CN110969844A - Method for calculating driving behavior similarity based on driving data and application - Google Patents
Method for calculating driving behavior similarity based on driving data and application Download PDFInfo
- Publication number
- CN110969844A CN110969844A CN201911132241.0A CN201911132241A CN110969844A CN 110969844 A CN110969844 A CN 110969844A CN 201911132241 A CN201911132241 A CN 201911132241A CN 110969844 A CN110969844 A CN 110969844A
- Authority
- CN
- China
- Prior art keywords
- driving behavior
- data
- driving
- behavior similarity
- vehicle
- 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
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/0104—Measuring and analyzing of parameters relative to traffic conditions
- G08G1/0125—Traffic data processing
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60W—CONJOINT CONTROL OF VEHICLE SUB-UNITS OF DIFFERENT TYPE OR DIFFERENT FUNCTION; CONTROL SYSTEMS SPECIALLY ADAPTED FOR HYBRID VEHICLES; ROAD VEHICLE DRIVE CONTROL SYSTEMS FOR PURPOSES NOT RELATED TO THE CONTROL OF A PARTICULAR SUB-UNIT
- B60W40/00—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models
- B60W40/08—Estimation or calculation of non-directly measurable driving parameters for road vehicle drive control systems not related to the control of a particular sub unit, e.g. by using mathematical models related to drivers or passengers
- B60W40/09—Driving style or behaviour
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07C—TIME OR ATTENDANCE REGISTERS; REGISTERING OR INDICATING THE WORKING OF MACHINES; GENERATING RANDOM NUMBERS; VOTING OR LOTTERY APPARATUS; ARRANGEMENTS, SYSTEMS OR APPARATUS FOR CHECKING NOT PROVIDED FOR ELSEWHERE
- G07C5/00—Registering or indicating the working of vehicles
- G07C5/08—Registering or indicating performance data other than driving, working, idle, or waiting time, with or without registering driving, working, idle or waiting time
- G07C5/0808—Diagnosing performance data
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G1/00—Traffic control systems for road vehicles
- G08G1/01—Detecting movement of traffic to be counted or controlled
- G08G1/017—Detecting movement of traffic to be counted or controlled identifying vehicles
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Automation & Control Theory (AREA)
- Mathematical Physics (AREA)
- Transportation (AREA)
- Mechanical Engineering (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention relates to a method for calculating driving behavior similarity based on driving data and application thereof, wherein the method comprises the steps of obtaining original data of a vehicle and processing the original data to extract a feature item for calculating the driving behavior similarity of a driver; establishing a driving behavior similarity calculation model of all conventional drivers of the vehicle, and training and verifying the driving behavior similarity calculation model; acquiring driving behavior data of a vehicle in a driving process in real time, and extracting a feature item for calculating the similarity of the driving behavior of the current driver; and inputting the characteristic items into the trained driving behavior similarity calculation model to obtain the driving behavior similarity of the current driver and all conventional drivers of the vehicle. According to the invention, by deploying the trained model, the similarity value of the driving behavior of the current driver and the driving behavior of the conventional driver can be given when the vehicle runs for a certain distance or after each driving journey is finished. By the method, the driving behavior is associated with the behavior similarity.
Description
Technical Field
The invention relates to the technical field of automobile safety, in particular to a method for calculating driving behavior similarity based on driving data and application thereof.
Background
With the rise of iot (internet of things) and iov (internet of vehicles), more and more automobiles have access to the cloud. Every minute and every second, the automobiles generate massive data, and how to mine and utilize the meaning and value of the data is a direction of positive attention of commercial organizations nowadays.
In the field of car networking (IoV), there have been many researches on driving behavior prediction or scoring and applications of landing, such as a hundredth driving behavior prediction method, and an adas (advanced driver Assistance systems) system-based real-time driving behavior scoring method and system of martian polar intelligence technology limited. As can be seen from the above abstract methods, the focus of each method for driving behavior research is different and mainly focuses on methods for predicting and judging the driving behavior. In fact, the mining based on the driving behavior data can also extend to many meaningful directions, such as the driving behavior similarity calculation method provided by the application.
