CN111717210A - Detection method for separation of driver from steering wheel in relative static state of hands - Google Patents

Detection method for separation of driver from steering wheel in relative static state of hands Download PDF

Info

Publication number
CN111717210A
CN111717210A CN202010485078.2A CN202010485078A CN111717210A CN 111717210 A CN111717210 A CN 111717210A CN 202010485078 A CN202010485078 A CN 202010485078A CN 111717210 A CN111717210 A CN 111717210A
Authority
CN
China
Prior art keywords
data
driver
acceleration
hand
steering wheel
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.)
Granted
Application number
CN202010485078.2A
Other languages
Chinese (zh)
Other versions
CN111717210B (en
Inventor
孙棣华
赵敏
高治平
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Chongqing University
Original Assignee
Chongqing University
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Chongqing University filed Critical Chongqing University
Priority to CN202010485078.2A priority Critical patent/CN111717210B/en
Publication of CN111717210A publication Critical patent/CN111717210A/en
Application granted granted Critical
Publication of CN111717210B publication Critical patent/CN111717210B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • 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
    • 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
    • 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
    • G07C5/0816—Indicating performance data, e.g. occurrence of a malfunction
    • 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
    • G07C5/0841—Registering performance data

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Automation & Control Theory (AREA)
  • Mathematical Physics (AREA)
  • Transportation (AREA)
  • Mechanical Engineering (AREA)
  • General Physics & Mathematics (AREA)
  • Emergency Alarm Devices (AREA)
  • Traffic Control Systems (AREA)

Abstract

The invention discloses a method for detecting the separation of a driver from a steering wheel in a state that the hands of the driver are relatively static, which adopts an intelligent watch and an intelligent mobile phone to collect information of a vehicle and the driver, and judges whether the hands of the driver are separated from the steering wheel by using an anomaly detection algorithm based on a Gauss multivariate model, thereby avoiding the problem that the current detection system scheme based on computer vision is influenced by external factors such as illumination and the like, improving the detection accuracy and stability, and simultaneously not causing psychological interference to the driver; the smart phone and the smart watch are used as sensors, so that the equipment cost is reduced, and the popularization is facilitated.

