阅读说明：本技术 一种车辆转弯和变道场景下驾驶员分神驾驶检测和预警系统及方法 (Driver distraction driving detection and early warning system and method under vehicle turning and lane changing scenes ) 是由 叶智锐 陈天 许跃如 刘民 王超 于 2019-10-28 设计创作，主要内容包括：本发明公开了一种车辆转弯和变道场景下驾驶员分神驾驶检测和预警系统及方法,所述系统包括数据采集单元、数据处理单元、判定单元以及预警单元。数据采集单元实时收集车辆以及驾驶员状态信息,据此预先建立车辆转弯/变道情景判定标准；以及预先通过聚类算法输出驾驶员三种状态的指标特征范围,并将其输入数据处理单元以供与收集到的实时数据对比。在车辆实际行驶的过程中进入转弯/变道的情景下,数据采集单元实时收集车辆以及驾驶员状态信息,并将其输入到数据处理单元,判定驾驶员是否存在分神驾驶状态。最终由预警单元发出两级预警信息。本发明能够准确地识别车辆进入转弯/变道的情景,并全面地对驾驶员分神状态进行准确判断和预警。(The invention discloses a detection and early warning system and a method for distracted driving of a driver in vehicle turning and lane changing scenes. The data acquisition unit collects the state information of the vehicle and the driver in real time, and accordingly a vehicle turning/lane changing situation judgment standard is established in advance; and outputting the index characteristic ranges of the three states of the driver through a clustering algorithm in advance, and inputting the index characteristic ranges into a data processing unit for comparison with the collected real-time data. Under the circumstance that the vehicle enters a turning/lane changing situation in the actual running process, the data acquisition unit collects the vehicle and the state information of the driver in real time, inputs the state information into the data processing unit and judges whether the driver has a distracted driving state or not. And finally, the early warning unit sends out two-stage early warning information. The method and the system can accurately identify the situation that the vehicle enters the turning/lane changing and comprehensively and accurately judge and early warn the distraction state of the driver.)

1. A driver distraction driving detection and early warning method under vehicle turning and lane changing scenes is characterized in that: the method comprises the following steps:

step 1, collecting vehicle index data and determining whether a vehicle enters a turning/lane changing situation;

step 2, building a vehicle and driver behavior data set by acquiring vehicle and driver state data after the vehicle enters a turning/lane changing scene in advance;

step 3, performing FCM clustering analysis on the collected vehicle driving state data and the driver driving state data to obtain a driver distraction state and a corresponding early warning level;

step 4, collecting vehicle index data and judging whether the vehicle enters a turning/lane changing state or not in the normal running process of the vehicle; if the vehicle enters a turning/lane changing state, entering the step 5, otherwise, continuously acquiring data;

step 5, after the vehicle is determined to enter a turning/lane changing state, acquiring vehicle driving state data and driver driving state index data; comparing the collected data with the characteristic range of the early warning hierarchy output by clustering analysis to determine the driving state of the vehicle driver;

and 6, judging whether to carry out driving distraction early warning or not according to the driving state of the vehicle driver determined in the step 5, and taking corresponding early warning measures according to early warning levels.

2. The driver distraction detection and early warning method in vehicle turning and lane changing scenes according to claim 1, wherein: in the step 1, the vehicle index data comprises vehicle steering lamp flicker, vehicle running track change and vehicle steering wheel rotation; when the vehicle steering lamp flickers, the vehicle running track changes, and the vehicle steering wheel rotates, the situation that the vehicle enters the turning/lane changing situation is judged.

3. The driver distraction detection and early warning method in vehicle turning and lane changing scenes according to claim 1 or 2, characterized in that: in the step 2, a vehicle and driver behavior data set is established by acquiring vehicle and driver state data after the vehicle enters a turning/lane changing scene in advance; the method comprises the following specific steps:

2-1, after judging that the vehicle enters the turning/lane changing scene, in a time period T₁Collecting vehicle running state data and driving state data of a driver;

the driving state data of the vehicle comprises an angle change range CT of the vehicle driving track direction and a rotation angle change range CD of a vehicle steering wheel;

the driving state data of the driver comprises a change range GP of the position of a fixation point of the driver and a change range GF of the grip strength of the driver to a steering wheel;

and 2-2, acquiring vehicle and driver state data of n time periods, grouping the vehicle driving state data and the driver driving state data in each time period, and establishing a vehicle and driver behavior data set.