In the internet field, there are many similarity calculation methods based on user usage behaviors, such as a user behavior similarity identification method based on a mobile terminal, and a user abnormal behavior detection method based on behavior similarity. For cross-domain reasons, it is difficult for traditional internet companies to learn some domain knowledge about car driving, which is less advantageous than traditional automotive industry enterprises. However, internet companies provide methods for calculating user behavior similarity based on their domain knowledge, and then use these methods to construct valuable functions, such as determining whether a user's account is stolen or not and whether the same account is used by multiple people or not. The method mainly collects operation traces of a user on a website, such as IP addresses of the user, click information and other data according to network information or some manual buried points. Since internet companies have difficulty in obtaining driving data of vehicles, it is difficult for them to construct a model regarding the similarity of driving behaviors, and even though many internet companies are actively participating IoV-related projects at present, most of them are working in cooperation with vehicle enterprises.
In academic circles, few methods or papers about driving behavior similarity calculation are published, but the methods or papers have the problems that the acquisition channel of driving data is limited, and the capability of mining the data is insufficient.
Disclosure of Invention
The invention provides a method for calculating driving behavior similarity based on driving data and application thereof, aiming at overcoming the defects in the prior art.
In order to solve the technical problems, the technical scheme of the invention is as follows:
a method for calculating driving behavior similarity based on driving data comprises the following steps:
acquiring original data of a vehicle and processing the original data to extract a feature item for calculating the similarity of driving behaviors of a driver;
establishing driving behavior similarity calculation models of all conventional drivers of the vehicle according to the feature items through a driving behavior similarity calculation algorithm;
training and verifying the established model to obtain a driving behavior similarity calculation model of all conventional drivers of the trained vehicle;
acquiring driving behavior data of a vehicle in a driving process in real time, and extracting a feature item for calculating the similarity of the driving behavior of the current driver;
and inputting the characteristic items into the trained driving behavior similarity calculation model to obtain the driving behavior similarity of the current driver and all conventional drivers of the vehicle.
Further, as a preferred technical scheme, the extracting the feature items for calculating the similarity of the driving behaviors of the driver specifically includes:
acquiring original data of a vehicle, and performing stroke division on the original data;
cleaning the original data of each journey in real time to obtain a basic index item of each journey, counting the basic index items of each journey in real time and guiding a vehicle owner to mark each journey;
and processing the statistical basic index items and extracting characteristic items for calculating the driving behavior similarity of the driver.
Further, as a preferred technical solution, the cleaning processing of the raw data of each trip specifically includes:
judging the abnormal data type in the original data of each stroke;
when the abnormal data is data missing, judging whether the missing data is a key data item, if so, discarding the data, otherwise, performing speculative recovery on the data;
when the abnormal data is abnormal data value, performing speculation recovery on the abnormal data;
and when the abnormal data is data repetition, performing deduplication processing on the abnormal data.
Further, as a preferred embodiment, the method for speculatively restoring the abnormal data includes a context completion method, an average value filling method, or a special value filling method.
Further, as a preferred technical solution, processing the statistical basic index items and extracting feature items for calculating the similarity of the driving behaviors of the driver specifically include:
processing the statistical basic index items through Spark;
and extracting statistics in a single trip greater than a certain preset distance as a characteristic item.
Further, as a preferred technical solution, establishing a driving behavior similarity calculation model of all conventional drivers of a vehicle specifically includes:
determining a driving behavior similarity calculation algorithm as a logistic regression function;
establishing a driving behavior similarity calculation model of all conventional drivers of the vehicle according to the characteristic items through a logistic regression function;
further, as a preferred technical scheme, the training and verification of the established model specifically comprises:
constructing a loss function and optimizing the loss function;
training the established model through the optimized loss function;
verifying the trained model by adopting a ten-fold cross-validation method by using Spark mllib;
and analyzing the accuracy, precision and recall rate of the model according to the verification result, and determining the optimal driving behavior similarity calculation model of the conventional driver according to the accuracy, precision and recall rate of the model.
Further, as a preferred technical solution, the obtaining of the driving behavior similarity of the current driver and all the conventional drivers of the vehicle specifically includes:
and assuming that the driving behavior of the current driver is a positive example, inputting the characteristic item of the driving behavior similarity of the current driver into all trained driving behavior similarity calculation models, and mapping the calculated probability value that the driving behavior of the current driver is the positive example to the driving behavior similarity values of all conventional drivers of the current driver and the vehicle.
