CN106448059A - Wrist strap instrument based driver fatigue detection method - Google Patents
Wrist strap instrument based driver fatigue detection method Download PDFInfo
- Publication number
- CN106448059A CN106448059A CN201610392374.1A CN201610392374A CN106448059A CN 106448059 A CN106448059 A CN 106448059A CN 201610392374 A CN201610392374 A CN 201610392374A CN 106448059 A CN106448059 A CN 106448059A
- Authority
- CN
- China
- Prior art keywords
- steering wheel
- fatigue
- subset
- index
- model
- 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
Classifications
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/02—Alarms for ensuring the safety of persons
- G08B21/06—Alarms for ensuring the safety of persons indicating a condition of sleep, e.g. anti-dozing alarms
Abstract
The invention provides a wrist strap instrument based driver fatigue detection method. The method includes: acquiring a steering wheel rotational angular velocity signal; computing fatigue feature indexes according to the steering wheel rotational angular velocity signal, wherein the fatigue feature indexes at least include steering wheel rotational angular velocity average, steering wheel rotational angular velocity standard deviation, steering wheel rotational angular velocity first moment, steering wheel rotational angular velocity maximum value, steering wheel rotational angular velocity quartile and steering wheel rotational angular velocity average larger than 75% fractile as well as steering wheel immobility times, steering wheel continuous immobility time and steering wheel reciprocating correction frequency; optimizing the extracted fatigue feature indexes so as to screen an optimal index subset from a fatigue judgement index universal set; using the optimal index subset as input to build an SVM (support vector machine) based fatigue detection model for fatigue detection.
Description
Technical field
The present invention relates to sensor technical field, particularly relates to a kind of driver fatigue detection side based on wrist strap instrument
Method.
Background technology
Sensor technology has had become as the basis in the fields such as automatic control since the appearance, is all control machinery " eyes
Eyeball ", and all Based Intelligent Control are required for relying on sensor technology.By taking context aware systems as an example, which is that the advanced auxiliary that drives is
The basis of system (Advanced Automotive Driving System, ADAS), intelligent automobile and pilotless automobile.Mesh
Before, the main method of environment sensing be by overhead rotary laser radar, laser radar, millimetre-wave radar, photographic head, ultrasound wave
The sensing systems such as sensor are perceived to environment, especially the static state in environment and dynamic object thing are positioned and with
Track.
With the high speed development of modern society's automobile industry, either family-sized car or long-haul truck, to automobile
Driving safety is proposed higher level requirement.Existing vehicle driving safety detection method is the driving shape to driver
State, travel condition of vehicle and environmental information etc. carry out real-time monitoring, analyze and predict driving dangerousness state and reminded in time,
Pro-active intervention or control vehicle motion if necessary.
Fatigue driving is to threaten vehicle driving safety, causes a main cause of vehicle accident.Research shows, Hen Duojiao
Interpreter former capital is derived from fatigue driving.Result is investigated and analysed to 2894 Road Traffic Injury to show, 9.68% accident occurs
Come from the fatigue driving of driver.Carnegie Mellon University of the U.S. studies on driving simulator and shows:The great thing of motor vehicles
Therefore rate 9.1% is owing to fatigue driving.The significant damage of driving fatigue makes it just cause the attention of society before 70 years,
Driver just must not continuously be driven 4 hours in nineteen thirty-nine and write into law by U.S. government, with avoid because long when drive and cause
Fatigue driving accident.China law clear stipulaties driver must not continuously operating motor vehicles do not stopped more than 4 hours rest or
Person stopped the time of having a rest less than 20 minutes.
In general the vehicle accident that fatigue driving is caused affects huge, causes great casualties and property loss.
2013, there is vehicle accident 198394 in China, cause 213724 people injury, 58539 people death, direct property loss
103896.6 ten thousand yuan.Driving fatigue not only endangers huge but also generally existing.The research worker of Loughborough university is to 996
Name truck and 4621 car male driver follow-up investigations, it is found that 29% driver is dozed off when driving, wherein
10% driver is because fatigue causes accident.The Tilley in the U.S. etc. is asked by 1500 drivers to randomly selecting
Volume investigation finds, wherein 64% people recognizes to live through driving fatigue in driving procedure, and the people more than 7% is because depth fatigue
And occurred of short duration sleeping in driving procedure, once there is vehicle accident because of driving fatigue in 10% people.Seko is in Japan
Similar research is also carried out, up to 75% driver had the experience of fatigue driving in his research.
