CN113221843A - Driving style classification method based on empirical mode decomposition characteristics - Google Patents
Driving style classification method based on empirical mode decomposition characteristics Download PDFInfo
- Publication number
- CN113221843A CN113221843A CN202110631751.3A CN202110631751A CN113221843A CN 113221843 A CN113221843 A CN 113221843A CN 202110631751 A CN202110631751 A CN 202110631751A CN 113221843 A CN113221843 A CN 113221843A
- Authority
- CN
- China
- Prior art keywords
- event
- characteristic
- driving style
- method based
- empirical mode
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/59—Context or environment of the image inside of a vehicle, e.g. relating to seat occupancy, driver state or inner lighting conditions
- G06V20/597—Recognising the driver's state or behaviour, e.g. attention or drowsiness
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- 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
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Probability & Statistics with Applications (AREA)
- Multimedia (AREA)
- Control Of Driving Devices And Active Controlling Of Vehicle (AREA)
Abstract
The invention belongs to the technical field of automatic driving, and particularly relates to a driving style classification method based on empirical mode decomposition characteristics, which comprises the following steps of: step S1, selecting a turning event as an event segment for judging the driving style; step S2, collecting data, and extracting turning events according to the steering wheel angle variance; and step S3, performing EMD on the lateral acceleration and speed characteristics in each turning event, performing characteristic parameter calculation on the obtained plurality of I MF, performing PCA dimension reduction on a characteristic matrix formed by the characteristic parameters, and screening the characteristic parameters according to the accumulated contribution rate. According to the invention, by adopting an EMD signal decomposition method, a high-frequency component which can represent the main characteristic energy of an original signal and is large is screened out, and original sample data is enriched. The transverse direction and the longitudinal direction are simultaneously considered during feature signal selection, feature parameters are selected through principal component contribution rate accumulated by PCA, and independence and accuracy of feature parameter extraction are guaranteed.
Description
Technical Field
The invention relates to the technical field of automatic driving, in particular to a driving style classification method based on Empirical Mode Decomposition (EMD) characteristics.
Background
The driving style classification can better understand the hobbies of drivers and related driving problems, can provide powerful support for the development of an intelligent auxiliary system, and is closely related to oil consumption.
At present, most of the research on the driving style is to calculate characteristic parameters for classification based on the whole driving process, most of the used data are simulator data, and the data volume is small and fixed. The EMD decomposition-based driver style identification method provided by the invention improves the inaccuracy of the identification result of the existing driving style identification method to a certain extent and improves the classification reliability. The invention provides a possibility of personalized development for future intelligent auxiliary systems and vehicle-mounted service systems, and meanwhile, a driver can know own driving behaviors more, adjust driving modes, realize oil saving and emission reduction and realize green travel.
Disclosure of Invention
Technical problem to be solved
Aiming at the defects of the prior art, the invention provides a driving style classification method based on empirical mode decomposition characteristics, which solves the problem of inaccuracy of the recognition result of the conventional driving style recognition method and improves the classification reliability.
The second technical proposal.
The invention specifically adopts the following technical scheme for realizing the purpose:
a driving style classification method based on empirical mode decomposition features comprises the following steps:
step S1, selecting a turning event as an event segment for judging the driving style;
step S2, collecting data, and extracting turning events according to the steering wheel angle variance;
step S3, EMD decomposition is carried out on the lateral acceleration and speed characteristics in each turning event, characteristic parameter calculation is carried out on a plurality of obtained intrinsic Mode functions IMF (intrinsic Mode function), Principal Component Analysis (PCA) (principal Component analysis) dimensionality reduction is carried out on a characteristic matrix formed by the characteristic parameters, and the characteristic parameters are screened according to the accumulated contribution rate;
and step S4, calculating the same parameters in the event, carrying out k-means clustering and determining the category center. And (3) extracting an event for the target driver, calculating characteristic parameters, repeating the series of same operations, determining an event center, comparing the distance between the event center and the category center, and determining the style of the event according to the distance. And then calculating the proportion of the style components to determine the final style.