Further, as a preferred technical solution, the method further comprises:
and optimizing the driving behavior similarity calculation model of all the conventional drivers of the trained vehicle according to the journey marked by the newly added vehicle owner.
Further, as a preferred technical solution, the characteristic items include: the number of accelerator steps per kilometer, the number of brake steps per kilometer, the average depth of an accelerator pedal, the average depth of a brake pedal, the average rotating speed of a steering wheel, the average vehicle speed, the rapid acceleration index, the rapid deceleration index and the rapid turning index.
The method for calculating the driving behavior similarity based on the driving data is applied to judging whether an account of a user is stolen or not or whether a vehicle is stolen.
Compared with the prior art, the technical scheme of the invention has the beneficial effects that:
according to the invention, by deploying the trained model, the similarity value of the driving behavior of the current driver and the driving behavior of the conventional driver can be given when the vehicle runs for a certain distance or after each driving journey is finished. By the method, the driving behavior is related to the behavior similarity, and the driving behavior similarity can be calculated by collecting and analyzing the driving data.
Meanwhile, the driving behavior similarity calculation is used as a basic method, so that valuable upper-layer applications are constructed. Such as: a taxi company monitors whether a registered driver of the car is using the car; to privately owned car owners to make decisions as to whether their vehicles are stolen, etc.
Drawings
FIG. 1 is a flow chart of the method steps of the present invention.
The drawings are for illustrative purposes only and are not to be construed as limiting the patent; for the purpose of better illustrating the embodiments, certain features of the drawings may be omitted, enlarged or reduced, and do not represent the size of an actual product; it will be understood by those skilled in the art that certain well-known structures in the drawings and descriptions thereof may be omitted; the same or similar reference numerals correspond to the same or similar parts; the terms describing positional relationships in the drawings are for illustrative purposes only and are not to be construed as limiting the patent.
Detailed Description
The following detailed description of the preferred embodiments of the present invention, taken in conjunction with the accompanying drawings, will make the advantages and features of the invention easier to understand for those skilled in the art and will therefore make the scope of the invention more clearly defined.
Example 1
A method for calculating driving behavior similarity based on driving data is disclosed, as shown in FIG. 1, and comprises the following steps:
and S10, acquiring the raw data of the vehicle and processing the raw data to extract the feature items for calculating the similarity of the driving behaviors of the driver.
And S101, acquiring original data of the vehicle, and performing stroke division on the original data.
The method specifically comprises the following steps: the method comprises the following steps that a vehicle collecting device collects driving data of a vehicle according to a certain time frequency, and then the driving data are uploaded to a cloud terminal for storage in real time through a data format specified by a terminal and the cloud terminal; the method comprises the steps of obtaining driving data of a vehicle stored in a cloud end and carrying out stroke division on the driving data; in this step, the vehicle starts to acquire and send the driving data after ignition, so that when the cloud does not receive the driving data of a certain vin for a certain time, the cloud automatically divides the driving data stored at present into a stroke. The driving data of the vehicle in the step is the original data of the vehicle; meanwhile, in the step, the vehicle starts to acquire and send the driving data after ignition, and the vehicle stops acquiring and sending the driving data before flameout, so that the acquisition and sending of the driving data are stopped at least 2 minutes between flameout and ignition of the vehicle, and when the cloud end does not receive the driving data of a certain vin in 2 minutes or more, the cloud end automatically divides the driving data stored at present into a stroke.
Wherein the raw data of the vehicle comprises: the vin number of the vehicle, the source of the identification data, the longitude of the current position of the vehicle, the latitude of the current position of the vehicle, the number of steering angles of a steering wheel, the rotating speed of the steering wheel, the current driving speed, the data acquisition time, the total mileage of the vehicle driven till now and the longitudinal acceleration of the vehicle; current engine speed, instantaneous fuel quantity, lateral acceleration of the vehicle, current brake pedal position, current accelerator pedal position, etc.
And S102, cleaning the original data of each stroke in real time to obtain a basic index item of each stroke, counting the basic index items of each stroke in real time and guiding a vehicle owner to mark each stroke.