Serious harm is had based on driving fatigue, universal existence and dependence laws and regulations and self abnegation are difficult to about
Shu Xing, becomes particularly important using technological means exploitation driver fatigue state recognition prior-warning device.Driver fatigue state
Detection technique has become an important development direction of advanced automobile security technology area.Research fatigue driving real-time detection side
Method is significant to improving traffic safety status.
For reducing the vehicle accident that fatigue driving causes, traffic safety is improved, various countries actively develop fatigue driving inspection
The SAVE-IT project of the research of survey method, the such as U.S., the AWAKE project in Europe, ASV project of Japan etc..Exploitation is high performance
Driver fatigue state inspection has become an important development direction of advanced automobile security technology area.More become now
Ripe giving fatigue pre-warning technology is generally based on machine vision, but recognizer is relative complex, and is easily subject to during data acquisition
Extraneous factor is disturbed.From current automobile using giving fatigue pre-warning system situation in terms of, typically all apply in medium-to-high grade vehicle, to general
For vehicle, cost becomes the key factor whether restriction installs giving fatigue pre-warning system.
Content of the invention
For unfavorable to fatigue driving monitoring problem present in prior art, the technical problem to be solved in the present invention is
There is provided a kind of low cost, easy for installation, to driver drive vehicle processes in without any interference and can be to fatigue driving state
Carry out the Driver Fatigue Detection based on wrist strap instrument of accurate warning.
In order to solve the above problems, the embodiment of the present invention proposes a kind of driver fatigue detection side based on wrist strap instrument
Method, including:
Step 1, acquisition steering wheel angle rate signal;
Step 2, fatigue characteristic index is calculated according to steering wheel angle rate signal, the wherein fatigue characteristic index includes
At least include:Steering wheel angle speed average, steering wheel angle velocity standard are poor, steering wheel angle speed first moment, steering wheel
Tarnsition velocity maximum value steering wheel angle speed quartile value, the steering wheel angle speed average more than 75% tantile;Also wrap
Include the lasting motionless number of times of steering wheel, steering wheel to continue dead time, steering wheel and back and forth revise frequency;
Step 3, the fatigue characteristic index to extracting are optimized, and are referred to filtering out optimum from tired discriminant criterion complete or collected works
Mark subset;
Step 4, by the use of the optimal index subset as input, set up for carry out fatigue detecting based on the tired of SVM
Labor detection model.
Wherein, the step 1 is specifically included:
Step 11, the acceleration transducer collection recent movement by being worn in the wrist strap instrument in driver's wrist
Acceleration signal;
Step 12, the acceleration signal of recent movement to receiving obtain steering wheel angle speed to time integral
Signal.
Wherein the step 2 is specifically included:
Step 21, by the steering wheel angle rate signal cutting for getting for predetermined time interval data segment;
In step 22, each data segment of acquisition, in 2s, steering wheel is rotated and is less than 0.5 °/s, 1 °/s, the data of 1.5 °/s, with
And in 4s, steering wheel is rotated less than 0.5 °/s, 1 °/s, the data of 1.5 °/s in each data segment;
Step 23, obtain in each data segment steering wheel correction threshold more than 1 °/s, 3 °/s, 5 °/s, 10 °/s, 15 °/s,
The data of 20 °/s;
Step 24, it is extracted maximum angle standard deviation and maximum zero-speed percentage ratio using dual-time window index extraction method, its
Described in slip window width and corresponding step-length take respectively 15s and 5s, 10s and 5s, 12s and 4s, 16s and 4s, 18s and 3s, 15s and
3s.
Wherein the step 3 is specifically included:
Step 31, the fatigue characteristic index optimization that sets up select formula:
Wherein, Y is tired discriminant criterion complete or collected works, and X is the nonvoid subset of Y, and J (X) is criterion function, i.e. tired discrimination model
Recognition accuracy, NXIt is test sample number, nXIt is the sample number of the correct identification of fatigue detecting model in test sample;
Step 32, nonvoid subset X is searched out from tired discriminant criterion complete or collected works Y using SFFS algorithm with solution formula (1);
Then by nonvoid subset X for being input into, using training sample, driver fatigue state-detection model, then profit are set up based on SVM algorithm
With the classification performance of test sample test model, value J (X) of criterion function is asked for;Just wherein meet the subset X of J (X) value maximum
It is the optimal solution of formula (1);
Wherein, SFFS algorithm is referred to from the beginning of empty set, is often taken turns selection a subset x in unselected feature, is made addition
Subset x post-evaluation function reaches optimum, then selects subset z in the feature for having selected, reaches rejecting subset z post-evaluation function
To optimum, optimal index subset X is obtained.