4. The driver distraction detection and early warning method in vehicle turning and lane changing scenes according to claim 3, wherein: in the step 3, FCM clustering analysis is carried out on the collected vehicle driving state data and the driver driving state data to obtain a driver distraction state and a corresponding early warning level; the method comprises the following specific steps:

3-1, using data indexes in the vehicle and driver behavior data set as input data, performing clustering estimation on the distraction state of the driver by adopting an FCM clustering method in combination with the behavior adopted by the driver, and dividing a clustering result into three early warning levels;

3-2, after the cluster analysis is finished, outputting an angle change range CT of the vehicle driving track direction, a rotation angle change range CD of a vehicle steering wheel, a change range GP of the driver gazing point position and a grip change range GF of the driver to the steering wheel, wherein the four indexes correspond to the characteristic ranges of three early warning levels;

3-3, taking the four indexes output in the step 3-2 as data processing and judging bases, and judging the distracted driving state of the driver; real-time vehicle and driver data collected in the normal driving process of the vehicle are compared with the real-time vehicle and the driver data to determine whether the driver has a distracted driving state and whether early warning is needed.

5. The method for detecting and warning distraction of the driver in the scene of vehicle turning and lane changing as claimed in claim 4, wherein: the three early warning levels in the step 3-1 are as follows:

early warning in class I: the driver does not make corresponding operation behavior when the vehicle turns/changes lanes;

II type early warning: the driver takes corresponding operation when the vehicle turns/changes lanes, but a distraction state exists;

no early warning is performed: the driver takes corresponding operation when the vehicle turns/changes lanes, and the distraction state does not exist.

6. The method for detecting and warning distraction of the driver in the scene of vehicle turning and lane changing according to claim 5, wherein: in the step 5, after the vehicle is determined to enter a turning/lane changing state, vehicle running state data and driver driving state index data are collected; comparing the collected data with the characteristic range of the early warning hierarchy output by clustering analysis to determine the driving state of the vehicle driver; the method comprises the following specific steps:

if the grip strength of the driver does not exceed the grip strength change range GF, the fixation point of the driver does not exceed the change range GP of the position of the fixation point, the vehicle track does not exceed the angle change range CT of the direction of the driving track, and the change of the steering wheel angle does not exceed the rotation angle change range CD, which indicates that the driver makes corresponding operation behavior when the vehicle turns/changes lanes and the distraction state does not exist;

if the grip strength of the driver does not exceed the grip strength change range GF, the fixation point of the driver does not exceed the change range GP of the position of the fixation point, and meanwhile, the angle change range CT that the vehicle track exceeds the direction of the driving track and the change of the steering wheel angle exceed the change range CD of the rotation angle indicate that the driver does not make corresponding operation behaviors when the vehicle turns/changes lanes;

if the grip strength of the driver exceeds the grip strength change range GF, the gaze point of the driver does not exceed the change range GP of the position of the gaze point, the vehicle track does not exceed the angle change range CT of the direction of the driving track, and the change of the steering wheel angle does not exceed the rotation angle change range CD, indicating that the driver takes corresponding operation when the vehicle turns/changes lanes, but the distraction state exists.

7. The method for detecting and warning distraction of the driver in the scene of vehicle turning and lane changing according to claim 6, wherein: step 6, judging whether to carry out driving distraction early warning according to the driving state of the vehicle driver determined in step 5, and taking corresponding early warning measures according to early warning levels; the method comprises the following specific steps:

if the driver of the vehicle takes corresponding operation after the vehicle enters the turning/lane changing scene and the distraction state does not exist, no early warning is carried out, and data acquisition is restarted;

if the driver of the vehicle does not take corresponding operation behavior after the vehicle enters the turning/lane changing scene, starting the class I early warning;

if the driver of the vehicle takes corresponding operation after the vehicle enters the turning/lane changing scene, but a distraction state exists during the operation, the II-type early warning is started.

8. The driver distraction detection and early warning method in vehicle turning and lane changing scenes according to claim 7, wherein: starting class I early warning, and reminding a driver by flashing a vehicle-mounted indicating lamp; and starting II-type early warning, and reminding a driver through vehicle-mounted indicating lamp flashing and voice prompt.