The specific cleaning treatment of the original data is as follows:
judging abnormal data types in the original data of each stroke, wherein the abnormal data types comprise data loss, data value abnormality, data repetition and the like;
when abnormal data is judged to be data missing, judging whether the missing data is a key data item, if so, discarding the data, otherwise, speculatively recovering the data through a context completion method, an average value filling method or a special value filling method and the like so as to complete the data;
when the abnormal data is judged to be abnormal data values, the abnormal data is speculatively restored by a context completion method, an average filling method or a special value filling method;
when the abnormal data is judged to be data repetition, carrying out duplicate removal processing on the abnormal data;
and obtaining a basic index item of each stroke after the cleaning is finished.
The method for counting the basic index items of each journey in real time and guiding the vehicle owner to mark each journey specifically comprises the following steps:
the vehicle owner marks each section of travel based on the counted basic index item of each travel so as to enable the section of travel to be associated with the corresponding driver, the safety of the vehicle is improved, the more travel data marked by the vehicle owner is, the more data used for training is, and the higher the accuracy of the established model is.
Wherein, the basic index item comprises:
driving mileage: and obtaining the data according to the kilometer of the original data, wherein the value of less than 1 kilometer is taken according to 1 kilometer. Subtracting the initial travel mileage of the travel from the travel mileage of the final data of the travel to obtain the travel mileage;
the number of accelerator steps: the number of times of stepping on the accelerator in the travel is determined according to the condition that the state value of the accelerator pedal is not zero;
depth of accelerator pedal per accelerator step: filtering data with an accelerator pedal state value not being zero in the stroke, and storing the data;
the braking frequency is: the number of times of stepping on the brake in the stroke is determined according to the condition that the state value of the brake pedal is not zero;
depth of brake pedal for each brake application: filtering data with a brake pedal state value not being zero in the stroke, and storing the data;
rotation speed per rotation of steering wheel: filtering out data with a steering wheel rotation value in a stroke, and storing the data;
number of steering wheel rotations: filtering data with a steering wheel rotation value in a stroke, and counting the total times;
average speed: dividing the sum of all speeds in the stroke by the number of all speeds;
the number of rapid acceleration times: when the longitudinal acceleration of the vehicle is more than 3m/s ^2, determining that the vehicle is accelerated rapidly for one time;
sharp turn times: when the transverse acceleration of the vehicle is more than 4m/s ^2, judging that the vehicle is in sharp turn for one time;
the number of rapid deceleration times: and when the longitudinal acceleration of the vehicle is less than-4.5 m/s 2, determining that the vehicle is suddenly decelerated.
And S103, processing the statistical basic index items and extracting characteristic items for calculating the driving behavior similarity of the driver.
The method specifically comprises the following steps:
processing the statistical basic index items through Spark;
and extracting statistics in a single trip greater than a certain preset distance as a characteristic item.
In the step, as the statistics are specific to each stroke, all the extracted feature items are subjected to statistics on data with a single stroke larger than a certain preset distance in order to avoid the deviation of the overall result caused by an excessively short stroke; for example, more than 2 km, the extracted feature items are all feature items capable of reflecting the current driving habits of the driver.
Wherein the characteristic items include: the number of times of stepping on the accelerator per kilometer, the number of times of stepping on the brake per kilometer, the average depth of the accelerator pedal, the average depth of the brake pedal, the average rotating speed of a steering wheel, the average vehicle speed, the rapid acceleration index, the rapid deceleration index, the rapid turning index and the like.
And S20, establishing a driving behavior similarity calculation model of all conventional drivers of the vehicle according to the characteristic items through a driving behavior similarity calculation algorithm.
The method comprises the following specific steps:
s201, determining a driving behavior similarity calculation algorithm as a logistic regression function;
in the invention, because the characteristic values are linear, the final required result is the similarity, and meanwhile, the similarity is converted into the binary problem that the same driver is not driving according to the calculated driving behavior similarity, the selected algorithm is a binary algorithm, in particular a logistic regression function.
And S202, establishing a driving behavior similarity calculation model of all conventional drivers of the vehicle by adopting a logistic regression function according to the marked characteristic items of the journey.
In this step, the model established by the logistic regression function is as follows:
wherein the content of the first and second substances,
a vector of eigenvalues representing the eigenvalues of the eigenvalues,
the weight coefficients representing the individual feature terms,
representing a constant.
And S30, training and verifying the established model to obtain a driving behavior similarity calculation model of all conventional drivers of the trained vehicle.
And S301, constructing a loss function through a maximum likelihood method.