Description

Detection method for separation of driver from steering wheel in relative static state of hands
Technical Field
The invention belongs to the technical field of traffic safety, and particularly relates to a method for detecting the separation of a driver from a steering wheel in a relatively static state of the driver's hand.
Background
In the driving process of a vehicle, the driving behavior of a driver plays a decisive role in road safety, and the driver should hold the steering wheel with both hands at the 9 o 'clock position and the 3 o' clock position in the driving process so as to prevent the driver from not reacting in time when an accident happens. However, due to boredom and fatigue after long driving or simply poor driving habits, drivers may not always comply with this regulation and most drivers are not aware that they are distracted when driving. When a driver intentionally or unintentionally leaves the steering wheel with one hand for a long time, places two hands at the lower end of the steering wheel or leaves the steering wheel with two hands at the same time, serious traffic accidents are easily caused. Therefore, a system that can accurately detect dangerous driving behavior of a driver and warn the driver can not only call the driver's alertness, but also enable the driver to improve driving skills by himself, help promote safer driving and reduce traffic accidents and contribute to social safety.
At present, a driver hand position identification method based on computer vision easily causes privacy problems and psychological interference to a driver, is easily influenced by factors such as external illumination and the like, and is not beneficial to popularization due to the fact that other sensors are installed on a steering wheel to form an extra cost problem. Therefore, a method for detecting the driver's hand disengagement from the steering wheel, which is not easily interfered by the outside, has low cost, does not interfere the driver and can accurately detect the driver's hand disengagement, is needed.
The invention provides a hand-off steering wheel detection model in a state that the hands of a driver are relatively static, and the hand state of the driver in a vehicle can be divided into a relatively static state and a motion state relative to the vehicle.
Disclosure of Invention
In view of the above, the present invention is directed to a method for detecting a driver's hand being out of steering wheel in a relatively stationary state.
The purpose of the invention is realized by the following technical scheme:
a detection method for a steering wheel separated from a driver in a relatively static state of hands comprises the following steps:
the method comprises the following steps: in the running process of a vehicle, acquiring the hand acceleration of a driver and the vehicle acceleration by using a sensor according to a fixed frequency, converting a sensor coordinate system into a world coordinate system, and recording the time t of acquiring the hand acceleration of the driver and the vehicle acceleration every time to form a data sequence;
step two: for driver hand acceleration data set DwAnd a vehicle acceleration data set DpPerforming sliding window processing to obtain a data fragment D ═ D1，D2，D3…Dn}，DnThe expression of (a) is as follows:
Figure BDA0002518846970000021
in the formula: dnFor the nth data segment, the data segment is,
Figure BDA0002518846970000022
the difference between the absolute values of the acceleration of the driver's hand and the acceleration of the vehicle at the first moment in the data segment,
Figure BDA0002518846970000023
is a time point t in the windowiA data value of time, and
Figure BDA0002518846970000024
the expression of (a) is as follows:
Figure BDA0002518846970000025
in the formula:
Figure BDA0002518846970000026
is indicative of the acceleration of the driver's hand,
Figure BDA0002518846970000027
represents vehicle acceleration;
then calculating the average value of all data in the sliding window to obtain each data segment DnAverage value D of middle datai，meanThe calculation formula is shown as the following formula:
Figure BDA0002518846970000028
in the formula: di，meanIs the average value of the data of the ith sliding window,
Figure BDA0002518846970000029
indicating the ith data in the window and l the length of a sliding window.
Whether the driver's hand is moving within each window is determined using the following equation. When D is presenti，meanWhen not greater than the threshold, it is considered that the driver's hand is relatively stationary within the window:
Figure BDA00025188469700000210
step three: extracting characteristic vectors capable of distinguishing the hand-held steering wheel of the driver from the hand-off steering wheel of the driver according to the vehicle acceleration and the acceleration information of the hands of the driver for the window data of which the hands of the driver are in the relatively static state in the step two;