Further, the turn event in step S1 requires the driver to complete deceleration, turn the turn lights on, turn the steering wheel, determine the trajectory of the vehicle, and then accelerate again within a few seconds (typically 10S).
Further, the collecting data in step S2 includes: steering wheel corner, lateral acceleration and speed, and the collected data adopts filtering processing, adopts a mean filtering method:wherein xk( k 1,2, 3.) is the original data point, ykFor the result point obtained by filtering, n is the number of data points (generally 100000).
Further, in step S3, the EMD decomposition algorithm obtains IMF components of the signal at different time characteristic scales by layer-by-layer screening, and the EMD decomposition mainly aims to perform Hilbert transform on the IMF components in order to smooth the signal, and further obtain instantaneous frequency components corresponding to the IMF components.
Further, the EMD decomposition in step S3 specifically includes:
step S31, finding all extreme points of the signal x (t);
step S32, fitting envelope lines emax (t) and emin (t) of upper and lower extreme points by using a 3-time spline curve, obtaining an average value m (t) of the upper and lower envelope lines, and subtracting h (t) from x (t), wherein h (t) is x (t) -m (t);
step S33, judging whether h (t) is IMF or not according to a preset criterion;
step S34, if not, replacing x (t) with h (t), repeating the above steps until h (t) meets the criterion, and then h (t) is the IMF (t) to be extracted.
Further, the k-means cluster in step S4 is: according to the accumulated contribution rate, X characteristic parameters (X is less than or equal to 6) are screened out, characteristic parameter calculation is carried out by taking each turning event as a whole to form characteristic points (P1, P2, …, PN), and N is less than or equal to 6.
The invention identifies the driving style based on the turning event, divides the whole driving process into paragraphs with practical significance, and can ensure that the characteristics of the driver are more complete and the classification accuracy is higher. The prior literature indicates that turning events involve more distinct driving actions and exhibit more distinct personalized features than other events. And moreover, Empirical Mode Decomposition (EMD) is carried out on the signals in the event, the most main component representing the original signals can be screened out, the sample size can be enriched to a certain extent, redundant signals can be eliminated, and the feature extraction accuracy is improved.
(III) advantageous effects
Compared with the prior art, the invention provides a driving style classification method based on empirical mode decomposition characteristics, which has the following beneficial effects:
according to the invention, by adopting an EMD signal decomposition method, a high-frequency component which can represent the main characteristic energy of an original signal and is large is screened out, and original sample data is enriched. The transverse direction and the longitudinal direction are simultaneously considered during feature signal selection, feature parameters are selected through principal component contribution rate accumulated by PCA, and independence and accuracy of feature parameter extraction are guaranteed. In consideration of the characteristic of fluctuation of the driving style, the invention divides the whole driving process into a plurality of meaningful driving event segments, shortens the time length of characteristic parameter calculation, highlights the characteristics of the driver and has more accurate classification result.
Drawings
FIG. 1 is a flow chart of the main body of the present invention;
FIG. 2 is a schematic diagram illustrating an EMD decomposition process of a signal during a turning event according to the present invention;
FIG. 3 is a schematic diagram illustrating the driving style classification and determination according to the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Examples
As shown in fig. 1-3, in the driving style classification method based on the empirical mode decomposition feature according to an embodiment of the present invention, a turning event is selected as an event segment for judging the driving style, data is collected, the turning event is extracted according to the steering wheel angle variance, then the EMD decomposition is performed on the lateral acceleration and speed features in each turning event, feature parameter calculation is performed on a plurality of obtained IMFs, PCA dimension reduction is performed on a feature matrix formed by the feature parameters, the feature parameters are screened according to the accumulated contribution rate, finally the same parameters are calculated in the event, k-means clustering is performed, and the category center is determined. And (3) extracting an event for the target driver, calculating a series of the same operations by using the characteristic parameters, determining an event center, comparing the distance between the event center and the category center, and determining the style of the event according to the distance. And then calculating the proportion of the style components to determine the final style. The method specifically comprises the following steps:
first, selection of driving event
The driving behavior is the control of a steering wheel, a brake pedal and an accelerator by hands and feet of a driver in the driving process, and a series of driving behaviors form a driving event. Examples of events that may be used to perform driving style recognition include braking events, acceleration and deceleration events, lane change events, and turning events. Under urban road conditions, turning events are selected herein for driver style recognition, which requires the driver to complete deceleration, turn the turn lights, turn the steering wheel, determine the vehicle trajectory, and then accelerate again within a few seconds (typically 10 seconds) compared to other events. The series of behaviors can capture more obvious actions of the driver, and the embodied personalized features are more obvious.