The established loss function is as follows:
wherein the content of the first and second substances,
equivalent to in logistic regression function
M represents the number of samples to be trained,
the y value predicted by the parameters theta and x; y represents the value of y in the training sample, i.e. the standard answer; the upper corner mark (i) indicates the ith sample.
The constructed loss function is optimized by gradient descent or newton's method.
Modulo of creation by optimized loss function pairThe patterns are trained to determine those in a logistic regression function
Value sum
Value, i.e. determining in the established model
Value sum
And (4) completing the training of the model.
In the process, when the overfitting condition occurs, a regular penalty term can be introduced to penalize the feature term, so that the overfitting condition is relieved.
The trained model was validated using Spark mllib with a ten-fold cross-validation method.
And analyzing the accuracy, precision and recall rate of the model according to the verification result, and determining the optimal driving behavior similarity calculation model of the conventional driver according to the accuracy, precision and recall rate of the model.
And S40, acquiring the driving behavior data of the vehicle in the driving process in real time, and extracting the feature items for calculating the similarity of the driving behavior of the current driver.
S401, acquiring driving behavior data uploaded to the cloud end by the vehicle in real time, and counting the characteristic item condition of each stroke according to the characteristic items required for calculating the driving behavior similarity;
and S402, when the vehicle runs for a certain mileage or a certain journey is finished, the cloud end extracts the characteristic items of the current journey or the previous journey for calculating the similarity of the driving behaviors of the current driver.
And S50, inputting the characteristic items into the trained driving behavior similarity calculation model to obtain the driving behavior similarity of the current driver and all conventional drivers of the vehicle.
The method specifically comprises the following steps: and obtaining the probability that the driving behavior of a certain trip is a positive example through a training model, and mapping the probability value to a driving behavior similarity value.
And assuming that the driving behavior of the current driver is a positive example, inputting the characteristic item of the driving behavior similarity of the current driver into all trained driving behavior similarity calculation models, and mapping the calculated probability value that the driving behavior of the current driver is the positive example to the driving behavior similarity values of all conventional drivers of the current driver and the vehicle.
The trained driving behavior similarity calculation model is as follows:
wherein the value of theta is obtained by training the model and is known,
namely, taking the driving behavior of the current driver as a positive example, calculating a probability value of y = 1 through a trained driving behavior similarity calculation model, and mapping the probability value of y = 1 to driving behavior similarity values of the current driver and all conventional drivers of the vehicle.
And S60, updating and optimizing the driving behavior similarity calculation model of all the conventional drivers of the trained vehicle according to the journey marked by the newly added vehicle owner.
The method specifically comprises the following steps: with the increase of the service time of the vehicle, the amount of marked data accumulated by the cloud end is increased, and the model is updated through the marked data amount so as to improve the accuracy of the model; the updating frequency of the model can be determined according to the newly added marking data volume of the user or a certain time period, and the model is reconstructed when the conditions are met, so that a closed loop is formed, and the model is more and more approximate to the driving condition of the actual user.
Example 2
The application of the method for calculating the driving behavior similarity based on the driving data is based on the method in the embodiment 1 and is suitable for judging whether the account of a user is stolen or not, or whether a vehicle is stolen or not and the like.
The method specifically comprises the following steps: by the method of the embodiment 1, whether the same account is used by multiple persons can be judged, so that whether the account of the user is stolen can be judged;
meanwhile, whether the same vehicle is driven by a plurality of drivers can be judged through the method of embodiment 1, so that whether the vehicle is stolen can be judged.
The method can also be used in other situations where the operation of the same person is required to be confirmed, so as to determine the safety of the used object.
It should be understood that the above-described embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the embodiments of the present invention. Other variations and modifications will be apparent to persons skilled in the art in light of the above description. And are neither required nor exhaustive of all embodiments. Any modification, equivalent replacement, and improvement made within the spirit and principle of the present invention should be included in the protection scope of the claims of the present invention.
Claims (11)
1. A method for calculating driving behavior similarity based on driving data is characterized by comprising the following steps:
acquiring original data of a vehicle and processing the original data to extract a feature item for calculating the similarity of driving behaviors of a driver;
establishing driving behavior similarity calculation models of all conventional drivers of the vehicle according to the feature items through a driving behavior similarity calculation algorithm;
training and verifying the established model to obtain a driving behavior similarity calculation model of all conventional drivers of the trained vehicle;
acquiring driving behavior data of a vehicle in a driving process in real time, and extracting a feature item for calculating the similarity of the driving behavior of the current driver;
and inputting the characteristic items into the trained driving behavior similarity calculation model to obtain the driving behavior similarity of the current driver and all conventional drivers of the vehicle.