step four: an algorithm for determining the optimal threshold value is constructed, and the specific process is as follows:
dividing the data processed in the third step into three parts at random: the first part is a training data set, the second part is a cross validation set, and the third part is a test set; detecting whether a driver receiving part leaves a steering wheel or not by adopting an anomaly detection algorithm, estimating parameter values of the anomaly detection algorithm by utilizing a training data set, then trying to identify whether data are abnormal or not on a cross validation set by using different thresholds, selecting an optimal threshold according to a test result, and finally testing a detection model on a test set until the detection accuracy of the detection model reaches the actual application requirement.
Step five: and detecting the data acquired in real time by using the trained anomaly detection algorithm, and if the detection result is that the hands of the driver leave the steering wheel, giving an early warning prompt to the driver.
Further, the sensor is a smart phone and a smart watch, the acceleration of the driver's hand is measured by the smart watch, the acceleration of the vehicle is measured by the smart phone, the smart phone is fixed in the carriage and collects the acceleration of the vehicle running through the smart watchIn-flight mobile phone acceleration sensor data
Figure BDA0002518846970000031
Gyroscope sensor data
Figure BDA0002518846970000032
Figure BDA0002518846970000033
Acceleration of gravity
Figure BDA0002518846970000034
Linear acceleration
Figure BDA0002518846970000035
The intelligent watch is bound at the wrist of a driver and used for collecting watch acceleration sensor data
Figure BDA0002518846970000036
Acceleration of gravity
Figure BDA0002518846970000037
Linear acceleration
Figure BDA0002518846970000038
Converting a smart phone coordinate system and a smart watch coordinate system into a world coordinate system by using a conversion matrix R, wherein R is a numerical matrix of 3x 3; the data sequence is stored in the form of
Figure BDA0002518846970000039
(i ═ 1,2,3 …) in which tiIs the acquisition time. Further, the specific process of the third step is as follows:
firstly, extracting a characteristic value from window data of an intelligent watch with a relatively static hand of a driver, and calculating a three-axis mean value wg of gravity acceleration of the intelligent watch through the following formulax,mean、wgy,mean、wgz,meanWhere l is the window length.
Figure BDA00025188469700000310
Figure BDA00025188469700000311
Figure BDA00025188469700000312
The vertical component of the vehicle vibration signal measured by the smart watch is calculated by:
Figure BDA00025188469700000313
in the formula:
Figure BDA00025188469700000314
representing the acceleration of gravity measured by the smart watch,
Figure BDA00025188469700000315
representing the linear acceleration measured by the smart watch, av,watchRepresenting the vertical component of the vehicle vibration signal measured by the smart watch. Then the vertical component of the vehicle vibration signal measured by the smartphone is calculated by:
Figure BDA00025188469700000316
in the formula:
Figure BDA00025188469700000317
representing the acceleration of gravity measured by the smart watch,
Figure BDA00025188469700000318
representing the linear acceleration measured by the smart watch, av,phoneRepresenting the vertical component of the vehicle vibration signal measured by the smart watch. Then, the variance Var (a) of the vertical component of the vibration signal is calculatedv,watch) And Var (a)v,phone) And the ratio of the twoFinally, the characteristic vector is obtained
Figure BDA0002518846970000041
Further, the anomaly detection algorithm of the step four is an anomaly detection algorithm based on a multivariate Gaussian model,
the concrete process of the step four is as follows:
and C, randomly selecting 60% of data of the hand-held steering wheel from the data processed in the step three to construct a training data set, forming a cross validation set by 20% of data of the hand-held steering wheel and 50% of data of the hand-off steering wheel, and constructing a test set by using the rest 20% of data of the hand-held steering wheel and 50% of data of the hand-off steering wheel. First, the multivariate height is estimated on the training set using the following equation
Parameters of the model:
Figure BDA0002518846970000042
in the formula: x is the number of(i)Represents one sample in the training set, and m is the number of samples in the training set. When a test specimen is given to it,
its probability density function is calculated by:
Figure BDA0002518846970000043