Wherein, the step 4 includes:
Step 41, using optimal index subset X as input, be to reduce big value index component to produce little value index component
Dominating role, improves the recognition accuracy of fatigue detecting model, the index order of magnitude is normalized;
Step 42, wherein the fatigue detecting model using the training of optimal index subset X based on SVM, the kernel function of the SVM
For Radial basis kernel function, penalty coefficient C and core variable σ are optimized
The object function that recognition accuracy using model on test set optimizes as variable, using grid data service grid
Search Grid-Search algorithm searches optimum variable combination (C, σ), and it is core variable that wherein C is penalty coefficient, σ;Concrete formula
For:
J (X) is criterion function, i.e. the recognition accuracy of tired discrimination model, NXIt is test sample number, nXIt is test sample
The sample number that middle fatigue detecting model is correctly recognized, C is penalty coefficient, and σ is core variable.
The target function value J (X) of model variable optimization and metrics evaluation criterion is asked for using cross-validation method.
The having the beneficial effect that of the technique scheme of the present invention:
Technical scheme gathers steering wheel angle acceleration letter by the acceleration transducer in wrist strap type device
Breath, avoids the acquisition modes of the data for detecting based on the physiological feature shortcoming stronger to driver's normal driving invasive, with
When evaded based on being easily subject to extraneous factor interference and recognizer relative complex during data acquisition in facial feature detection
The problems such as.The method signals collecting of the embodiment of the present invention is easy, with a high credibility, it is only necessary to by simple in wrist strap type device,
Low cost acceleration transducer and by the Bluetooth transmission GES between vehicle and wrist strap type device, you can realize drive
The purpose of people's giving fatigue pre-warning, is remarkably improved the pervasive degree of giving fatigue pre-warning.The embodiment of the present invention is solved because external environment is (as light
According to) difference cause prohibited data detection true, and because early warning system for prompting there is a problem of cause interference with driving.The present invention
The detection of driver's vehicle-mounted fatigue is applied to, is capable of the fatigue state of real-time reminding driver, farthest avoids leading because of fatigue
The generation of the vehicle accident of cause;Apply also for the report of driver fatigue state post-project evaluating and feed back.
Description of the drawings
Fig. 1 is the method flow diagram in the embodiment of the present invention;
Fig. 2 is the flow chart of index optimization selection algorithm SFFS;
Fig. 3 is the Establishing process figure of metrics evaluation criterion;
Fig. 4 is the precision curve of index optimization selection algorithm.
Specific embodiment
For making the technical problem to be solved in the present invention, technical scheme and advantage clearer, below in conjunction with accompanying drawing and tool
Body embodiment is described in detail.
In the embodiment of the present invention, driver fatigue detection is carried out by method as shown in Figure 1.Specifically include:
Step 1, data collection steps:
Recent movement signal is gathered by the wrist strap instrument being worn in driver's wrist, and gathers GES.?
In an embodiment of the invention, in the wrist strap instrument, acceleration transducer is provided with to gather the acceleration of recent movement
Signal;Then speed information is gathered by blue-tooth device.
Step 2, data prediction step:
Steering wheel angle rate signal is obtained according to the recent movement signal for obtaining, can specifically be adopted to acceleration
Sensor acquisition to steering wheel acceleration signal obtain steering wheel angle rate signal to time integral.
Step 3, according to steering wheel angle rate signal calculate fatigue characteristic index:
Find under study for action, the fluctuation range of steering wheel angle speed and frequency are that driver obtains main fatigue characteristic, this
Using the steering wheel angle rate signal in the scheduled time as a sample (for example with the data in 60s in inventive embodiments
For a sample), from the steering wheel angle speed of each sample, then it is extracted average (Mean), standard deviation (Std), single order
Square (FOM), maximum (Max), quartile value (Q1, Q2, Q3) and the average (Mean_Q3) more than 75% tantile etc. are basic
Statistic is used as statistics class index.