2. The method for calculating the driving behavior similarity based on the driving data according to claim 1, wherein the extracting the feature items for calculating the driving behavior similarity of the driver specifically comprises:
acquiring original data of a vehicle, and performing stroke division on the original data;
cleaning the original data of each journey in real time to obtain a basic index item of each journey, counting the basic index items of each journey in real time and guiding a vehicle owner to mark each journey;
and processing the statistical basic index items and extracting characteristic items for calculating the driving behavior similarity of the driver.
3. The method for calculating the driving behavior similarity based on the driving data as claimed in claim 2, wherein the cleaning of the raw data of each trip specifically comprises:
judging the abnormal data type in the original data of each stroke;
when the abnormal data is data missing, judging whether the missing data is a key data item, if so, discarding the data, otherwise, performing speculative recovery on the data;
when the abnormal data is abnormal data value, performing speculation recovery on the abnormal data;
and when the abnormal data is data repetition, performing deduplication processing on the abnormal data.
4. The method for calculating the driving behavior similarity based on the driving data as claimed in claim 3, wherein the method for speculatively restoring the abnormal data comprises a context completion method, an average value filling method or a special value filling method.
5. The method for calculating the driving behavior similarity based on the driving data as claimed in claim 2, wherein the step of processing the statistical basic index items and extracting the feature items for calculating the driving behavior similarity of the driver specifically comprises the steps of:
processing the statistical basic index items through Spark;
and extracting statistics in a single trip greater than a certain preset distance as a characteristic item.
6. The method for calculating the driving behavior similarity based on the driving data according to claim 1, wherein the establishing of the driving behavior similarity calculation model of all the regular drivers of the vehicle specifically comprises:
determining a driving behavior similarity calculation algorithm as a logistic regression function;
and establishing a driving behavior similarity calculation model of all conventional drivers of the vehicle according to the characteristic items through a logistic regression function.
7. The method for calculating the driving behavior similarity based on the driving data as claimed in claim 6, wherein the training and verification of the established model specifically comprises:
constructing a loss function and optimizing the loss function;
training the established model through the optimized loss function;
verifying the trained model by adopting a ten-fold cross-validation method by using Spark mllib;
and analyzing the accuracy, precision and recall rate of the model according to the verification result, and determining the optimal driving behavior similarity calculation model of the conventional driver according to the accuracy, precision and recall rate of the model.
8. The method for calculating the driving behavior similarity based on the driving data as claimed in claim 1, wherein the obtaining of the driving behavior similarity between the current driver and all the regular drivers of the vehicle specifically comprises:
and assuming that the driving behavior of the current driver is a positive example, inputting the characteristic item of the driving behavior similarity of the current driver into all trained driving behavior similarity calculation models, and mapping the calculated probability value that the driving behavior of the current driver is the positive example to the driving behavior similarity values of all conventional drivers of the current driver and the vehicle.
9. The method for calculating driving behavior similarity based on driving data according to claim 1, further comprising:
and optimizing the driving behavior similarity calculation model of all the conventional drivers of the trained vehicle according to the journey marked by the newly added vehicle owner.
10. The method for calculating the driving behavior similarity based on the driving data according to claim 1, wherein the characteristic items comprise: the number of accelerator steps per kilometer, the number of brake steps per kilometer, the average depth of an accelerator pedal, the average depth of a brake pedal, the average rotating speed of a steering wheel, the average vehicle speed, the rapid acceleration index, the rapid deceleration index and the rapid turning index.