in the formula: x and μ are vectors of 1 × d; .
And then, trying to identify whether the data are abnormal by using different thresholds on a cross validation set, selecting an optimal threshold according to an F1 value of a test result, and finally testing the detection model on the test set until the detection accuracy of the detection model reaches the actual application requirement.
Due to the adoption of the technical scheme, the invention has the following beneficial effects:
the invention judges whether the hand of the driver is separated from the steering wheel by using the anomaly detection algorithm, thereby avoiding the problem that the current detection system scheme based on computer vision is influenced by external factors such as illumination and the like, improving the detection accuracy and stability and simultaneously causing no psychological interference to the driver; the smart phone and the smart watch are used as sensors, so that the equipment cost is reduced, and the popularization is facilitated.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention. The objectives and other advantages of the invention may be realized and attained by the means of the instrumentalities and combinations particularly pointed out hereinafter.
Drawings
The drawings of the present invention are described below.
FIG. 1 is a schematic flow chart of a method for detecting a driver's hand being out of steering wheel in a relatively stationary state;
fig. 2 is a diagram of standard placement poses for a smartphone.
Detailed Description
The invention is further illustrated by the following figures and examples.
Example 1
As shown in fig. 1-2, the method for detecting the separation of the steering wheel of the driver in the relatively static state of the hand provided by the embodiment includes the following steps:
the method comprises the following steps: in the running process of a vehicle, acquiring the hand acceleration of a driver and the vehicle acceleration by using a sensor according to a fixed frequency, converting a sensor coordinate system into a world coordinate system, and recording the time t of acquiring the hand acceleration of the driver and the vehicle acceleration every time to form a data sequence;
step two: for driver hand acceleration data set DwAnd a vehicle acceleration data set DpPerforming sliding window processing to obtain a data fragment D ═ D1，D2，D3…Dn}，DnThe expression of (a) is as follows:
Figure BDA0002518846970000051
in the formula: dnFor the nth data segment, the data segment is,
Figure BDA0002518846970000052
the difference between the absolute values of the acceleration of the driver's hand and the acceleration of the vehicle at the first moment in the data segment,
Figure BDA0002518846970000053
is a time point t in the windowiA data value of time, and
Figure BDA0002518846970000054
the expression of (a) is as follows:
Figure BDA0002518846970000055
in the formula:
Figure BDA0002518846970000056
is indicative of the acceleration of the driver's hand,
Figure BDA0002518846970000057
represents vehicle acceleration;
then calculating the average value of all data in the sliding window to obtain each data segment DnAverage value D of middle datai，meanThe calculation formula is shown as the following formula:
Figure BDA0002518846970000058
in the formula: di，meanIs the average value of the data of the ith sliding window,
Figure BDA0002518846970000059
indicating the ith data in the window and l the length of a sliding window.
Whether the driver's hand is moving within each window is determined using the following equation. When D is presenti，meanWhen not greater than the threshold, it is considered that the driver's hand is relatively stationary within the window:
Figure BDA00025188469700000510
step three: extracting characteristic vectors capable of distinguishing the hand-held steering wheel of the driver from the hand-off steering wheel of the driver according to the vehicle acceleration and the acceleration information of the hands of the driver for the window data of which the hands of the driver are in the relatively static state in the step two;
step four: an algorithm for determining the optimal threshold value is constructed, and the specific process is as follows:
dividing the data processed in the third step into three parts at random: the first part is a training data set, the second part is a cross validation set, and the third part is a test set; detecting whether a driver receiving part leaves a steering wheel or not by adopting an anomaly detection algorithm, estimating parameter values of the anomaly detection algorithm by utilizing a training data set, then trying to identify whether data are abnormal or not on a cross validation set by using different thresholds, selecting an optimal threshold according to a test result, and finally testing a detection model on a test set until the detection accuracy of the detection model reaches the actual application requirement.
