CN107095670A - Time of driver's reaction Forecasting Methodology - Google Patents
Time of driver's reaction Forecasting Methodology Download PDFInfo
- Publication number
- CN107095670A CN107095670A CN201710391410.7A CN201710391410A CN107095670A CN 107095670 A CN107095670 A CN 107095670A CN 201710391410 A CN201710391410 A CN 201710391410A CN 107095670 A CN107095670 A CN 107095670A
- Authority
- CN
- China
- Prior art keywords
- mrow
- msub
- driver
- time
- reaction
- 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
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/369—Electroencephalography [EEG]
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/16—Devices for psychotechnics; Testing reaction times ; Devices for evaluating the psychological state
- A61B5/162—Testing reaction times
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/16—Devices for psychotechnics; Testing reaction times ; Devices for evaluating the psychological state
- A61B5/18—Devices for psychotechnics; Testing reaction times ; Devices for evaluating the psychological state for vehicle drivers or machine operators
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7203—Signal processing specially adapted for physiological signals or for diagnostic purposes for noise prevention, reduction or removal
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/725—Details of waveform analysis using specific filters therefor, e.g. Kalman or adaptive filters
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7271—Specific aspects of physiological measurement analysis
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B2503/00—Evaluating a particular growth phase or type of persons or animals
- A61B2503/20—Workers
- A61B2503/22—Motor vehicles operators, e.g. drivers, pilots, captains
Landscapes
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Veterinary Medicine (AREA)
- Public Health (AREA)
- Psychiatry (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Animal Behavior & Ethology (AREA)
- Surgery (AREA)
- Artificial Intelligence (AREA)
- Signal Processing (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Psychology (AREA)
- Educational Technology (AREA)
- Developmental Disabilities (AREA)
- Hospice & Palliative Care (AREA)
- Child & Adolescent Psychology (AREA)
- Social Psychology (AREA)
- Evolutionary Computation (AREA)
- Fuzzy Systems (AREA)
- Mathematical Physics (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The invention discloses a kind of time of driver's reaction Forecasting Methodology, the time of driver's reaction Forecasting Methodology includes step：Obtain the EEG signals of driver；The brain electrical feature parameter in the EEG signals after filtered processing is extracted by wavelet transformation；Using the brain electrical feature parameter of driver as input layer, using the reaction time as output layer, BP neural network forecast model is built.In the present invention, brain electrical feature parameter is extracted by wavelet transformation and is used as objective prediction index, with reference to BP neural network method, a kind of prediction algorithm of time of driver's reaction is constructed, so that the exploitation for the dangerous driving condition early warning system of vehicle-mounted real-time driver provides theoretical foundation with design.
Description
Technical field
The present invention relates to specific people's reaction time field of measuring technique, more particularly to a kind of time of driver's reaction prediction
Method.
Background technology
The continuous lifting of high ferro speed proposes requirements at the higher level, wherein EMUs driver to the work capacity of EMUs driver
It is to influence the key factor of driving operation security reliability on the respond of accident.
The correlative study predicted both at home and abroad with regard to EMUs driver (and automobile driver) reaction time at present is still rare,
Existing research is inquired into mainly for the reaction time with neuro-physiological signals correlation.Jap et al. have studied in analog loop
Trainman EEG signals and the correlation in reaction time in dull driving procedure for a long time under border, result of study show brain telecommunications
Low-frequency band presents positive correlation with the reaction time in number, and high frequency band is presented negatively correlated with the reaction time.Darwent et al.
The alertness of trainman is monitored under true environment, as a result shown with the decline for driving alertness, to burst
The reaction time of event constantly extends, and is closely connected while demonstrating EEG signals and having been existed with the reaction time.Haga et al.
Experiment of the trainman to the respond of signal lamp is devised, machine can directly be reflected by demonstrating the change of EEG signals
The respond of car driver.Lin et al. utilizes EEG signals and the correlation of driving behavior performance (reaction time, speed etc.),
Real-time monitoring has been carried out to driving Sustained attention level.
More than research demonstrate reaction time and neuro-physiological signals from different perspectives there is correlation, but fail to use two
Person's correlation was effectively predicted the reaction time.Therefore how by the neuro-physiological signals of EMUs driver, motor-car is realized
Effective prediction of the group driver to the reaction time of accident, is that the dangerous driving condition of the vehicle-mounted real-time EMUs driver of structure is pre-
The key technique of alert system.
The content of the invention
In view of this, the present invention is intended to provide a kind of time of driver's reaction Forecasting Methodology, i.e., extracted by wavelet transformation
Brain electrical feature parameter, with reference to BP neural network method, constructs a kind of the pre- of time of driver's reaction as objective prediction index
Method of determining and calculating, so that the exploitation for the dangerous driving condition early warning system of vehicle-mounted real-time driver provides theoretical foundation with design.
Specifically, time of driver's reaction Forecasting Methodology of the invention includes step：Obtain the EEG signals of driver；
The brain electrical feature parameter in the EEG signals after filtered processing is extracted by wavelet transformation；With the brain electrical feature parameter of driver
As input layer, using the reaction time as output layer, BP neural network forecast model is built.