9. The utility model provides a driver's distraction driving detection and early warning system under vehicle turn and lane change scene which characterized in that: the system comprises a data acquisition unit, a data processing unit, a judgment unit and an early warning unit;

the data acquisition unit comprises a vehicle steering lamp flicker detection device, a vehicle steering wheel rotation detection device, a vehicle running camera, a non-invasive eye tracker and a grip strength sensor; the system is used for collecting the state information of a driver and a vehicle in the driving process and transmitting the state information to the data processing unit;

the data processing unit adopts a related vehicle-mounted processor terminal and is used for receiving the data transmitted by the data acquisition unit and transmitting the data to a corresponding judgment unit; the early warning system is used for comparing the pre-input index range of the early warning level with the state information of the vehicle and the driver in the actual driving process;

the determination unit includes a vehicle state determination unit and a driver state determination unit for determining whether the vehicle enters a turning/lane change driving state and whether the driver has a distracted driving state; the judging unit adopts a relevant vehicle-mounted processor terminal, is coupled with the data processing unit and is used for receiving the data transmitted by the data processing unit;

the early warning unit adopts relevant on-vehicle stereo set and on-vehicle pilot lamp for carry out corresponding early warning to driver's distraction driving state, including pilot lamp scintillation, voice prompt.

10. The system of claim 9, wherein the driver distraction detection and warning system is configured to detect the driver distraction in the vehicle turning and lane changing scene: the vehicle steering lamp flicker detection device is assembled in a vehicle steering system and is used for collecting vehicle steering lamp flicker data in real time; the vehicle steering wheel rotation detection device is assembled in a vehicle direction control system and is used for acquiring the rotation angle data of the vehicle steering wheel in real time; the vehicle running cameras are assembled on rearview mirrors on two sides of a vehicle and used for acquiring vehicle direction and track change data in real time; the non-invasive eye tracker is assembled above a dial plate of a vehicle driving position instrument, and a camera of the non-invasive eye tracker is aligned to the eye part of the face of a driver and is used for acquiring position data of an eye gazing point and change data of the eye gazing point position in the process of driving the vehicle of the driver in real time; the grip sensor is assembled on a vehicle steering wheel and is used for acquiring grip data of a driver's operating hand on the steering wheel in real time in the process of driving the vehicle.

Technical Field

The invention belongs to the field of active safety prevention and control of vehicle drivers, and particularly relates to a driver distraction driving detection and early warning system and method in vehicle turning and lane changing scenes.

Background

In the process of continuous development of the modern automobile industry and continuous emergence of emerging technologies, the safety control problems of vehicle driving and vehicle drivers are always the first problems concerned by people. Statistical data of traffic accidents show that human factors of drivers, such as distraction driving and other behaviors are one of main reasons for inducing the traffic accidents; while a significant portion of traffic accidents occur in vehicle turning and lane changing scenarios. Reducing distraction under special situations such as vehicle turning and lane changing during the driving process of a driver is one of important problems to be solved in the field of active safety control of vehicle drivers.

The expression forms of the distracted driving behaviors of the driver tend to be diversified, and the distracted driving behaviors have obvious differences under different driving situations. At present, the existing relevant research for detecting and early warning the distracted driving state of the driver mainly focuses on detecting and judging the distracted driving state of the driver by considering a predefined judgment index in the normal driving process of a vehicle. The detection of driver distraction driving under a specific scene in the driving process of the vehicle is less researched, and the possible difference of the definition standard of the driver distraction state under the specific scene in the driving process of the vehicle is not considered.

Disclosure of Invention

The purpose of the invention is as follows: aiming at the problems, the invention provides a driving system and a method for detecting and early warning the distraction driving behavior of a driver when a vehicle turns and changes lanes, which solve the problem that the difference possibly existing in the definition standard of the distraction state of the driver under the specific scene of turning and changing lanes of the vehicle is not considered in the existing method and system for detecting the distraction driving state of the driver of the vehicle; and three-level early warning information is provided, and detection and early warning are comprehensively carried out on the distracted driving state of the vehicle driver.

The technical scheme is as follows: in order to realize the purpose of the invention, the technical scheme adopted by the invention is as follows:

a driver distraction driving detection and early warning system under vehicle turning and lane changing scenes comprises a data acquisition unit, a data processing unit, a judgment unit and an early warning unit.

The data acquisition unit comprises a vehicle steering lamp flicker detection device, a vehicle steering wheel rotation detection device, a vehicle running camera, a non-invasive eye tracker and a grip strength sensor; the data acquisition unit is used for acquiring the state information of a driver and a vehicle in the driving process and transmitting the state information to the data processing unit.