Second, driving event extraction
And selecting a driving route, wherein the urban road at least comprises 10 curves, and acquiring real vehicle data. The tested drivers (generally more than 20 drivers) with skilled driving skills are selected, wherein male drivers and female drivers account for 50% of the total number respectively, the average age is 30 years (the upper part and the lower part of the driver have 2 years of floating space respectively), and the average driving age is more than five years. In the experimental process, the tested driver is healthy and can freely run along the tested road section for half an hour, and the collected data comprise the steering wheel rotation angle, the lateral acceleration and the speed. In order to eliminate the interference of data such as vibration, noise and the like, the collected data is filtered, and a mean value filtering method is adopted.
Wherein xk( k 1,2, 3.) is the original data point, ykFor the result point obtained by filtering, n is the number of data points (generally 100000).
A turning event requires that three conditions be met simultaneously: direction change of 70 °, total duration less than 10s, stable direction at least 5s before turning. In order to completely extract the turning event, an initial time window is set to be 10s, the step length is set to be 3s, and the variance of the steering wheel angle in each window is calculated by utilizing a sliding time window. Setting a variance threshold value as m, and preliminarily extracting turning event segments. And then, carrying out time domain analysis on the steering wheel corner signal in the extracted turning event by utilizing Fourier transform, setting a corner threshold value to be n, and detecting whether the signal waveform meets the trend from 0 to n to 0 (absolute equality is not required and the approximation is required). If the requirements are met, the turning event is the turning event, otherwise, the window is slid back and forth by taking 3s as the step length until the requirements are met.
Three, EMD decomposition and feature extraction
The EMD time-frequency analysis method is fundamentally different from the traditional signal time-frequency analysis method as a new method for processing nonlinear non-stationary signals, and has good effect in practical application. The EMD algorithm obtains IMF components of signals with different time characteristic scales through layer-by-layer screening. The EMD decomposition mainly aims to smooth signals, Hilbert transform is carried out on IMF components, instantaneous frequency components corresponding to the IMF components are further obtained, and the obtained instantaneous frequency has reasonable physical significance.
The empirical mode decomposition method comprises the following steps:
(1) finding all the extreme points of the signal x (t);
(2) fitting envelope lines emax (t) and emin (t) of upper and lower extreme points by using a 3-time spline curve, obtaining an average value m (t) of the upper and lower envelope lines, and subtracting h (t) from x (t), wherein h (t) is x (t) -m (t);
(3) judging whether h (t) is IMF or not according to a preset criterion;
(4) if not, replacing x (t) with h (t), repeating the steps until h (t) meets the criterion, and then h (t) is the IMF (t) needing to be extracted;
(5) every time one-order IMF is obtained, deducting it from the original signal, and repeating the above steps; the only remaining part of the signal is a monotonic or constant sequence until the end.
EMD decomposition is often used as a preprocessing stage for signal feature extraction, and can be used for decomposition of any signal in theory, and has obvious advantages in processing non-stationary and non-linear data. The EMD can adaptively perform principal component analysis of the signal, and each decomposed component represents each frequency component in the original signal. The advantages of the EMD analysis method are mainly expressed in the aspects of principal component analysis, self-adaptive time-frequency analysis and signal local transient characteristic characterization.