Further, the filtering process is specifically included：EEG signals carry out overall filtering process with 0-35Hz bandwidth.
Further, brain electrical feature parameter is extracted by wavelet transformation to specifically include：
A, for the EEG signals after filtered processing, be designated as u (n), then its wavelet transformation is defined as：
In formula,For wavelet function；I is frequency factor；M is the time-shifting factor；N is signal duration；
Signal u (n) after b, wavelet transformation carries out finite layer decomposition, using Mallat algorithms：
In formula, AH is approximation component；Ci is the details coefficients under different scale；H is Decomposition order；Pass through above-mentioned finite layer
Decompose the wavelet coefficient for obtaining 3 kinds of different frequency ranges of θ, α, β；
C, the energy value to above-mentioned frequency extraction wavelet coefficient are used as brain electrical feature parameter：
In formula, PXFor the energy value of corresponding band；SX(t) it is the wavelet coefficient of corresponding band；T is the time；hiFor respective tones
The amplitude of section；
D, the EEG signals to q electrode are handled, and accordingly obtain 3 × q brain electrical feature parameters.
Further, methods described also includes：Resulting brain electrical feature parameter is normalized by formula, makes brain
The numerical value of electrical feature parameter is between [0,1], to eliminate noise present in data：
In formula, ximaxWith ximinFor brain electrical feature parameter xiMaxima and minima.
Further, BP neural network forecast model is built to specifically include：Using the brain electrical feature parameter of driver as defeated
Enter layer, using reaction time predicted value as output layer, build the BP neural network forecast model containing 1 hidden layer；Wherein, input
Node layer number is determined by the brain electrical feature number of parameters of driver；Hidden layer node number s is then carried out by model training result
Preferentially choose；Output layer node number is 1.
Further, in the model input layer to hidden layer, hidden layer to the connection weight coefficient between output layer and
Biasing is respectively wik,wk1,bik,bk1(i=1,2 ..., 3 × q, k=1,2 ..., s), for input layer arbitrary node o extremely
The arbitrary node p of hidden layer output：
yop=f (xiwop+bop)
In formula, f () is Sigmoid functions, i.e.,
Output layer output result is：
In formula, yiExported for Neural Network model predictive result；W1For the connection weight number system number square of input layer to hidden layer
Battle array；W2For the connection weight number system matrix number of hidden layer to output layer；xiFor driver's brain electrical feature parameter；b1It is input layer to hidden
Bias matrix containing layer；b2For the bias matrix of hidden layer to output layer.
Further, the EEG signals for obtaining driver are specifically included：Gather the brain electricity of the driver of predetermined number
Signal.
Further, eeg signal acquisition frequency is 10Hz；Electroencephalogramdata data collector continuous acquisition driver's eeg data.
Further, the reaction time predicted value obtained by BP neural network model is real with testing the gathered reaction time
The comparison of actual value fitting degree chooses maximum absolute error M1 and relative mean square error M2 and is used as prediction effect assessment indicator：
The present invention extracts brain electrical feature parameter by wavelet transformation is used as objective prediction index, with reference to BP neural network side
Method, constructs a kind of prediction algorithm of time of driver's reaction, realizes driver accurate pre- to the accident reaction time
Survey, this provides theoretical foundation to the dangerous driving condition early warning system exploitation of vehicle-mounted real-time driver with design.
Brief description of the drawings
The accompanying drawing for being incorporated into specification and constituting a part for specification shows embodiments of the invention, and with
Description is used for the principle for explaining the present invention together.In the drawings, similar reference is used to represent similar key element.Under
Accompanying drawing in the description of face is some embodiments of the present invention, rather than whole embodiments.Come for those of ordinary skill in the art
Say, on the premise of not paying creative work, other accompanying drawings can be obtained according to these accompanying drawings.
Fig. 1 is the schematic flow sheet of time of driver's reaction Forecasting Methodology provided in an embodiment of the present invention；
There is the schematic diagram of position for experimental duties moderate stimulation point in the embodiment of the present invention in Fig. 2；
The schematic diagram that Fig. 3 flashes for detectable signal in experimental duties in the embodiment of the present invention；
Fig. 4 is the BP neural network model structure schematic diagram in the embodiment of the present invention；
Fig. 5 is BP neural network model prediction reaction time and the contrast signal of real reaction time in the embodiment of the present invention
Figure；
Fig. 6 is BP neural network model prediction reaction time error schematic diagram in the embodiment of the present invention.
Embodiment
To make the purpose, technical scheme and advantage of the embodiment of the present invention clearer, below in conjunction with the embodiment of the present invention
In accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is
A part of embodiment of the present invention, rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art
The every other embodiment obtained under the premise of creative work is not made, belongs to the scope of protection of the invention.Need
Illustrate, in the case where not conflicting, the feature in embodiment and embodiment in the application can be mutually combined.