The data processing unit adopts a relevant vehicle-mounted processor terminal and is used for receiving data transmitted by the data acquisition unit, namely the vehicle steering lamp flicker detection device, the vehicle steering wheel rotation detection device, the vehicle running camera, the non-invasive eye tracker and the grip strength sensor, and transmitting the data to a corresponding judgment unit. The data processing unit is used for comparing the pre-input index range of the early warning level with the state information of the vehicle and the driver in the actual driving process.

The determination unit includes a vehicle state determination unit and a driver state determination unit for determining whether the vehicle enters a turning/lane change running state and whether the driver has a distracted driving state. The judging unit adopts a relevant vehicle-mounted processor terminal, is coupled with the data processing unit and is used for receiving the data transmitted by the data processing unit and judging whether the vehicle enters a turning/lane changing scene or not and whether the driver has a state of distraction driving or not.

The early warning unit is used for carrying out corresponding early warning on the distracted driving state of the driver, such as flashing of an indicator light and voice prompt. The early warning unit adopts relevant on-vehicle stereo set and on-vehicle pilot lamp for carry out pilot lamp scintillation, voice prompt early warning to I, II level driver distraction state.

Further, the vehicle steering lamp flicker detection device is assembled in a vehicle steering system and used for collecting vehicle steering lamp flicker data in real time; the vehicle steering wheel rotation detection device is assembled in a vehicle direction control system and is used for acquiring the rotation angle data of the vehicle steering wheel in real time; the vehicle running cameras are assembled on rearview mirrors on two sides of a vehicle and used for acquiring vehicle direction and track change data in real time; the non-invasive eye tracker is assembled above a dial plate of a vehicle driving position instrument, and a camera of the non-invasive eye tracker is aligned to the eye part of the face of a driver and is used for acquiring position data of an eye gazing point and change data of the eye gazing point position in the process of driving the vehicle of the driver in real time; the grip sensor is assembled on a vehicle steering wheel and is used for acquiring grip data of a driver's operating hand on the steering wheel in real time in the process of driving the vehicle.

A driver distraction driving detection and early warning method under vehicle turning and lane changing scenes comprises the following steps:

step 1, collecting vehicle index data and determining whether a vehicle enters a turning/lane changing situation;

step 2, building a vehicle and driver behavior data set by acquiring vehicle and driver state data after the vehicle enters a turning/lane changing scene in advance;

step 3, performing FCM clustering analysis on the collected vehicle driving state data and the driver driving state data to obtain a driver distraction state and a corresponding early warning level;

step 4, collecting vehicle index data and judging whether the vehicle enters a turning/lane changing state or not in the normal running process of the vehicle; if the vehicle enters a turning/lane changing state, entering the step 5, otherwise, continuously acquiring data;

step 5, after the vehicle is determined to enter a turning/lane changing state, acquiring vehicle driving state data and driver driving state index data; comparing the collected data with the characteristic range of the early warning hierarchy output by clustering analysis to determine the driving state of the vehicle driver;

and 6, judging whether to carry out driving distraction early warning or not according to the driving state of the vehicle driver determined in the step 5, and taking corresponding early warning measures according to early warning levels.

Further, in the step 1, the vehicle index data includes vehicle turn light flicker, vehicle track change and vehicle steering wheel rotation; when the vehicle steering lamp flickers, the vehicle running track changes, and the vehicle steering wheel rotates, the situation that the vehicle enters the turning/lane changing situation is judged.

Further, in the step 2, a vehicle and driver behavior data set is established by acquiring vehicle and driver state data after the vehicle enters a turning/lane changing scene in advance; the method comprises the following specific steps:

2-1, after judging that the vehicle enters the turning/lane changing scene, in a time period T₁Collecting vehicle running state data and driving state data of a driver; the driving state data of the vehicle comprises an angle change range CT of the vehicle driving track direction and a rotation angle change range CD of a vehicle steering wheel; the driving state data of the driver comprises a change range GP of the position of a fixation point of the driver and a change range GF of the grip strength of the driver to a steering wheel;

and 2-2, acquiring vehicle and driver state data of n time periods, grouping the vehicle driving state data and the driver driving state data in each time period, and establishing a vehicle and driver behavior data set.