Since the turning event is an event integrating the lateral and longitudinal driving behaviors, the lateral acceleration and the speed are taken as characteristic signals, and the mean value, the maximum value and the standard deviation of the lateral acceleration and the speed are taken as characteristic parameters. In order to eliminate the correlation among the characteristic parameters and reduce the data dimension, principal component analysis is carried out on all the calculated characteristic parameters by PCA, and a plurality of characteristic parameters with the cumulative contribution rate exceeding z (generally 80%) are screened out for subsequent analysis. The EMD decomposition process is explained by taking the lateral acceleration as an example, the lateral acceleration signal is subjected to EMD decomposition in each turning event segment, a plurality of IMF (eigenmode function) components (the last one is noise in the signal and can be ignored) are obtained through decomposition, and in order to ensure that the sample size is increased and the calculated amount is not excessively increased, the first w IMFs (generally 5) are selected to calculate the characteristic parameters (mean value, maximum value and standard deviation). Event extraction is performed according to the method by using the collected data, and assuming that a (generally 10) turning events are extracted from each driver during driving, a feature matrix with the size of 10a × 3w can be obtained through the decomposition calculation. Similarly, the velocity signal is subjected to EMD decomposition, and the first three IMFs are selected for feature parameter calculation, so that a feature matrix with a size of 10a × 3w is formed. In combination, the final size of the feature matrix is 10a × 6w, which serves as the input for the subsequent PCA principal component analysis. The specific process is shown in figure 2.
Fourthly, classifying driving styles
According to the accumulated contribution rate, X characteristic parameters (X is less than or equal to 6) are screened out, characteristic parameter calculation is carried out by taking each turning event as a whole to form characteristic points (P1, P2, …, PN), and N is less than or equal to 6. And performing k-means clustering on all the feature points, and setting the clustering number to be 2. The clustering result is two cluster graphs in a coordinate system, the larger the transverse acceleration value and the velocity value are, the more the style is aggressive, and the statistical values (mean value, maximum value and variance) of the transverse acceleration value and the velocity value also have the property. Therefore, the images far away from the origin in the result graph represent aggressive styles, and the parts near the origin represent conservative styles, and the cluster centers m1 and m2 of the images are taken as two types of classification criteria, namely, the category centers. And finishing judging the driving style classification standard.
And collecting the steering wheel angle, the lateral acceleration and the vehicle speed of the target driver through a CAN bus. And (3) carrying out turning event extraction on the whole driving process of the target driver, and dividing the whole driving process into a plurality of turning event segments. The driving style reflects a characteristic inherent to the driver, so the driving style of the driver is not invariable. In consideration of the variability of the driving style, the style judging method provided by the invention does not directly judge the whole driving process, but judges the style of each turning event segment of the target driver firstly and then judges the whole style. And calculating corresponding characteristic parameter values in each turning event by using the characteristic parameters screened by the process, wherein each event segment forms an event center, and the event center coordinates consist of all the calculated characteristic parameter values. Calculating the distances d1 and d2 between the event center and the centers m1 and m2 of the two categories, wherein if d1 is less than d2, the event segment is an aggressive style; if d1 > d2, the event fragment is conservative. And finally, calculating the proportion of the two styles, wherein the larger the proportion is, the more the driving style of the target driver has the style tendency. The specific process is shown in figure 3.
According to the driving style classification method based on the empirical mode decomposition characteristics, the road section collection data are set, the steering wheel corner variance is calculated by using the sliding time window, and a plurality of turning event segments are extracted according to the threshold value. Then, considering the transverse and longitudinal driving behaviors, performing EMD signal decomposition on the transverse acceleration and speed characteristics in each turning event segment, performing characteristic parameter calculation (mean value, maximum value and variance) on a plurality of eigenmode Function (IMF) components obtained by decomposition, performing dimensionality reduction Analysis on a formed characteristic matrix by using Principal Component Analysis (PCA), and screening out a plurality of characteristic parameters with the cumulative contribution rate exceeding z (generally 80%). And calculating the same characteristic parameters in each turning event, and then performing K-means clustering to determine a clustering center, namely a category center. And finally, extracting events of the target driver, calculating parameters, determining the event center of each event fragment, and calculating the distance between the event center and the class center, wherein the class belongs to the class with the small distance. And finally, calculating the style component proportion of the event segment, wherein the style with high proportion is the final driving style of the target driver.