Step five: and detecting the data acquired in real time by using the trained anomaly detection algorithm, and if the detection result is that the hands of the driver leave the steering wheel, giving an early warning prompt to the driver.
In this embodiment, the sensor is smart phone and smart watch, driver's hand acceleration is the acceleration that smart watch measured, vehicle acceleration is the acceleration that smart phone measured, smart phone is fixed in the carriage, gathers the cell-phone acceleration sensor data of vehicle driving in-process
Figure BDA0002518846970000061
Gyroscope sensor data
Figure BDA0002518846970000062
Figure BDA0002518846970000063
Acceleration of gravity
Figure BDA0002518846970000064
Linear acceleration
Figure BDA0002518846970000065
The intelligent watch is bound at the wrist of a driver and used for collecting watch acceleration sensor data
Figure BDA0002518846970000066
Acceleration of gravity
Figure BDA0002518846970000067
Linear acceleration
Figure BDA0002518846970000068
Converting a smart phone coordinate system and a smart watch coordinate system into a world coordinate system by using a conversion matrix R, wherein R is a numerical matrix of 3x 3; the data sequence is stored in the form of
Figure BDA0002518846970000069
(i ═ 1,2,3 …) in which tiIs the acquisition time.
In this embodiment, the specific process of the third step is as follows:
firstly, extracting a characteristic value from window data of an intelligent watch with a relatively static hand of a driver, and calculating a three-axis mean value wg of gravity acceleration of the intelligent watch through the following formulax,mean、wgy,mean、wgz,meanWhere l is the window length.
Figure BDA00025188469700000610
Figure BDA00025188469700000611
Figure BDA00025188469700000612
The vertical component of the vehicle vibration signal measured by the smart watch is calculated by:
Figure BDA0002518846970000071
in the formula:
Figure BDA0002518846970000072
representing the acceleration of gravity measured by the smart watch,
Figure BDA0002518846970000073
representing the linear acceleration measured by the smart watch, av,watchRepresenting the vertical component of the vehicle vibration signal measured by the smart watch. Then the vertical component of the vehicle vibration signal measured by the smartphone is calculated by:
Figure BDA0002518846970000074
in the formula:
Figure BDA0002518846970000075
representing the acceleration of gravity measured by the smart watch,
Figure BDA0002518846970000076
representing the linear acceleration measured by the smart watch, av,phoneRepresenting the vertical component of the vehicle vibration signal measured by the smart watch. Then, the variance Var (d) of the vertical component of the vibration signal is calculatedv,watch) And Var (a)v,phone) And the ratio of the two
Figure BDA0002518846970000077
Finally, the characteristic vector is obtained
Figure BDA0002518846970000078
In this embodiment, the anomaly detection algorithm in the fourth step isAnd the abnormality detection algorithm based on the multivariate Gaussian model comprises the following specific processes in the fourth step:
and C, randomly selecting 60% of data of the hand-held steering wheel from the data processed in the step three to construct a training data set, forming a cross validation set by 20% of data of the hand-held steering wheel and 50% of data of the hand-off steering wheel, and constructing a test set by using the rest 20% of data of the hand-held steering wheel and 50% of data of the hand-off steering wheel. First, the parameters of a multivariate gaussian model are estimated on a training set using the following equation:
Figure BDA0002518846970000079
in the formula: x is the number of(i)Represents one sample in the training set, and m is the number of samples in the training set. When a test sample is given, its probability density function is calculated by:
Figure BDA00025188469700000710
in the formula: x and μ are vectors of 1 × d; .
And then, trying to identify whether the data are abnormal by using different thresholds on a cross validation set, selecting an optimal threshold according to an F1 value of a test result, and finally testing the detection model on the test set until the detection accuracy of the detection model reaches the actual application requirement.
Finally, the above embodiments are only for illustrating the technical solutions of the present invention and not for limiting, although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that modifications or equivalent substitutions may be made on the technical solutions of the present invention without departing from the spirit and scope of the technical solutions, and all of them should be covered in the protection scope of the present invention.