Further, in the step 3, FCM cluster analysis is performed on the collected vehicle driving state data and the driver driving state data to obtain the driver distraction state and the corresponding early warning level; the method comprises the following specific steps:

3-1, using data indexes in the vehicle and driver behavior data set as input data, performing clustering estimation on the distraction state of the driver by adopting an FCM clustering method in combination with the behavior adopted by the driver, and dividing a clustering result into three early warning levels;

3-2, after the cluster analysis is finished, outputting an angle change range CT of the vehicle driving track direction, a rotation angle change range CD of a vehicle steering wheel, a change range GP of the driver gazing point position and a grip change range GF of the driver to the steering wheel, wherein the four indexes correspond to the characteristic ranges of three early warning levels;

3-3, taking the four indexes output in the step 3-2 as data processing and judging bases, and judging the distracted driving state of the driver; real-time vehicle and driver data collected in the normal driving process of the vehicle are compared with the real-time vehicle and the driver data to determine whether the driver has a distracted driving state and whether early warning is needed.

Further, the three early warning levels in the step 3-1 are as follows:

early warning in class I: the driver does not make corresponding operation behavior when the vehicle turns/changes lanes;

II type early warning: the driver takes corresponding operation when the vehicle turns/changes lanes, but a distraction state exists;

no early warning is performed: the driver takes corresponding operation when the vehicle turns/changes lanes, and the distraction state does not exist.

Further, in the step 5, after the vehicle is determined to enter the turning/lane changing state, vehicle driving state data and driver driving state index data are collected; comparing the collected data with the characteristic range of the early warning hierarchy output by clustering analysis to determine the driving state of the vehicle driver; the method comprises the following specific steps:

if the grip strength of the driver does not exceed the grip strength change range GF, the gaze point of the driver does not exceed the change range GP of the position of the gaze point, the vehicle track does not exceed the angle change range CT of the direction of the driving track, and the change of the steering wheel angle does not exceed the rotation angle change range CD, indicating that the driver makes corresponding operation behavior when the vehicle turns/changes lanes and no distraction state exists.

If the grip strength of the driver does not exceed the grip strength change range GF, the gaze point of the driver does not exceed the change range GP of the position of the gaze point, and simultaneously the angle change range CT that the vehicle track exceeds the direction of the driving track and the change of the steering wheel angle exceeds the change range CD of the rotation angle, the driver does not make corresponding operation behaviors when the vehicle turns/changes lanes.

If the grip strength of the driver exceeds the grip strength change range GF, the gaze point of the driver does not exceed the change range GP of the position of the gaze point, the vehicle track does not exceed the angle change range CT of the direction of the driving track, and the change of the steering wheel angle does not exceed the rotation angle change range CD, indicating that the driver takes corresponding operation when the vehicle turns/changes lanes, but the distraction state exists.

Further, step 6, according to the driving state of the vehicle driver determined in step 5, judging whether to perform driving distraction early warning, and taking corresponding early warning measures according to early warning levels; the method comprises the following specific steps:

if the driver of the vehicle takes corresponding operation after the vehicle enters the turning/lane changing scene and the distraction state does not exist, no early warning is carried out, and data acquisition is restarted;

if the driver of the vehicle does not take corresponding operation behavior after the vehicle enters the turning/lane changing scene, starting the class I early warning;

if the driver of the vehicle takes corresponding operation after the vehicle enters the turning/lane changing scene, but a distraction state exists during the operation, the II-type early warning is started.

Further, starting class I early warning, and reminding a driver by flashing a vehicle-mounted indicator light; and starting II-type early warning, and reminding a driver through vehicle-mounted indicating lamp flashing and voice prompt.

Has the advantages that: compared with the prior art, the technical scheme of the invention has the following beneficial technical effects:

at present, the existing relevant research for detecting and early warning the distracted driving state of a driver mainly focuses on detecting and judging the distracted driving state of the driver through artificially predefined judgment indexes in the normal driving process of a vehicle. The detection of driver distraction driving under a specific scene in the driving process of the vehicle is less researched, and the possible difference of the definition standard of the driver distraction state under the specific scene in the driving process of the vehicle is not considered. The technical scheme of the invention solves the problem that the difference possibly exists in the definition standard of the distraction state of the driver under the specific scene of vehicle turning and lane changing is not considered in the traditional method and system for detecting the distraction driving state of the vehicle driver; and three-level early warning information is provided, and detection and early warning are comprehensively carried out on the distracted driving state of the vehicle driver.