Below in conjunction with the accompanying drawings and scene describe in detail the embodiment of the present invention time of driver's reaction Forecasting Methodology.
Shown in Figure 1, the method for the time of driver's reaction Forecasting Methodology includes step：Obtain the brain telecommunications of driver
Number；The brain electrical feature parameter in the EEG signals after filtered processing is extracted by wavelet transformation；With the brain electrical feature of driver
Parameter is as input layer, using the reaction time as output layer, builds BP neural network forecast model.
It is detailed with reference to specific experiment scene, example in order to more illustrate the advantage, principle and effect of the method for the invention
It is thin to illustrate.
1 experimental method
1.1 subjects are chosen
Choose 20 some EMUs Drivers Crew male students and be used as subject.Age, average was 36.1 between 34-38 Sui
Year, standard deviation is 1.8 years old；Driving age, average was 7.2 years in 6-11, and standard deviation is 1.1 years.Selected subject sleep quality is good,
Health is good, no bad habit (smoking, be addicted to drink), no anomalous trichromatism or colour blindness, and eyesight or correct vision are 1.0.It is real
Test 4h before beginning to forbid being tested the reviver such as coffee for drinking or tea product, and be fully understood by《Informed consent form》Under the premise of
Voluntarily sign.
1.2 experimental facilities
1.2.1 bullet train simulator
Experiment is using the CRH380 type bullet train simulators with 6DOF kinematic system, and the simulator is big using single channel
To visual system before screen, its screen resolution is 1920 × 1200pix, and horizontal view angle is up to 160°.Locomotive master operating station is by arranging
Car automatic protective system (Automatic Train Protection, ATP), comprehensive special digital GSM and
The compositions such as DMI display screens, switch, indicator lamp, speed setting controller.Simulator acoustic environment is by 7.1 DAB sounding systems
System produce, the system can high fidelity simulation emulation motor-car operation when background sound environment.The validity of the simulator passes through
Systematic testing, its fidelity can meet requirement of experiment.
1.2.2 electroencephalogramdata data collector
Use the 64 of the production of Compumedics companies of Australia to lead Neuroscan electroencephalographs in experimentation in real time to connect
The EEG signals data of continuous collection subject.The electroencephalograph can be general using international electroencephalography 10-20 systems electrode cap, its
Electrode position has been set, and is chosen FCz electrodes and is used as reference electrode.
1.3 experimental duties
Experiment uses Foochow to Hefei southern station circuit, and total track length is 808km.Train is by way of 22 stations, stop of arriving at a station
The station operation of Shi Jinhang regular callings.When train is in each interval operation, driven by the way of random signal detection come real-time detection
Sail and the reaction time is tested in operation.Subject driver behavior during, in the screen of front five possible positions will present with
The time is presented for 120 ± 10s in machine signal (red point) (Fig. 2), signal.When there is (Fig. 3) in some position in signal, it is desirable to be tested
Made a response by the way that key mode is as fast as possible.If after 1000ms occurs in signal, being tested non-button and being then considered as this secondary response
It is invalid.
1.4 experiment flows and data acquisition
24h allows subject to understand experimental duties, operation rules before experiment starts, then according to drive routine custom operation mould
Intend device, until subject being capable of the skilled operation simulator.In order to ensure that subject has higher alertness level, the unified peace of experiment
Come the morning 8:00 is carried out.Before formal experiment starts, in order that subject adapts to drive simulation environment and enters experimental state,
Give the exercise of 15min drive simulations.In formal experimentation, indoor light illumination is 300lx, and temperature is 24 ± 1 DEG C.It is required that
It is tested and keeps motor-car operation, a length of 2h during driving operation to be not less than 220km/h speed.
In the reaction time that synchronous recording subject is stimulated random signal in driving procedure, data acquiring frequency is 10Hz.
Meanwhile, electroencephalogramdata data collector continuous acquisition subject eeg data.To remove other information interference, recording level and vertical eye electricity, flesh
Electricity.EEG signals sample rate is set, frequency acquisition band a width of 128Hz, 0.5-100Hz, it is desirable to which all electrode impedances must not exceed
5kΩ。
2. the reaction time forecast model based on EEG signals
The eeg data of the 20 EMUs drivers gathered for above-mentioned experiment, filtered place is extracted using wavelet transformation
Every brain electrical feature parameter in eeg data after reason, with reference to using brain electrical feature parameter as input pointer, reaction time conduct
The BP neural network of output-index, builds EMUs driver to accident reaction time forecast model.Concrete model builds step
It is rapid as follows：
2.1 brain electrical feature parameter extractions
EEG signals can reflect cerebral cortex active state, its brain electricity frequency when EMUs driver is in hypoarouse
Spectral structure is intended to low-frequency band, otherwise then EEG spectrum distribution is intended to high frequency band when high arousal level.Study
As a result θ in EEG signals (4~8Hz) is shown, α (8~13Hz), 3 kinds of frequency ranges of β (13~30Hz) have height phase with the reaction time
Guan Xing, can as the reaction time objective prediction index.Therefore the embodiment of the present invention will extract above-mentioned 3 kinds of frequencies by wavelet transformation
The wavelet coefficient energy value of section is used as brain electrical feature parameter.Its calculating process is as follows：
(1) the EEG signals gathered to experiment carry out overall filtering process with 0-35Hz bandwidth, remove power frequency electric and part
The artefact such as myoelectricity composition is disturbed.