Finally, it should be noted that: although the present invention has been described in detail with reference to the foregoing embodiments, it will be apparent to those skilled in the art that changes may be made in the embodiments and/or equivalents thereof without departing from the spirit and scope of the invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (6)
1. A driving style classification method based on empirical mode decomposition features is characterized by comprising the following steps: the method comprises the following steps:
step S1, selecting a turning event as an event segment for judging the driving style;
step S2, collecting data, and extracting turning events according to the steering wheel angle variance;
step S3, EMD decomposition is carried out on the transverse acceleration and speed characteristics in each turning event, characteristic parameter calculation is carried out on the obtained IMFs, PCA dimension reduction is carried out on a characteristic matrix formed by the characteristic parameters, and the characteristic parameters are screened according to the accumulated contribution rate;
and step S4, calculating the same parameters in the event, carrying out k-means clustering and determining the category center. And (3) extracting an event for the target driver, calculating characteristic parameters, repeating the series of same operations, determining an event center, comparing the distance between the event center and the category center, and determining the style of the event according to the distance. And then calculating the proportion of the style components to determine the final style.
2. The driving style classification method based on empirical mode decomposition features according to claim 1, characterized in that: the turn event in step S1 requires the driver to complete deceleration, turn the turn lights on, turn the steering wheel, determine the vehicle trajectory, and then accelerate again within a few seconds (typically 10S).
3. The driving style classification method based on empirical mode decomposition features according to claim 1, characterized in that: collecting data in step S2 includes: steering wheel corner, lateral acceleration and speed, and the collected data adopts filtering processing, adopts a mean filtering method:wherein xk(k 1,2, 3.) is the original data point, ykFor the result point obtained by filtering, n is the number of data points (generally 100000).
4. The driving style classification method based on empirical mode decomposition features according to claim 1, characterized in that: in the step S3, the EMD decomposition algorithm obtains IMF components of the signal at different time characteristic scales by layer-by-layer screening, and the EMD decomposition mainly aims to perform the stabilization processing on the signal, perform Hilbert transform on the IMF components, and further obtain instantaneous frequency components corresponding to the IMF components.
5. The driving style classification method based on empirical mode decomposition features according to claim 1, characterized in that: the EMD decomposition in step S3 specifically includes:
step S31, finding all extreme points of the signal x (t);
step S32, fitting envelope lines emax (t) and emin (t) of upper and lower extreme points by using a 3-time spline curve, obtaining an average value m (t) of the upper and lower envelope lines, and subtracting h (t) from x (t), wherein h (t) is x (t) -m (t);
step S33, judging whether h (t) is IMF or not according to a preset criterion;
step S34, if not, replacing x (t) with h (t), repeating the above steps until h (t) meets the criterion, and then h (t) is the IMF (t) to be extracted;
step S35, when each IMF is obtained, subtracting the IMF from the original signal, and repeating the steps; the only remaining part of the signal is a monotonic or constant sequence until the end.