Except counting class index, it is lasting that the embodiment of the present invention is also extracted the motionless number of times of steering wheel (NMRHOLD), steering wheel
Dead time (THRSHLD), steering wheel back and forth revise the experience class index such as frequency.
(1), the motionless number of times of steering wheel and time refer to steering wheel rotate threshold value take respectively 0.5 °/s, 1 °/s, 1.5 °/s,
Persistent period takes 2s and two times of 4s respectively, so then extracts this two category feature totally 2 × 3 × 2=12 index altogether.
(2), steering wheel back and forth revises frequency, take reciprocal correction amplitude thresholds be 1 °/s, 3 °/s, 5 °/s, 10 °/s, 15 °/
S, 20 °/s, such totally 6 indexs.
(3), using dual-time window index extraction method, maximum angle standard deviation (MSDA) and maximum zero-speed percentage ratio are extracted
Two indices (MPNS), slip window width and corresponding step-length take respectively 15s and 5s, 10s and 5s, 12s and 4s, 16s and 4s, 18s and
3s, 15s and 3s, so totally 2 × 6=12 index.8+12+6+12=38 index altogether.
Step 4, the fatigue characteristic index extracted by step 3 are in optimized selection, and set up based on SVM (Support
Vector Machine, support vector machine) fatigue detecting model:
Increase the dimension of discriminant criterion indiscriminately, the differentiation result of PRS can not be improved.Actually ask
In topic, one, with stronger dependency between tired discriminant criterion, there is bulk redundancy information;Two, tired discriminant criterion quantity
Various, affect operation efficiency and the structure complexity of fatigue detecting model;Three, disturbed by redundancy, poorly efficient index, complete with index
The method in collection construction feature space often deteriorates the classification performance of fatigue detecting model.Therefore, the purpose that index optimization is selected
As:From various, redundancy, the tired discriminant criterion of association, choose one group of optimum index to build Fatigue pattern classification
Feature space, realizes optimizing the purpose of fatigue detecting system performance.
The purpose that index optimization is selected is that to filter out quantity tired discriminant criterion complete or collected works from quantity for D be d (D>D)
Subset, makes the classification performance of the fatigue detecting model of foundation reach optimum.Need to solve two problems for this, one is to build one kind
Search strategy, the rule that agriculture products are selected into and reject, to select optimal index combination within the time for allowing;Two is to set up
A kind of interpretational criteria, can be evaluated exactly to the classification capacity of different fatigue discriminant criterion combination.
Contrasted by analysis, the present invention combines illumination scan using Wrapper method interpretational criteria and sets up approximately most
Excellent index selection algorithm.I.e. step 4 is specifically included:
Step 41, the index optimization that sets up select formula:
Wherein, Y is tired discriminant criterion complete or collected works, and X is the nonvoid subset of Y, and J (X) is criterion function, i.e. tired discrimination model
Recognition accuracy, NXIt is test sample number, nXIt is the sample number of the correct identification of fatigue detecting model in test sample.
Step 42, nonvoid subset X is searched out from index complete or collected works Y using SFFS algorithm with solution formula (1);Then will be non-
Vacuous subset X is input, sets up driver fatigue state-detection model using training sample based on SVM algorithm, recycles test specimens
The classification performance of this test model, asks for value J (X) of criterion function;The subset X for wherein meeting J (X) value maximum is exactly formula
(1) optimal solution;The subset can make the recognition accuracy highest of the driver fatigue state-detection model that sets up based on SVM.
Illumination scan is floated using forward sequence and selects (SFFS) algorithm, and SFFS algorithm is often taken turns from the beginning of empty set
A subset x is selected in unselected feature, make addition subset x post-evaluation function that optimum is reached, then in the spy for having selected
Middle selection subset z is levied, is made rejecting subset z post-evaluation function that optimum is reached, obtain optimal index subset;As shown in Figure 2.
Step 43, using optimal index subset as input, set up and driver fatigue state-detection model set up based on SVM.