(2) for the EEG signals after filtered processing, u (n) is designated as, then its wavelet transformation is defined as：
In formula,For wavelet function；I is frequency factor；M is the time-shifting factor；N is signal duration.
(3), in order to carry out finite layer decomposition to the signal u (n) after wavelet transformation, the embodiment of the present invention introduces Mallat algorithms
[9], i.e.,
In formula, AH is approximation component；Ci is the details coefficients under different scale；H is Decomposition order, and it is 3 that the number of plies is taken in text.
Therefore, the wavelet coefficient for obtaining 3 kinds of different frequency ranges of θ, α, β can be decomposed by above-mentioned finite layer.
(4) the energy value of wavelet coefficient can reflect the frequency domain character of EEG signals, therefore extract corresponding to above-mentioned wave band
Energy value is used as brain electrical feature parameter.
In formula, PXFor the energy value of corresponding band；SX(t) it is the wavelet coefficient of corresponding band；T is the time；hiFor respective tones
The amplitude of section.
According to (1)-(4) step the EEG signals of q electrode are handled, then accordingly obtain 3 × q brain electrical features ginseng
Number, is designated as xi(i=1,2 ..., 3 × q).Because the dimension of every brain electrical feature parameter is different, brain electrical feature parameter is pressed into formula
(4) it is normalized, so that the numerical value of brain electrical feature parameter is between [0,1], to eliminate noise present in data.
In formula, ximaxWith ximinFor brain electrical feature parameter xiMaxima and minima.
2.2 forecast models based on BP neural network are built
Using the brain electrical feature parameter of EMUs driver as input layer, using reaction time predicted value as output layer, build
BP neural network forecast model containing 1 hidden layer.Wherein input layer number by EMUs driver brain electrical feature parameter
3 × q of number is determined；Hidden layer node number s is then preferentially chosen by model training result；Output layer node number is 1, its
Structural representation is as shown in Figure 4.
If input layer is respectively to hidden layer, hidden layer to the connection weight coefficient between output layer and biasing in the model
wik,wk1,bik,bk1(i=1,2 ..., 3 × q, k=1,2 ..., s), for arbitrary node o the appointing to hidden layer of input layer
Meaning node p output：
yop=f (xiwop+bop) ⑸
In formula, f () is Sigmoid functions, i.e.,
Output layer output result is：
In formula, yiFor the output of neutral net, i.e. model prediction result；W1For the connection weight number system of input layer to hidden layer
Matrix number；W2For the connection weight number system matrix number of hidden layer to output layer；xiFor EMUs driver's brain electrical feature parameter；b1To be defeated
Enter layer to the bias matrix of hidden layer；b2For the bias matrix of hidden layer to output layer.
For EMUs driver's experiment sample Xi=(x1,x2,...xi,...,x3×q；yi), wherein yiRepresent EMUs
In reaction time obtained by driver's ith button, it is inputted neutral net and is trained, the network output of the experiment sample is missed
Difference is defined as：
In formula,For EMUs driver's real reaction time.And the overall error to all experiment samples of EMUs driver is determined
Justice is：
In formula, N is experiment sample number.Adjustment connection flexible strategy training neutral net, Zhi Daozong are inversely propagated by error
Untill error reaches minimum, so as to complete neural metwork training.
2.3 prediction effect assessment indicators
It is actual with testing the gathered reaction time to evaluate the reaction time predicted value obtained by BP neural network model
It is worth the quality of fitting degree, the embodiment of the present invention is chosen maximum absolute error M1 and relative mean square error M2 and surveyed as prediction effect
Comment index.
3 beneficial effects and analysis
The eeg data in each reaction time is gathered for experiment, is obtained through normalization using Section 2.1 of method
96 brain electrical feature parameters after processing, as the input pointer of BP neural network forecast model.In order to verify that brain electricity is special
Levying parameter and reaction time has correlation, and the embodiment of the present invention is entered using Pearson came correlation test to correlation between the two
Performing check, so as to provide theoretical premise for reaction time prediction.
(1) correlation analysis in brain electrical feature parameter and reaction time.During by tri- brain electrical feature parameters of θ, α, β with reaction
Between carry out Pearson came correlation test, it is as shown in the table for its result.
Each brain electrical feature parameter of table 1 and the correlation in reaction time
Note：| r | be the absolute value of Pearson came relative coefficient, | r | the size of value reflect correlation between the two
Power.* represents that correlation is extremely notable in 0.01 significance；* the correlation under 0.05 significance is represented
Significantly.