11. Use of a method according to any of claims 1-10 for calculating driving behavior similarity based on driving data, adapted to determine whether a user's account is stolen or whether a vehicle is stolen.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911132241.0A CN110969844A (en) | 2019-11-19 | 2019-11-19 | Method for calculating driving behavior similarity based on driving data and application |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201911132241.0A CN110969844A (en) | 2019-11-19 | 2019-11-19 | Method for calculating driving behavior similarity based on driving data and application |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110969844A true CN110969844A (en) | 2020-04-07 |
Family
ID=70030864
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201911132241.0A Pending CN110969844A (en) | 2019-11-19 | 2019-11-19 | Method for calculating driving behavior similarity based on driving data and application |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110969844A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111460076A (en) * | 2020-04-20 | 2020-07-28 | 广州亚美信息科技有限公司 | Driving route familiarity determination method, device, computer equipment and storage medium |
| CN112418569A (en) * | 2021-01-25 | 2021-02-26 | 北京云真信科技有限公司 | User portrait determination method, electronic equipment and computer readable storage medium |
| CN115439954A (en) * | 2022-08-29 | 2022-12-06 | 上海寻序人工智能科技有限公司 | Data closed-loop method based on cloud large model |
| CN116821805A (en) * | 2023-06-28 | 2023-09-29 | 运脉云技术有限公司 | Vehicle service platform system for monitoring driving behavior and driving behavior monitoring method |
| CN116911610A (en) * | 2023-07-20 | 2023-10-20 | 上海钢联物流股份有限公司 | Method and system for monitoring, evaluating and early warning of driving safety risk of transport vehicle |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9067565B2 (en) * | 2006-05-22 | 2015-06-30 | Inthinc Technology Solutions, Inc. | System and method for evaluating driver behavior |
| CN104765995A (en) * | 2015-04-24 | 2015-07-08 | 福建师范大学 | Intelligent device identity authentication method and client side based on touch screen operation |
| CN105069448A (en) * | 2015-09-29 | 2015-11-18 | 厦门中控生物识别信息技术有限公司 | True and false face identification method and device |
| CN105761329A (en) * | 2016-03-16 | 2016-07-13 | 成都信息工程大学 | Method of identifying driver based on driving habits |
| CN106945631A (en) * | 2017-04-17 | 2017-07-14 | 成都雅骏新能源汽车科技股份有限公司 | A kind of anti-stealing method for vehicles based on driving behavior |
| CN107323423A (en) * | 2016-04-29 | 2017-11-07 | 宇龙计算机通信科技(深圳)有限公司 | A kind of anti-stealing method for vehicles and system |
| CN108229567A (en) * | 2018-01-09 | 2018-06-29 | 北京荣之联科技股份有限公司 | Driver identity recognition methods and device |
| CN108280482A (en) * | 2018-01-30 | 2018-07-13 | 广州小鹏汽车科技有限公司 | Driver's recognition methods based on user behavior, apparatus and system |
| CN108512815A (en) * | 2017-02-28 | 2018-09-07 | 腾讯科技（北京）有限公司 | Door chain detection method, door chain detection device and server |
| CN108944799A (en) * | 2017-05-18 | 2018-12-07 | 腾讯科技（深圳）有限公司 | Vehicle drive abnormal behavior treating method and apparatus |
| CN109583508A (en) * | 2018-12-10 | 2019-04-05 | 长安大学 | A kind of vehicle abnormality acceleration and deceleration Activity recognition method based on deep learning |
| CN109606311A (en) * | 2017-09-30 | 2019-04-12 | 比亚迪汽车工业有限公司 | Vehicle authentication method, device and storage medium |
-
2019
- 2019-11-19 CN CN201911132241.0A patent/CN110969844A/en active Pending
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9067565B2 (en) * | 2006-05-22 | 2015-06-30 | Inthinc Technology Solutions, Inc. | System and method for evaluating driver behavior |
| CN104765995A (en) * | 2015-04-24 | 2015-07-08 | 福建师范大学 | Intelligent device identity authentication method and client side based on touch screen operation |
| CN105069448A (en) * | 2015-09-29 | 2015-11-18 | 厦门中控生物识别信息技术有限公司 | True and false face identification method and device |
| CN105761329A (en) * | 2016-03-16 | 2016-07-13 | 成都信息工程大学 | Method of identifying driver based on driving habits |
| CN107323423A (en) * | 2016-04-29 | 2017-11-07 | 宇龙计算机通信科技(深圳)有限公司 | A kind of anti-stealing method for vehicles and system |
| CN108512815A (en) * | 2017-02-28 | 2018-09-07 | 腾讯科技（北京）有限公司 | Door chain detection method, door chain detection device and server |
| CN106945631A (en) * | 2017-04-17 | 2017-07-14 | 成都雅骏新能源汽车科技股份有限公司 | A kind of anti-stealing method for vehicles based on driving behavior |