Nicety of grading using model on test set is used as the interpretational criteria of measurement index classification performance.The meter of criterion function
Flow process is calculated, including three links:One, input element, with the tired discriminant criterion subset X of screening as input, is to reduce big value to refer to
The dominating role that mark component is produced to little value index component, improves the recognition accuracy of fatigue detecting model, to the index order of magnitude
It is normalized;Two, model training link, using fatigue detecting model of the index subset X training based on SVM, the present invention is selected
Radial basis kernel function needs to be optimized penalty coefficient C and core variable σ as the kernel function of SVM.Wherein penalty coefficient C is controlled
The recognition accuracy of driver fatigue detection model processed and generalization ability, core variable σ determines the nonlinear problem of luv space
Transform to the linear separability after higher dimensional space.The purpose of variable optimization is just to determine suitable C and σ, to ensure fatigue detecting
The recognition accuracy of model and generalization ability.The target letter that recognition accuracy using model on test set optimizes as variable
Number, searches optimum variable using grid data service grid search (Grid-Search) method and combines (C, σ), as formula (2) institute
Show;Three, model measurement link, the target function value J of model variable optimization and metrics evaluation criterion is asked for using cross-validation method
(X).
J (X) is criterion function, i.e. the recognition accuracy of tired discrimination model, NXIt is test sample number, nXIt is test sample
The sample number that middle fatigue detecting model is correctly recognized, C is penalty coefficient, and σ is core variable.
In this example, the precision curve of index Optimization Selection Algorithm is as shown in Figure 4.
Step 5, using the fatigue detecting model based on SVM, forecasting fatigue is carried out to the real time data for receiving.
The above is the preferred embodiment of the present invention, it is noted that for those skilled in the art
For, on the premise of without departing from principle of the present invention, some improvements and modifications can also be made, these improvements and modifications
Should be regarded as protection scope of the present invention.
Claims (5)
1. a kind of Driver Fatigue Detection based on wrist strap instrument, it is characterised in that include:
Step 1, acquisition steering wheel angle rate signal;
Step 2, fatigue characteristic index is calculated according to steering wheel angle rate signal, the wherein fatigue characteristic index includes at least
Including:Steering wheel angle speed average, steering wheel angle velocity standard are poor, steering wheel angle speed first moment, steering wheel angle
Speed maximum value steering wheel angle speed quartile value, the steering wheel angle speed average more than 75% tantile;The also side of including
Continue dead time, steering wheel to the motionless number of times of disk, steering wheel and back and forth revise frequency;
Step 3, the fatigue characteristic index to extracting are optimized, to filter out optimal index from tired discriminant criterion complete or collected works
Collection;
Step 4, by the use of the optimal index subset as input, set up the fatigue inspection based on SVM for carrying out fatigue detecting
Survey model.
2. the Driver Fatigue Detection based on wrist strap instrument according to claim 1, it is characterised in that the step
1 specifically includes:
Step 11, gathered by the acceleration transducer that is worn in the wrist strap instrument in driver's wrist recent movement plus
Rate signal;
Step 12, the acceleration signal of recent movement to receiving obtain steering wheel angle speed letter to time integral
Number.
3. the Driver Fatigue Detection based on wrist strap instrument according to claim 1, it is characterised in that the step
2 specifically include:
Step 21, by the steering wheel angle rate signal cutting for getting for predetermined time interval data segment;
In step 22, each data segment of acquisition, in 2s, steering wheel is rotated and is less than 0.5 °/s, 1 °/s, the data of 1.5 °/s, and per
In one data segment, in 4s, steering wheel is rotated and is less than 0.5 °/s, 1 °/s, the data of 1.5 °/s;
Step 23, obtain in each data segment steering wheel correction threshold more than 1 °/s, 3 °/s, 5 °/s, 10 °/s, 15 °/s, 20 °/s
Data;
Step 24, it is extracted maximum angle standard deviation and maximum zero-speed percentage ratio, wherein institute using dual-time window index extraction method
State slip window width and corresponding step-length takes 15s and 5s, 10s and 5s, 12s and 4s, 16s and 4s, 18s and 3s, 15s and 3s respectively.