As can be seen from the table, three brain electrical feature parameters are presented significantly correlated with the reaction time to varying degrees
Property, EEG signals in forefathers' research are demonstrated again there is high correlation with the reaction time.From the point of view of correlation power, show
Right brain electrical feature parameter alpha and reaction time are more relevant compared with other two, while the most notable.
It in addition, there will be research and show that corticocerebral activity change can directly reflect the driving state of mind, its brain telecommunications
Number with the state of mind of driver is (tired, drowsiness etc.) has high correlation.In the classical cognitive psychology examination such as visual detection
Show that the change for also demonstrating EEG signals and reaction time have high correlation in testing, and this research passes through brain electrical feature
Parameter and the correlation analysis in reaction time, are demonstrated in actual driving task operating process, the reaction with EMUs driver
Time also has correlation.
(2) model output result is analyzed.In order to determine the optimum network structure of BP neural network forecast model, while in order to
The globality and representativeness of training sample are embodied, the embodiment of the present invention is taken out at random respectively to the experiment sample of 20 EMUs drivers
7 samples are taken, to resulting 140 samples as training sample, by the final produced maximum absolute error that predicts the outcome
Evaluation criterion is used as with relative mean square error.After multiple repetition training, it is determined that optimum network structure is 96-20-1
(input layer-hidden layer-output layer).With reference to optimum network structure, to above-mentioned 140 samples, (sample for randomly selecting 75% is made
For training sample, remainder is used as test sample) substitute into model and re-start training and test, it predicts the outcome such as Fig. 5 and Fig. 6 institutes
Show.
Can be seen that from Fig. 5 and Fig. 6, model prediction result is closer to actual result, show the forecast result of model compared with
It is good.Therefore each driver's experiment sample is equally used and randomly selects 75% as training sample, the remaining side as test sample
Method, is trained to neural network model with testing respectively, while the precision height in order to understand carried model, the present invention is used
Existing bayes predictive model is predicted to the reaction time, and resulting predicts the outcome as shown in table 2.
Each driver of table 2 is resulting in forecast model to predict the outcome
In general, using maximum absolute error average value in the result obtained by artificial neural network
(11.01%) it is above bayes predictive model (14.6%, 10.7%) with relative mean square error average value (8.15%).Together
When, reaction time of each driver is predicted the outcome and be can be seen that, it is equal using the maximum absolute error obtained by artificial neural network
Less than 15%, and maximum relative mean square error is 10.89%, and minimum relative mean square error is 6.93%；And use Bayes pre-
Survey in the result that model is obtained, maximum absolute error is between 11% to 20%, and maximum relative mean square error is 15.36%,
Minimum relative mean square error is 8.14%, illustrates that neural network prediction precision is higher than bayes predictive model.Therefore originally
Inventive embodiments, which carry forecast model, has reliability, can Accurate Prediction EMUs driver to the accident reaction time so that
Effectively reduce accident rate.
From the above it can be seen that motor-car drive simulating experiment of the embodiment of the present invention based on 2h, with regard to EMUs driver to machine event of dashing forward
The prediction in reaction time is studied, and its useful achievement and conclusion are as follows：
(1) the EEG signals based on EMUs driver, are extracted using wavelet transformation and can be used for EMUs driver reaction level to survey
Comment and reaction time prediction the electric indexs of 3 brains of θ, α, β.With reference to BP neural network, when constructing a kind of EMUs driver reaction
Between forecast model.
(2) it can be seen that from result, 3 brain electrical feature parameters are respectively provided with significant correlation with the reaction time, illustrate EEG signals
The state of mind of driver can be directly reacted, so as to provide research foundation for the prediction in reaction time.
(3) final result shows, when the reaction time that the driver of model prediction is stimulated random signal is with driver's real reaction
Between maximum absolute error be 11.01% (1.75%), and relative mean square error be 8.51% (1.37%), less than other prediction
Model, shows that this method has degree of precision.
The embodiment of the present invention realizes Accurate Prediction of the EMUs driver to the accident reaction time, the achievement in research pair
The vehicle-mounted dangerous driving condition early warning system exploitation of real-time EMUs driver provides theoretical foundation with design.From now on can be to the party
Applicability of the method under motor-car actual motion state gives further checking research.
It will appreciated by the skilled person that realizing all or part of step/units/modules of above-described embodiment
It can be completed by the related hardware of programmed instruction, foregoing routine can be stored in computer read/write memory medium, should
Upon execution, perform includes the step of correspondence in above-described embodiment each unit to program；And foregoing storage medium includes：ROM、
RAM, magnetic disc or laser disc etc. are various can be with the medium of store program codes.
Particular embodiments described above, has been carried out further in detail to the purpose of the present invention, technical scheme and beneficial effect
Describe in detail it is bright, should be understood that the foregoing is only the present invention specific embodiment, be not intended to limit the invention, it is all
Within the spirit and principles in the present invention, any modification, equivalent substitution and improvements done etc., should be included in the guarantor of the present invention
Within the scope of shield.