| CN108944799A (en) * | 2017-05-18 | 2018-12-07 | 腾讯科技（深圳）有限公司 | Vehicle drive abnormal behavior treating method and apparatus |
| CN109606311A (en) * | 2017-09-30 | 2019-04-12 | 比亚迪汽车工业有限公司 | Vehicle authentication method, device and storage medium |
| CN108229567A (en) * | 2018-01-09 | 2018-06-29 | 北京荣之联科技股份有限公司 | Driver identity recognition methods and device |
| CN108280482A (en) * | 2018-01-30 | 2018-07-13 | 广州小鹏汽车科技有限公司 | Driver's recognition methods based on user behavior, apparatus and system |
| CN109583508A (en) * | 2018-12-10 | 2019-04-05 | 长安大学 | A kind of vehicle abnormality acceleration and deceleration Activity recognition method based on deep learning |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111460076A (en) * | 2020-04-20 | 2020-07-28 | 广州亚美信息科技有限公司 | Driving route familiarity determination method, device, computer equipment and storage medium |
| CN111460076B (en) * | 2020-04-20 | 2023-05-26 | 亚美智联数据科技有限公司 | Driving route familiarity determination method, driving route familiarity determination device, computer device, and storage medium |
| CN112418569A (en) * | 2021-01-25 | 2021-02-26 | 北京云真信科技有限公司 | User portrait determination method, electronic equipment and computer readable storage medium |
| CN115439954A (en) * | 2022-08-29 | 2022-12-06 | 上海寻序人工智能科技有限公司 | Data closed-loop method based on cloud large model |
| CN116821805A (en) * | 2023-06-28 | 2023-09-29 | 运脉云技术有限公司 | Vehicle service platform system for monitoring driving behavior and driving behavior monitoring method |
| CN116911610A (en) * | 2023-07-20 | 2023-10-20 | 上海钢联物流股份有限公司 | Method and system for monitoring, evaluating and early warning of driving safety risk of transport vehicle |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110969844A (en) | Method for calculating driving behavior similarity based on driving data and application | |
| CN111062240B (en) | Monitoring method and device for automobile driving safety, computer equipment and storage medium | |
| US11688212B2 (en) | Machine learning techniques for classifying driver behavior | |
| CN111688713B (en) | Driving behavior analysis method and device | |
| CN109670970B (en) | Driving behavior scoring method and device and computer readable storage medium | |
| CN105045788A (en) | Method of processing and analysing vehicle driving big data and system thereof | |
| Castignani et al. | An evaluation study of driver profiling fuzzy algorithms using smartphones | |
| Matousek et al. | Detecting anomalous driving behavior using neural networks | |
| CN106097709A (en) | Driving behavior recognition methods based on intelligent vehicle mounted terminal | |
| CN112700201A (en) | Goods source recommendation method, electronic device and storage medium | |
| Wu et al. | Clustering of several typical behavioral characteristics of commercial vehicle drivers based on GPS data mining: Case study of highways in China | |
| Hassan et al. | Road anomaly classification for low-cost road maintenance and route quality maps | |
| CN113060146B (en) | Longitudinal tracking control method, device, equipment and storage medium | |
| Das et al. | Driver behaviour profiling in VANETs: comparison of ensemble machine learning techniques | |
| Park et al. | This car is mine!: Automobile theft countermeasure leveraging driver identification with generative adversarial networks | |
| EP3382570A1 (en) | Method for characterizing driving events of a vehicle based on an accelerometer sensor | |
| Yarlagadda et al. | Heterogeneity in the driver behavior: an exploratory study using real-time driving data | |
| Van Hinsbergh et al. | Vehicle point of interest detection using in-car data | |
| Attal et al. | Riding patterns recognition for Powered two-wheelers users' behaviors analysis | |
| CN108268678B (en) | Driving behavior analysis method, device and system | |
| Fazio et al. | A new application for analyzing driving behavior and environment characterization in transportation systems based on a fuzzy logic approach | |
| CN107608270A (en) | The analysis method and device of automotive performance | |
| CN114802264A (en) | Vehicle control method and device and electronic equipment | |
| Jain et al. | Review of computational techniques for modelling eco-safe driving behavior | |
| Reichenbach et al. | A model for traffic incident prediction using emergency braking data |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20200407 |