4. the Driver Fatigue Detection based on wrist strap instrument according to claim 1, it is characterised in that the step
3 specifically include:
Step 31, the fatigue characteristic index optimization that sets up select formula:
Wherein, Y is tired discriminant criterion complete or collected works, and X is the nonvoid subset of Y, and J (X) is criterion function, i.e. the knowledge of tired discrimination model
Other accuracy rate, NXIt is test sample number, nXIt is the sample number of the correct identification of fatigue detecting model in test sample;
Step 32, nonvoid subset X is searched out from tired discriminant criterion complete or collected works Y using SFFS algorithm with solution formula (1);Then
Nonvoid subset X is input, using training sample, driver fatigue state-detection model is set up based on SVM algorithm, recycle and survey
The classification performance of this test model of sample, asks for value J (X) of criterion function;The subset X for wherein meeting J (X) value maximum is exactly public
The optimal solution of formula (1);
Wherein, SFFS algorithm is referred to from the beginning of empty set, is often taken turns selection a subset x in unselected feature, is made addition subset x
Post-evaluation function reaches optimum, then selects subset z in the feature for having selected, so that rejecting subset z post-evaluation function is reached most
Excellent, obtain optimal index subset X.
5. the Driver Fatigue Detection based on wrist strap instrument according to claim 1, it is characterised in that the step
4 include:
Step 41, using optimal index subset X as input, be to reduce the domination that big value index component is produced to little value index component
Effect, improves the recognition accuracy of fatigue detecting model, the index order of magnitude is normalized;
Step 42, the fatigue detecting model using the training of optimal index subset X based on SVM, the wherein kernel function of the SVM are footpath
To base kernel function, penalty coefficient C and core variable σ are optimized
The object function that recognition accuracy using model on test set optimizes as variable, using grid data service grid search
Grid-Search algorithm searches optimum variable combination (C, σ), and it is core variable that wherein C is penalty coefficient, σ;Specifically formula is:
J (X) is criterion function, i.e. the recognition accuracy of tired discrimination model, NXIt is test sample number, nXIt is tired in test sample
The sample number that labor detection model is correctly recognized, C is penalty coefficient, and σ is core variable.
The target function value J (X) of model variable optimization and metrics evaluation criterion is asked for using cross-validation method.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610392374.1A CN106448059A (en) | 2016-06-06 | 2016-06-06 | Wrist strap instrument based driver fatigue detection method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610392374.1A CN106448059A (en) | 2016-06-06 | 2016-06-06 | Wrist strap instrument based driver fatigue detection method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106448059A true CN106448059A (en) | 2017-02-22 |
Family
ID=58183962
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610392374.1A Pending CN106448059A (en) | 2016-06-06 | 2016-06-06 | Wrist strap instrument based driver fatigue detection method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106448059A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108682119A (en) * | 2018-05-29 | 2018-10-19 | 重庆大学 | Method for detecting fatigue state of driver based on smart mobile phone and smartwatch |
CN110171426A (en) * | 2019-05-23 | 2019-08-27 | 重庆大学 | A kind of recognition methods of driver's dangerous driving behavior and system based on smart machine |
CN110877516A (en) * | 2018-09-05 | 2020-03-13 | 上海蔚来汽车有限公司 | Fragrance controller, vehicle-mounted fragrance device, vehicle-mounted fragrance system and control method thereof |
CN114066297A (en) * | 2021-11-24 | 2022-02-18 | 西南交通大学 | Method for identifying working state of high-speed railway traffic dispatcher |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090019153A1 (en) * | 2007-07-12 | 2009-01-15 | Viasat, Inc. | Methods and systems for performing a prefetch abort operation |
CN103462618A (en) * | 2013-09-04 | 2013-12-25 | 江苏大学 | Automobile driver fatigue detecting method based on steering wheel angle features |
CN103552560A (en) * | 2013-11-01 | 2014-02-05 | 扬州瑞控汽车电子有限公司 | Driver driving state recognition-based lane departure alarming method |
CN104036619A (en) * | 2013-03-04 | 2014-09-10 | 德尔福电子(苏州)有限公司 | Fatigue detection method and detection apparatus based on turn angle data of steering wheel |