Claims (9)
1. a kind of time of driver's reaction Forecasting Methodology, it is characterised in that methods described includes step：
Obtain the EEG signals of driver；
The brain electrical feature parameter in the EEG signals after filtered processing is extracted by wavelet transformation；
Using the brain electrical feature parameter of driver as input layer, using the reaction time as output layer, BP neural network prediction is built
Model.
2. time of driver's reaction Forecasting Methodology as claimed in claim 1, it is characterised in that the filtering process is specifically wrapped
Include：EEG signals carry out overall filtering process with 0-35Hz bandwidth.
3. time of driver's reaction Forecasting Methodology as claimed in claim 2, it is characterised in that brain electricity is extracted by wavelet transformation
Characteristic parameter is specifically included：
A, for the EEG signals after filtered processing, be designated as u (n), then its wavelet transformation is defined as：
In formula,For wavelet function；I is frequency factor；M is the time-shifting factor；N is signal duration；
Signal u (n) after b, wavelet transformation carries out finite layer decomposition, using Mallat algorithms：
<mrow>
<mi>u</mi>
<mrow>
<mo>(</mo>
<mi>n</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<msub>
<mi>A</mi>
<mi>H</mi>
</msub>
<mo>+</mo>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>H</mi>
</munderover>
<msub>
<mi>C</mi>
<mi>i</mi>
</msub>
</mrow>
In formula, AH is approximation component；Ci is the details coefficients under different scale；H is Decomposition order；Decomposed by above-mentioned finite layer
Obtain the wavelet coefficient of 3 kinds of different frequency ranges of θ, α, β；
C, the energy value to above-mentioned frequency extraction wavelet coefficient are used as brain electrical feature parameter：
<mrow>
<msub>
<mi>P</mi>
<mi>X</mi>
</msub>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mi>t</mi>
</mfrac>
<msup>
<mrow>
<mo>|</mo>
<msub>
<mi>S</mi>
<mi>X</mi>
</msub>
<mrow>
<mo>(</mo>
<mi>t</mi>
<mo>)</mo>
</mrow>
<mo>|</mo>
</mrow>
<mn>2</mn>
</msup>
<mi>d</mi>
<mi>t</mi>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mi>n</mi>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>n</mi>
</munderover>
<msup>
<mrow>
<mo>|</mo>
<msub>
<mi>h</mi>
<mi>i</mi>
</msub>
<mo>|</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
In formula, PXFor the energy value of corresponding band；SX(t) it is the wavelet coefficient of corresponding band；T is the time；hiFor corresponding band
Amplitude；
D, the EEG signals to q electrode are handled, and accordingly obtain 3 × q brain electrical feature parameters.
4. time of driver's reaction Forecasting Methodology as claimed in claim 3, it is characterised in that methods described also includes：To institute
Obtained brain electrical feature parameter is normalized by formula, makes the numerical value of brain electrical feature parameter between [0,1], to eliminate number
The noise present in：
<mrow>
<mi>x</mi>
<mo>=</mo>
<mfrac>
<mrow>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
</msub>
</mrow>
<mrow>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
</mrow>
</msub>
<mo>-</mo>
<msub>
<mi>x</mi>
<mrow>
<mi>i</mi>
<mi>m</mi>
<mi>i</mi>
<mi>n</mi>
</mrow>
</msub>
</mrow>
</mfrac>
</mrow>
In formula, xi maxWith xi minFor brain electrical feature parameter xiMaxima and minima.
5. the time of driver's reaction Forecasting Methodology as described in any one of Claims 1-4, it is characterised in that build BP nerves
Network Prediction Model is specifically included：Using the brain electrical feature parameter of driver as input layer, using reaction time predicted value as defeated
Go out layer, build the BP neural network forecast model containing 1 hidden layer；Wherein, input layer number is special by the brain electricity of driver
Levy number of parameters decision；Hidden layer node number s is then preferentially chosen by model training result；Output layer node number is 1.
6. time of driver's reaction Forecasting Methodology as claimed in claim 5, it is characterised in that input layer is to hidden in the model
It is respectively w containing layer, hidden layer to the connection weight coefficient between output layer and biasingik,wk1,bik,bk1(i=1,2 ..., 3 ×
Q, k=1,2 ..., s), for input layer arbitrary node o to hidden layer arbitrary node p output：
yop=f (xiwop+bop)
In formula, f () is Sigmoid functions, i.e.,
<mrow>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mi>&sigma;</mi>
<mo>)</mo>
</mrow>
<mo>=</mo>
<mfrac>
<mn>1</mn>
<mrow>
<mn>1</mn>
<mo>+</mo>
<msup>
<mi>e</mi>
<mrow>
<mo>-</mo>
<mi>&sigma;</mi>
</mrow>
</msup>
</mrow>
</mfrac>
</mrow>
Output layer output result is：
<mrow>
<msub>
<mi>y</mi>
<mi>i</mi>
</msub>
<mo>=</mo>
<mi>f</mi>
<mrow>
<mo>(</mo>
<msubsup>
<mi>W</mi>
<mn>2</mn>
<mi>T</mi>
</msubsup>
<mo>(</mo>
<mrow>
<mi>f</mi>
<mrow>
<mo>(</mo>
<mrow>
<msubsup>
<mi>W</mi>
<mn>1</mn>
<mi>T</mi>
</msubsup>
<msub>
<mi>x</mi>
<mi>i</mi>
</msub>
</mrow>
<mo>)</mo>
</mrow>
<mo>+</mo>
<msub>
<mi>b</mi>
<mn>1</mn>
</msub>
</mrow>
<mo>)</mo>
<mo>+</mo>
<msub>
<mi>b</mi>
<mn>2</mn>
</msub>
<mo>)</mo>
</mrow>
</mrow>
In formula, yiExported for Neural Network model predictive result；W1For the connection weight number system matrix number of input layer to hidden layer；W2
For the connection weight number system matrix number of hidden layer to output layer；xiFor driver's brain electrical feature parameter；b1For input layer to hidden layer
Bias matrix；b2For the bias matrix of hidden layer to output layer.