Claims (4)

1. A detection method for detecting the separation of a steering wheel of a driver in a relatively static state of hands is characterized by comprising the following steps:
the method comprises the following steps: in the running process of a vehicle, acquiring the hand acceleration of a driver and the vehicle acceleration by using a sensor according to a fixed frequency, converting a sensor coordinate system into a world coordinate system, and recording the time t of acquiring the hand acceleration of the driver and the vehicle acceleration every time to form a data sequence;
step two: for driver hand acceleration data set DwAnd a vehicle acceleration data set DpPerforming sliding window processing to obtain a data fragment D ═ D1，D2，D3…Dn}，DnThe expression of (a) is as follows:
Figure FDA0002518846960000011
in the formula: dnFor the nth data segment, the data segment is,
Figure FDA0002518846960000012
the difference between the absolute values of the acceleration of the driver's hand and the acceleration of the vehicle at the first moment in the data segment,
Figure FDA0002518846960000013
is a time point t in the windowiA data value of time, and
Figure FDA0002518846960000014
the expression of (a) is as follows:
Figure FDA0002518846960000015
in the formula:
Figure FDA0002518846960000016
is indicative of the acceleration of the driver's hand,
Figure FDA0002518846960000017
represents vehicle acceleration;
then calculating the average value of all data in the sliding window to obtainRespective data segment DnAverage value D of middle datai，meanThe calculation formula is shown as the following formula:
Figure FDA0002518846960000018
in the formula: di，meanIs the average value of the data of the ith sliding window,
Figure FDA0002518846960000019
represents the ith data in the window, and l represents the length of a sliding window;
whether the driver's hand is moving within each window is determined using the following equation. When D is presenti，meanWhen not greater than the threshold, it is considered that the driver's hand is relatively stationary within the window:
Figure FDA00025188469600000110
step three: extracting characteristic vectors capable of distinguishing the hand-held steering wheel of the driver from the hand-off steering wheel of the driver according to the vehicle acceleration and the acceleration information of the hands of the driver for the window data of which the hands of the driver are in the relatively static state in the step two;
step four: an algorithm for determining the optimal threshold value is constructed, and the specific process is as follows:
dividing the data processed in the third step into three parts at random: the first part is a training data set, the second part is a cross validation set, and the third part is a test set; detecting whether a driver receiving part leaves a steering wheel or not by adopting an anomaly detection algorithm, estimating parameter values of the anomaly detection algorithm by utilizing a training data set, then trying to identify whether data are abnormal or not on a cross validation set by using different thresholds, selecting an optimal threshold according to a test result, and finally testing a detection model on a test set until the detection accuracy of the detection model reaches the actual application requirement.
Step five: and detecting the data acquired in real time by using the trained anomaly detection algorithm, and if the detection result is that the hands of the driver leave the steering wheel, giving an early warning prompt to the driver.
2. The detection method according to claim 1, wherein the specific process of the first step is as follows: the sensor is smart phone and smart watch, driver's hand acceleration is the acceleration that smart watch measured, vehicle acceleration is the acceleration that smart phone measured, smart phone is fixed in the carriage, gathers the cell-phone acceleration sensor data of vehicle driving in-process
Figure FDA0002518846960000021
Gyroscope sensor data
Figure FDA0002518846960000022
Acceleration of gravity
Figure FDA0002518846960000023
Linear acceleration
Figure FDA0002518846960000024
The intelligent watch is bound at the wrist of a driver and used for collecting watch acceleration sensor data
Figure FDA0002518846960000025
Acceleration of gravity
Figure FDA0002518846960000026
Figure FDA0002518846960000027
Linear acceleration
Figure FDA0002518846960000028
Converting a smart phone coordinate system and a smart watch coordinate system into a world coordinate system by using a conversion matrix R, wherein R is a numerical matrix of 3x 3; the data sequence is stored in the form of
Figure FDA0002518846960000029
Figure FDA00025188469600000210
Where ti is the acquisition time.
3. The detection method according to claim 1 or 2, wherein the specific process of the third step is as follows:
firstly, extracting a characteristic value from window data of an intelligent watch with a relatively static hand of a driver, and calculating a three-axis mean value wg of gravity acceleration of the intelligent watch through the following formulax，mean、wgy，mean、wgz，meanWherein l is the window length;
Figure FDA00025188469600000211
Figure FDA00025188469600000212
Figure FDA00025188469600000213
the vertical component of the vehicle vibration signal measured by the smart watch is calculated by:
Figure FDA00025188469600000214
in the formula:
Figure FDA00025188469600000215
representing the acceleration of gravity measured by the smart watch,
Figure FDA00025188469600000216
representing the linear acceleration measured by the smart watch, av，watchIndicating what the smart watch measuresThe vertical component of the resulting vehicle vibration signal; then the vertical component of the vehicle vibration signal measured by the smartphone is calculated by:
Figure FDA0002518846960000031
in the formula:
Figure FDA0002518846960000032
representing the acceleration of gravity measured by the smart watch,
Figure FDA0002518846960000033
representing the linear acceleration measured by the smart watch, av，phoneRepresenting a vertical component of the vehicle vibration signal measured by the smart watch; then, the variance Var (a) of the vertical component of the vibration signal is calculatedv，watch) And Var (a)v，phone) And the ratio of the two
Figure FDA0002518846960000034
Finally, the characteristic vector is obtained
Figure FDA0002518846960000035
4. The detection method according to claim 1, wherein the anomaly detection algorithm of the fourth step is an anomaly detection algorithm based on a multivariate Gaussian model, and the specific process of the fourth step is as follows:
randomly selecting 60% of data of the hand-held steering wheel from the data processed in the third step to construct a training data set, wherein 20% of data of the hand-held steering wheel and 50% of data of the hand leaving the steering wheel form a cross validation set, and constructing a test set by using the remaining 20% of data of the hand-held steering wheel and 50% of data of the hand leaving the steering wheel; first, the parameters of a multivariate gaussian model are estimated on a training set using the following equation:
Figure FDA0002518846960000036
in the formula: x is the number of(i)Representing a sample in the training set, wherein m is the number of samples in the training set; when a test sample is given, its probability density function is calculated by:
Figure FDA0002518846960000037
in the formula: x and μ are vectors of 1 × d;
and then, trying to identify whether the data are abnormal by using different thresholds on a cross validation set, selecting an optimal threshold according to an F1 value of a test result, and finally testing the detection model on the test set until the detection accuracy of the detection model reaches the actual application requirement.
CN202010485078.2A 2020-06-01 2020-06-01 Detection method for separation of driver from steering wheel in relative static state of hands Active CN111717210B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010485078.2A CN111717210B (en) 2020-06-01 2020-06-01 Detection method for separation of driver from steering wheel in relative static state of hands

