ||
GR01 | Patent grant |