7. time of driver's reaction Forecasting Methodology as claimed in claim 6, it is characterised in that the brain electricity of the acquisition driver
Signal is specifically included：Gather the EEG signals of the driver of predetermined number.
8. time of driver's reaction Forecasting Methodology as claimed in claim 7, it is characterised in that eeg signal acquisition frequency is
10Hz；Electroencephalogramdata data collector continuous acquisition driver's eeg data.
9. time of driver's reaction Forecasting Methodology as claimed in claim 7, it is characterised in that obtained by BP neural network model
Comparison of the reaction time predicted value with testing gathered reaction time actual value fitting degree use maximum absolute error M1
Prediction effect assessment indicator is used as with relative mean square error M2：
<mrow>
<msub>
<mi>M</mi>
<mn>1</mn>
</msub>
<mo>=</mo>
<munder>
<mrow>
<mi>m</mi>
<mi>a</mi>
<mi>x</mi>
</mrow>
<mi>i</mi>
</munder>
<mrow>
<mo>|</mo>
<mfrac>
<mrow>
<msub>
<mi>y</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<msubsup>
<mi>y</mi>
<mi>i</mi>
<mo>*</mo>
</msubsup>
</mrow>
<msub>
<mi>y</mi>
<mi>i</mi>
</msub>
</mfrac>
<mo>|</mo>
</mrow>
<mo>&times;</mo>
<mn>100</mn>
<mi>%</mi>
</mrow>
<mrow>
<msub>
<mi>M</mi>
<mn>2</mn>
</msub>
<mo>=</mo>
<msqrt>
<mrow>
<mfrac>
<mn>1</mn>
<mi>N</mi>
</mfrac>
<munderover>
<mo>&Sigma;</mo>
<mrow>
<mi>i</mi>
<mo>=</mo>
<mn>1</mn>
</mrow>
<mi>N</mi>
</munderover>
<msup>
<mrow>
<mo>(</mo>
<mfrac>
<mrow>
<msub>
<mi>y</mi>
<mi>i</mi>
</msub>
<mo>-</mo>
<msubsup>
<mi>y</mi>
<mi>i</mi>
<mo>*</mo>
</msubsup>
</mrow>
<msub>
<mi>y</mi>
<mi>i</mi>
</msub>
</mfrac>
<mo>)</mo>
</mrow>
<mn>2</mn>
</msup>
</mrow>
</msqrt>
<mo>&times;</mo>
<mn>100</mn>
<mi>%</mi>
<mo>.</mo>
</mrow>
2
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710391410.7A CN107095670A (en) | 2017-05-27 | 2017-05-27 | Time of driver's reaction Forecasting Methodology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201710391410.7A CN107095670A (en) | 2017-05-27 | 2017-05-27 | Time of driver's reaction Forecasting Methodology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN107095670A true CN107095670A (en) | 2017-08-29 |
Family
ID=59660411
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201710391410.7A Pending CN107095670A (en) | 2017-05-27 | 2017-05-27 | Time of driver's reaction Forecasting Methodology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107095670A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107595302A (en) * | 2017-09-06 | 2018-01-19 | 西安交通大学 | A kind of device and method that user's state of mind is monitored using EEG signals |
| CN108836321A (en) * | 2018-05-03 | 2018-11-20 | 江苏师范大学 | A kind of EEG signals preprocess method based on adaptive noise cancel- ation system |
| CN108874137A (en) * | 2018-06-15 | 2018-11-23 | 北京理工大学 | A kind of gesture motion based on EEG signals is intended to the universal model of detection |
| CN109204307A (en) * | 2018-10-30 | 2019-01-15 | 吉林大学 | Obtain method, storage medium and the system of adaptive anticollision early warning distance |
| CN110313916A (en) * | 2019-07-17 | 2019-10-11 | 军事科学院系统工程研究院军需工程技术研究所 | A kind of patient with diabetic feet plantar pressure characteristic recognition method |
| CN111803065A (en) * | 2020-06-23 | 2020-10-23 | 北方工业大学 | Dangerous traffic scene identification method and system based on electroencephalogram data |
| CN112365705A (en) * | 2020-08-27 | 2021-02-12 | 招商局重庆交通科研设计院有限公司 | Method for determining road traffic volume |
| CN113633296A (en) * | 2021-10-15 | 2021-11-12 | 西南交通大学 | Reaction time prediction model construction method, device, equipment and readable storage medium |
| CN114081504A (en) * | 2021-11-23 | 2022-02-25 | 青岛理工大学 | Driving intention identification method and system based on electroencephalogram signals |