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010485078.2A CN111717210B (en) 2020-06-01 2020-06-01 Detection method for separation of driver from steering wheel in relative static state of hands

Publications (2)

Publication Number Publication Date
CN111717210A true CN111717210A (en) 2020-09-29
CN111717210B CN111717210B (en) 2022-11-11

Family

ID=72565684

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010485078.2A Active CN111717210B (en) 2020-06-01 2020-06-01 Detection method for separation of driver from steering wheel in relative static state of hands

Country Status (1)

Country Link
CN (1) CN111717210B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112706710A (en) * 2021-01-27 2021-04-27 南京工程学院 Intelligent recognition and warning system and method for safe driving of motor vehicle
CN114468494A (en) * 2022-02-27 2022-05-13 重庆长安汽车股份有限公司 Automobile-used intelligent wrist-watch and mounting structure thereof

Citations (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102010013243A1 (en) * 2010-03-29 2011-09-29 Audi Ag Method for determining information relating to the direction of vision of a driver and the position of the driver's hands with respect to the steering wheel in a motor vehicle and motor vehicle
US20110254956A1 (en) * 2010-04-19 2011-10-20 Denso Corporation Driving assistance apparatus
CN104276080A (en) * 2014-10-16 2015-01-14 北京航空航天大学 Bus driver hand-off-steering-wheel detection warning system and warning method
US20160325779A1 (en) * 2015-05-05 2016-11-10 Ford Global Technologies, Llc Hands-off steering wheel governed by pedestrian detection
CN107082072A (en) * 2016-02-12 2017-08-22 通用汽车环球科技运作有限责任公司 By means of the detection enhancing of slipping out of the hand of composite signal
US20180003593A1 (en) * 2016-06-30 2018-01-04 Massachusetts lnstitute of Technology Applying motion sensor data to wheel imbalance detection, tire pressure monitoring, and/or tread depth measurement
CN107662613A (en) * 2017-09-27 2018-02-06 西北工业大学 A kind of extreme driving behavior recognition methods and system based on mobile intelligent perception
CN107685729A (en) * 2016-08-04 2018-02-13 丰田自动车株式会社 Controller of vehicle
CN108528444A (en) * 2017-03-06 2018-09-14 丰田自动车株式会社 Warning system and drive assist system
CN108665680A (en) * 2018-05-10 2018-10-16 武汉理工大学 A kind of turnpike driving people's fatigue monitoring early warning system based on Intelligent bracelet
CN108773377A (en) * 2018-07-17 2018-11-09 重庆大学 A kind of automobile fuel consumption real-time estimation method and device based on mobile terminal
CN109034111A (en) * 2018-08-17 2018-12-18 北京航空航天大学 A kind of driver's hand based on deep learning is from steering wheel detection method and system
CN109927731A (en) * 2017-12-15 2019-06-25 蔚来汽车有限公司 Driver lets go detection method, device, controller and storage medium
US20190209050A1 (en) * 2016-08-31 2019-07-11 Goertek Inc. Wearable device and method of using same to monitor motion state
CN110135327A (en) * 2019-05-11 2019-08-16 东南大学 A kind of driving behavior recognition methods based on multi-region feature learning model
CN110298257A (en) * 2019-06-04 2019-10-01 东南大学 A kind of driving behavior recognition methods based on human body multiple location feature
CN110309723A (en) * 2019-06-04 2019-10-08 东南大学 A kind of driving behavior recognition methods based on characteristics of human body's disaggregated classification
CN110386060A (en) * 2018-04-20 2019-10-29 比亚迪股份有限公司 Steering wheel both hands are detached from based reminding method, device, system and the vehicle with it
DE102019217280A1 (en) * 2018-11-08 2020-05-14 Mando Corporation DRIVER ASSISTANCE DEVICE, METHOD FOR THEIR CONTROL, AND DRIVER ASSISTANCE SYSTEM

Patent Citations (19)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE102010013243A1 (en) * 2010-03-29 2011-09-29 Audi Ag Method for determining information relating to the direction of vision of a driver and the position of the driver's hands with respect to the steering wheel in a motor vehicle and motor vehicle
US20110254956A1 (en) * 2010-04-19 2011-10-20 Denso Corporation Driving assistance apparatus
CN104276080A (en) * 2014-10-16 2015-01-14 北京航空航天大学 Bus driver hand-off-steering-wheel detection warning system and warning method
US20160325779A1 (en) * 2015-05-05 2016-11-10 Ford Global Technologies, Llc Hands-off steering wheel governed by pedestrian detection
CN107082072A (en) * 2016-02-12 2017-08-22 通用汽车环球科技运作有限责任公司 By means of the detection enhancing of slipping out of the hand of composite signal
US20180003593A1 (en) * 2016-06-30 2018-01-04 Massachusetts lnstitute of Technology Applying motion sensor data to wheel imbalance detection, tire pressure monitoring, and/or tread depth measurement
CN107685729A (en) * 2016-08-04 2018-02-13 丰田自动车株式会社 Controller of vehicle
US20190209050A1 (en) * 2016-08-31 2019-07-11 Goertek Inc. Wearable device and method of using same to monitor motion state
CN108528444A (en) * 2017-03-06 2018-09-14 丰田自动车株式会社 Warning system and drive assist system
CN107662613A (en) * 2017-09-27 2018-02-06 西北工业大学 A kind of extreme driving behavior recognition methods and system based on mobile intelligent perception
CN109927731A (en) * 2017-12-15 2019-06-25 蔚来汽车有限公司 Driver lets go detection method, device, controller and storage medium
CN110386060A (en) * 2018-04-20 2019-10-29 比亚迪股份有限公司 Steering wheel both hands are detached from based reminding method, device, system and the vehicle with it
CN108665680A (en) * 2018-05-10 2018-10-16 武汉理工大学 A kind of turnpike driving people's fatigue monitoring early warning system based on Intelligent bracelet
CN108773377A (en) * 2018-07-17 2018-11-09 重庆大学 A kind of automobile fuel consumption real-time estimation method and device based on mobile terminal
CN109034111A (en) * 2018-08-17 2018-12-18 北京航空航天大学 A kind of driver's hand based on deep learning is from steering wheel detection method and system
DE102019217280A1 (en) * 2018-11-08 2020-05-14 Mando Corporation DRIVER ASSISTANCE DEVICE, METHOD FOR THEIR CONTROL, AND DRIVER ASSISTANCE SYSTEM
CN110135327A (en) * 2019-05-11 2019-08-16 东南大学 A kind of driving behavior recognition methods based on multi-region feature learning model
CN110298257A (en) * 2019-06-04 2019-10-01 东南大学 A kind of driving behavior recognition methods based on human body multiple location feature
CN110309723A (en) * 2019-06-04 2019-10-08 东南大学 A kind of driving behavior recognition methods based on characteristics of human body's disaggregated classification