Drawings

FIG. 1 is a flowchart of the overall operation of the detection and early warning method of the present invention;

fig. 2 is a flowchart of the method operation of determining whether the vehicle has entered a turning/lane-changing state in the present invention.

Detailed Description

The technical solution of the present invention is further described below with reference to the accompanying drawings and examples.

The invention provides a driving system for detecting and early warning distracted driving behaviors of a driver during vehicle turning and lane changing.

The data acquisition unit comprises a vehicle steering lamp flicker detection device, a vehicle steering wheel rotation detection device, a vehicle running camera, a non-invasive eye tracker and a grip strength sensor; the data acquisition unit is used for acquiring the state information of a driver and a vehicle in the driving process and transmitting the state information to the data processing unit.

The data processing unit adopts a relevant vehicle-mounted processor terminal and is used for receiving data transmitted by the data acquisition unit, namely the vehicle steering lamp flicker detection device, the vehicle steering wheel rotation detection device, the vehicle running camera, the non-invasive eye tracker and the grip strength sensor, and transmitting the data to a corresponding judgment unit.

The data processing unit inputs in advance an angle change range CT based on the vehicle driving track direction, a rotation angle change range CD of a vehicle steering wheel, a change range GP of the driver gazing point position, a result of FCM cluster analysis of the driver distraction driving state based on four indexes of a grip change range GF of the driver to the steering wheel and index ranges of three early warning levels so as to compare the information with the vehicle and the driver state in the actual driving process.

The determination unit includes a vehicle state determination unit and a driver state determination unit for determining whether the vehicle enters a turning/lane change running state and whether the driver has a distracted driving state. The judging unit adopts a relevant vehicle-mounted processor terminal and is coupled with the data processing unit and used for receiving the data transmitted by the data processing unit and judging whether the vehicle enters a turning/lane changing scene or not and whether the driver has a state of distraction driving or not. The work flow of the vehicle state determination unit is shown in fig. 2.

The early warning unit is used for carrying out corresponding early warning on the distracted driving state of the driver, such as flashing of an indicator light and voice prompt. The early warning unit adopts relevant on-vehicle stereo set and on-vehicle pilot lamp, carries out pilot lamp scintillation, voice prompt early warning to I, II level driver distraction state.

The vehicle steering lamp flicker detection device is coupled with a vehicle steering system and used for collecting vehicle steering lamp flicker condition data in real time and providing data basis for judging whether a vehicle enters a turning/lane changing situation.

The vehicle steering wheel rotation detection device is coupled with a vehicle steering control system, is assembled with a vehicle steering wheel facility and is used for collecting the data of the rotation angle of the vehicle steering wheel in real time and providing data basis for judging whether the vehicle enters a turning/lane changing scene and whether a driver has a distracted driving state.

The vehicle running camera and rearview mirror facilities on two sides of the vehicle are assembled together and used for collecting vehicle direction and track change data in real time and providing data basis for judging whether the vehicle enters a turning/lane changing scene and whether a driver has a distracted driving state.

The non-invasive eye tracker is assembled above a dial plate facility of a vehicle driving position instrument, and a camera of the non-invasive eye tracker is aligned to the eye position of the face of a driver and used for acquiring position data of an eye gazing point and change data of the eye gazing point position of the driver in the vehicle driving process in real time and providing data basis for judging whether a distracted driving state exists in the driver.

The grip force sensor is coupled with a vehicle direction control system, is assembled with a vehicle steering wheel facility, is used for real-time grip force data of a steering wheel of a manipulator of a driver in the process of driving the vehicle, and provides data basis for judging whether the driver has a distracted driving state.

A driver distraction driving detection and early warning method under the scene of vehicle turning and lane changing is disclosed, the working flow is shown in figure 1, and the method comprises the following steps:

step 1, collecting vehicle index data and determining whether a vehicle enters a turning/lane changing situation; the vehicle index data comprises vehicle steering lamp flicker, vehicle track change and vehicle steering wheel rotation. As shown in fig. 2, when the vehicle turn signal blinks and the vehicle trajectory changes while the vehicle steering wheel is turned, it is determined that the vehicle enters a turning/lane changing situation.