CN104688252A (en) * | 2015-03-16 | 2015-06-10 | 清华大学 | Method for detecting fatigue status of driver through steering wheel rotation angle information |
CN105631485A (en) * | 2016-03-28 | 2016-06-01 | 苏州阿凡提网络技术有限公司 | Fatigue driving detection-oriented steering wheel operation feature extraction method |
-
2016
- 2016-06-06 CN CN201610392374.1A patent/CN106448059A/en active Pending
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20090019153A1 (en) * | 2007-07-12 | 2009-01-15 | Viasat, Inc. | Methods and systems for performing a prefetch abort operation |
CN104036619A (en) * | 2013-03-04 | 2014-09-10 | 德尔福电子(苏州)有限公司 | Fatigue detection method and detection apparatus based on turn angle data of steering wheel |
CN103462618A (en) * | 2013-09-04 | 2013-12-25 | 江苏大学 | Automobile driver fatigue detecting method based on steering wheel angle features |
CN103552560A (en) * | 2013-11-01 | 2014-02-05 | 扬州瑞控汽车电子有限公司 | Driver driving state recognition-based lane departure alarming method |
CN104688252A (en) * | 2015-03-16 | 2015-06-10 | 清华大学 | Method for detecting fatigue status of driver through steering wheel rotation angle information |
CN105631485A (en) * | 2016-03-28 | 2016-06-01 | 苏州阿凡提网络技术有限公司 | Fatigue driving detection-oriented steering wheel operation feature extraction method |
Non-Patent Citations (1)
Title |
---|
屈肖蕾: "基于转向操作和车辆状态的疲劳驾驶检测方法研究", 《中国优秀硕士学位论文全文数据库工程科技II辑》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108682119A (en) * | 2018-05-29 | 2018-10-19 | 重庆大学 | Method for detecting fatigue state of driver based on smart mobile phone and smartwatch |
CN110877516A (en) * | 2018-09-05 | 2020-03-13 | 上海蔚来汽车有限公司 | Fragrance controller, vehicle-mounted fragrance device, vehicle-mounted fragrance system and control method thereof |
CN110171426A (en) * | 2019-05-23 | 2019-08-27 | 重庆大学 | A kind of recognition methods of driver's dangerous driving behavior and system based on smart machine |
CN110171426B (en) * | 2019-05-23 | 2021-07-02 | 重庆大学 | Driver dangerous driving behavior identification method and system based on intelligent equipment |
CN114066297A (en) * | 2021-11-24 | 2022-02-18 | 西南交通大学 | Method for identifying working state of high-speed railway traffic dispatcher |
CN114066297B (en) * | 2021-11-24 | 2023-04-18 | 西南交通大学 | Method for identifying working state of high-speed railway traffic dispatcher |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104269026B (en) | Fatigue driving real-time monitoring and early warning method based on Android platform | |
US10414407B1 (en) | Driving analysis using vehicle-to-vehicle communication | |
CN107000750A (en) | For the improved device and method of vehicle economy | |
CN106448059A (en) | Wrist strap instrument based driver fatigue detection method | |
CN105564436A (en) | Advanced driver assistance system | |
CN106314438A (en) | Method and system for detecting abnormal track in driver driving track | |
CN103552560A (en) | Driver driving state recognition-based lane departure alarming method | |
CN108248610B (en) | It is a kind of adaptively divert one's attention driving behavior monitoring and tampering devic | |
EP3498559B1 (en) | Method for recognizing the driving style of a driver of a land vehicle, and corresponding apparatus | |
CN104408915B (en) | A kind of method of estimation of traffic state data and system | |
CN105574537A (en) | Multi-sensor-based dangerous driving behavior detection and evaluation method | |
CN101002239A (en) | System and method for monitoring driving | |
Sun et al. | An integrated solution for lane level irregular driving detection on highways | |
WO2015081335A2 (en) | Advanced context-based driver scoring | |
CN111311093B (en) | Road intersection risk assessment and early warning method based on driver physiological data | |
CN113378741B (en) | Auxiliary sensing method and system for aircraft tractor based on multi-source sensor | |
CN110766982B (en) | Vehicle collision detection system based on vehicle-mounted sensor | |
CN106650157A (en) | Method, device and system for vehicle part fault probability estimation | |
El Masri et al. | Toward self-policing: Detecting drunk driving behaviors through sampling CAN bus data | |
CN112356849A (en) | Early warning method and device for dangerous driving | |
He et al. | An individual driving behavior portrait approach for professional driver of HDVs with naturalistic driving data | |
Yardy et al. | Detecting malicious driving with machine learning | |
Fogue et al. | Using data mining and vehicular networks to estimate the severity of traffic accidents | |
CN105303837A (en) | Method and system for detecting following behavior characteristic parameter of driver | |
CN106447497A (en) | Safe driving monitoring method, vehicle, electronic device and insurance server |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into 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: 20170222 |