Non-Patent Citations (6)

* Cited by examiner, † Cited by third party
Title
SUN DIHUA等: "Recognition of Fatigue Driving Based on Steering Operation Using Wearable Smart Watch", 《GREEN, SMART AND CONNECTED TRANSPORTATION SYSTEMS 》 *
周聪颖: "基于可穿戴设备的驾驶行为检测技术研究", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 *
张凡等: "基于身体动作的分心驾驶预警方法研究", 《汽车实用技术》 *
殷晓玲等: "基于智能手机内置传感器的人体运动状态识别", 《通信学报》 *
王二虎: "基于手机传感器的人体运动姿态识别研究", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 *
范菁等: "基于智能手机的车辆行为实时判别与渐进矫正方法研究", 《计算机科学》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112706710A (en) * 2021-01-27 2021-04-27 南京工程学院 Intelligent recognition and warning system and method for safe driving of motor vehicle
CN114468494A (en) * 2022-02-27 2022-05-13 重庆长安汽车股份有限公司 Automobile-used intelligent wrist-watch and mounting structure thereof
CN114468494B (en) * 2022-02-27 2023-06-06 重庆长安汽车股份有限公司 Automobile-used intelligent wrist-watch and mounting structure thereof

Also Published As

Publication number Publication date
CN111717210B (en) 2022-11-11

Similar Documents

Publication Publication Date Title
JP5444898B2 (en) Status detection device, status detection method and program
US20160016590A1 (en) Method for detecting driving events of a vehicle based on a smartphone
EP3105085B1 (en) Vehicle event assessment
JP5392686B2 (en) Driving support apparatus and method
CN108682119B (en) Driver fatigue state detection method based on smart phone and smart watch
CN111717210B (en) Detection method for separation of driver from steering wheel in relative static state of hands
CN105564436A (en) Advanced driver assistance system
KR101473957B1 (en) Apparatus and method for determining insurance premium based on driving pattern recognition
Zhao et al. Join driving: A smart phone-based driving behavior evaluation system
JP2008507721A (en) Operation monitoring system and method
JP2016057836A (en) Mobile object analysis system and mobile object directional axis estimation method
JP2019112049A (en) Method for recognizing driving style of driver of land vehicle and corresponding device
CN107358248B (en) Method for improving falling detection system precision
CN107662613B (en) A kind of extreme driving behavior recognition methods and system based on mobile intelligent perception
CN112848816B (en) Tire pressure detection method, tire pressure detection equipment, storage medium and tire pressure detection device based on pressure sensor
Engelbrecht et al. Performance comparison of dynamic time warping (DTW) and a maximum likelihood (ML) classifier in measuring driver behavior with smartphones
CN104182037A (en) Gesture recognition method based on coordinate conversion
Saiprasert et al. Detecting driving events using smartphone
Chaovalit et al. A method for driving event detection using SAX on smartphone sensors
US9846174B2 (en) Computer-implemented methods and computer systems/machines for identifying dependent and vehicle independent states
EP3382570A1 (en) Method for characterizing driving events of a vehicle based on an accelerometer sensor
CN109978026B (en) Elevator position detection method and system based on LSTM network
CN112129290A (en) System and method for monitoring riding equipment
CN107539038B (en) Vehicle tire pressure state monitoring method and device
Vavouranakis et al. Recognizing driving behaviour using smartphones

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