Step 2, building a vehicle and driver behavior data set by acquiring vehicle and driver state data after the vehicle enters a turning/lane changing scene in advance; the method comprises the following specific steps:

2-1, after judging that the vehicle enters the turning/lane changing scene, in a time period T₁Collecting vehicle running state data and driving state data of a driver; the driving state data of the vehicle comprises an angle change range CT of the vehicle driving track direction and a rotation angle change range CD of a vehicle steering wheel; the driving state data of the driver comprises a change range GP of the position of a fixation point of the driver and a change range GF of the grip strength of the driver to a steering wheel;

and 2-2, acquiring vehicle and driver state data of n time periods, grouping the vehicle driving state data and the driver driving state data in each time period, and establishing a vehicle and driver behavior data set.

Step 3, performing FCM clustering analysis on the collected vehicle driving state data and the driver driving state data to obtain a driver distraction state and a corresponding early warning level; the method comprises the following specific steps:

3-1, using data indexes in the vehicle and driver behavior data set as input data, using an FCM clustering method to perform clustering estimation on the distraction state of the driver in combination with the behavior of the driver, and dividing a clustering result into the following three early warning levels:

early warning in class I: the driver does not make corresponding operation behavior when the vehicle turns/changes lanes;

II type early warning: the driver takes corresponding operation when the vehicle turns/changes lanes, but a distraction state exists;

no early warning is performed: the driver takes corresponding operation when the vehicle turns/changes lanes, and the distraction state does not exist.

3-2, after the cluster analysis is finished, outputting an angle change range CT of the vehicle driving track direction, a rotation angle change range CD of a vehicle steering wheel, a change range GP of the driver gazing point position and a grip change range GF of the driver to the steering wheel, wherein the four indexes correspond to the characteristic ranges of three early warning levels;

3-3, taking the four indexes output in the step 3-2 as data processing and judging bases, and judging the distracted driving state of the driver; real-time vehicle and driver data collected in the normal driving process of the vehicle are compared with the real-time vehicle and the driver data to determine whether the driver has a distracted driving state and whether early warning is needed.

Step 4, collecting vehicle index data and judging whether the vehicle enters a turning/lane changing state or not in the normal running process of the vehicle; if the vehicle enters a turning/lane changing state, the step 5 is entered, otherwise, the data collection is continued.

Step 5, after the vehicle is determined to enter a turning/lane changing state, acquiring vehicle driving state data and driver driving state index data; comparing the collected data with the characteristic range of the early warning hierarchy output by clustering analysis to determine the driving state of the vehicle driver; the method comprises the following specific steps:

if the grip strength of the driver does not exceed the grip strength change range GF, the gaze point of the driver does not exceed the change range GP of the position of the gaze point, the vehicle track does not exceed the angle change range CT of the direction of the driving track, and the change of the steering wheel angle does not exceed the rotation angle change range CD, indicating that the driver makes corresponding operation behavior when the vehicle turns/changes lanes and no distraction state exists.

If the grip strength of the driver does not exceed the grip strength change range GF, the gaze point of the driver does not exceed the change range GP of the position of the gaze point, and simultaneously the angle change range CT that the vehicle track exceeds the direction of the driving track and the change of the steering wheel angle exceeds the change range CD of the rotation angle, the driver does not make corresponding operation behaviors when the vehicle turns/changes lanes.

If the grip strength of the driver exceeds the grip strength change range GF, the gaze point of the driver does not exceed the change range GP of the position of the gaze point, the vehicle track does not exceed the angle change range CT of the direction of the driving track, and the change of the steering wheel angle does not exceed the rotation angle change range CD, indicating that the driver takes corresponding operation when the vehicle turns/changes lanes, but the distraction state exists.

Step 6, judging whether to carry out driving distraction early warning according to the driving state of the vehicle driver determined in the step 5, and taking corresponding early warning measures according to early warning levels; the method comprises the following specific steps:

if the driver of the vehicle takes corresponding operation after the vehicle enters the turning/lane changing scene and the distraction state does not exist, no early warning is carried out, and data acquisition is restarted;

if the driver of the vehicle does not take corresponding operation behavior after the vehicle enters the turning/lane changing scene, starting a class I early warning, and reminding the driver by flashing a vehicle-mounted indicator lamp;

if the driver of the vehicle takes corresponding operation after the vehicle enters the turning/lane changing scene, but the distraction state exists during the operation, the II-type early warning is started, and the driver is reminded through the flashing of the vehicle-mounted indicating lamp and the voice prompt.