6. The driving style classification method based on empirical mode decomposition features according to claim 1, characterized in that: the k-means clustering in step S4 is: according to the accumulated contribution rate, X characteristic parameters (X is less than or equal to 6) are screened out, characteristic parameter calculation is carried out by taking each turning event as a whole to form characteristic points (P1, P2, …, PN), and N is less than or equal to 6.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110631751.3A CN113221843A (en) | 2021-06-07 | 2021-06-07 | Driving style classification method based on empirical mode decomposition characteristics |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110631751.3A CN113221843A (en) | 2021-06-07 | 2021-06-07 | Driving style classification method based on empirical mode decomposition characteristics |
Publications (1)
Publication Number | Publication Date |
---|---|
CN113221843A true CN113221843A (en) | 2021-08-06 |
Family
ID=77083138
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110631751.3A Pending CN113221843A (en) | 2021-06-07 | 2021-06-07 | Driving style classification method based on empirical mode decomposition characteristics |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113221843A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113850554A (en) * | 2021-10-15 | 2021-12-28 | 上海吉七物联网科技有限公司 | Event identification method and device, electronic equipment and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104112335A (en) * | 2014-07-25 | 2014-10-22 | 北京机械设备研究所 | Multi-information fusion based fatigue driving detecting method |
CN105105774A (en) * | 2015-10-09 | 2015-12-02 | 吉林大学 | Driver alertness monitoring method and system based on electroencephalogram information |
CN108937921A (en) * | 2018-04-11 | 2018-12-07 | 杭州电子科技大学 | In conjunction with the driving fatigue feature extracting method of empirical mode decomposition and energy spectral density |
CN112052895A (en) * | 2020-09-03 | 2020-12-08 | 同济大学 | Pure electric vehicle driving style clustering method |
CN112613786A (en) * | 2020-12-30 | 2021-04-06 | 武汉理工大学 | Individualized and differentiated driving risk evaluation method based on driving safety event |
-
2021
- 2021-06-07 CN CN202110631751.3A patent/CN113221843A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104112335A (en) * | 2014-07-25 | 2014-10-22 | 北京机械设备研究所 | Multi-information fusion based fatigue driving detecting method |
CN105105774A (en) * | 2015-10-09 | 2015-12-02 | 吉林大学 | Driver alertness monitoring method and system based on electroencephalogram information |
CN108937921A (en) * | 2018-04-11 | 2018-12-07 | 杭州电子科技大学 | In conjunction with the driving fatigue feature extracting method of empirical mode decomposition and energy spectral density |
CN112052895A (en) * | 2020-09-03 | 2020-12-08 | 同济大学 | Pure electric vehicle driving style clustering method |
CN112613786A (en) * | 2020-12-30 | 2021-04-06 | 武汉理工大学 | Individualized and differentiated driving risk evaluation method based on driving safety event |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113850554A (en) * | 2021-10-15 | 2021-12-28 | 上海吉七物联网科技有限公司 | Event identification method and device, electronic equipment and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN107492251B (en) | Driver identity recognition and driving state monitoring method based on machine learning and deep learning | |
CN109635852B (en) | User portrait construction and clustering method based on multi-dimensional attributes | |
CN111126438B (en) | Driving behavior recognition method and system | |
WO2021093012A1 (en) | Method and apparatus for extracting time series data formation mode and terminal device | |
CN115186594A (en) | Energy-saving speed optimization method under influence of man-vehicle-road coupling | |
CN113581188A (en) | Commercial vehicle driver driving style identification method based on Internet of vehicles data | |
CN113297795A (en) | Method for constructing running condition of pure electric vehicle | |
CN114169444A (en) | Driving style classification method considering risk potential field distribution under vehicle following working condition | |
CN113401130B (en) | Driving style recognition method and device based on environmental information and storage medium | |
CN113221843A (en) | Driving style classification method based on empirical mode decomposition characteristics | |
CN113642114B (en) | Personified random following driving behavior modeling method capable of making mistakes | |
CN117972488A (en) | Operation area-oriented full-class traffic conflict detection method | |
CN114048798A (en) | Automobile driving condition construction method based on improved noise reduction self-encoder | |
US10807602B2 (en) | Evaluating apparatus | |
CN112396118A (en) | GM-HMM-based driver acceleration intention modeling method | |
CN115774942A (en) | Driving style identification model modeling and statistical method based on Internet of vehicles real vehicle data and SVM | |
CN115718886A (en) | Construction method and device for pure electric vehicle running condition | |
CN115366891A (en) | Driving style recognition method, system and storage medium | |
CN114707573A (en) | Unsupervised driving style analysis method based on basic driving operation event | |
CN110667597B (en) | Driving style state identification method based on vehicle controller local area network data information | |
CN114987504A (en) | Dynamic driver identity recognition method and system based on incremental learning | |
CN113591780A (en) | Method and system for identifying driving risk of driver | |
Gao et al. | Driver identification based on stop-and-go events using naturalistic driving data | |
Ma et al. | Lane change analysis and prediction using mean impact value method and logistic regression model | |
CN116432108B (en) | Driving behavior evaluation and driving style online identification method under following scene |
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 | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210806 |
|
RJ01 | Rejection of invention patent application after publication |