| CN114224359A (en) * | 2021-12-18 | 2022-03-25 | 西南交通大学 | Electroencephalogram fatigue index determination method for high-speed rail dispatcher |
| CN114343661A (en) * | 2022-03-07 | 2022-04-15 | 西南交通大学 | Method, device and equipment for estimating reaction time of high-speed rail driver and readable storage medium |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1253762A (en) * | 1999-11-19 | 2000-05-24 | 中国科学院上海生理研究所 | Full-automatic quantitive detection device of epileptoid discharge in electroencephalogram |
| CN101221554A (en) * | 2008-01-25 | 2008-07-16 | 北京工业大学 | Brain wave characteristic extraction method based on wavelet translation and BP neural network |
| CN101352337A (en) * | 2007-12-25 | 2009-01-28 | 天津大学 | Method for capturing signals and extracting characteristics of stand imagination action brain wave |
| CN101690666A (en) * | 2009-10-13 | 2010-04-07 | 北京工业大学 | Driving working load calculation method of automobile driver |
| CN102254226A (en) * | 2011-06-15 | 2011-11-23 | 河海大学 | Method for identifying human activities based on BP (Back Propagation) neural network in intelligent family environment |
| CN103431859A (en) * | 2012-01-10 | 2013-12-11 | 西安交通大学 | Experimental method for determining brain load in multitask visual cognition |
| CN103845052A (en) * | 2014-02-20 | 2014-06-11 | 清华大学 | Human body faint early warning method based on acquired electroencephalogram signals |
| CN104665849B (en) * | 2014-12-11 | 2017-03-01 | 西南交通大学 | A kind of high ferro dispatcher based on the interaction of many physiological signals multi-model stress detection method |
-
2017
- 2017-05-27 CN CN201710391410.7A patent/CN107095670A/en active Pending
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1253762A (en) * | 1999-11-19 | 2000-05-24 | 中国科学院上海生理研究所 | Full-automatic quantitive detection device of epileptoid discharge in electroencephalogram |
| CN101352337A (en) * | 2007-12-25 | 2009-01-28 | 天津大学 | Method for capturing signals and extracting characteristics of stand imagination action brain wave |
| CN101221554A (en) * | 2008-01-25 | 2008-07-16 | 北京工业大学 | Brain wave characteristic extraction method based on wavelet translation and BP neural network |
| CN101690666A (en) * | 2009-10-13 | 2010-04-07 | 北京工业大学 | Driving working load calculation method of automobile driver |
| CN102254226A (en) * | 2011-06-15 | 2011-11-23 | 河海大学 | Method for identifying human activities based on BP (Back Propagation) neural network in intelligent family environment |
| CN103431859A (en) * | 2012-01-10 | 2013-12-11 | 西安交通大学 | Experimental method for determining brain load in multitask visual cognition |
| CN103845052A (en) * | 2014-02-20 | 2014-06-11 | 清华大学 | Human body faint early warning method based on acquired electroencephalogram signals |
| CN104665849B (en) * | 2014-12-11 | 2017-03-01 | 西南交通大学 | A kind of high ferro dispatcher based on the interaction of many physiological signals multi-model stress detection method |
Non-Patent Citations (3)
| Title |
|---|
| CHIN-TENG LIN ET AL: "Wireless and Wearable EEG System for Evaluating Driver Vililance", 《IEEE TRANSACTION ON BIOMEDICAL CIRCUITS AND SYSTEMS》 * |
| 张骏等: "基于SVR模型的驾驶简单反应时间预测方法", 《中国公路学报》 * |
| 林相波: "脑电信号特征分析和特征提取的研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107595302A (en) * | 2017-09-06 | 2018-01-19 | 西安交通大学 | A kind of device and method that user's state of mind is monitored using EEG signals |
| CN108836321A (en) * | 2018-05-03 | 2018-11-20 | 江苏师范大学 | A kind of EEG signals preprocess method based on adaptive noise cancel- ation system |
| CN108874137A (en) * | 2018-06-15 | 2018-11-23 | 北京理工大学 | A kind of gesture motion based on EEG signals is intended to the universal model of detection |
| CN109204307A (en) * | 2018-10-30 | 2019-01-15 | 吉林大学 | Obtain method, storage medium and the system of adaptive anticollision early warning distance |
| CN110313916A (en) * | 2019-07-17 | 2019-10-11 | 军事科学院系统工程研究院军需工程技术研究所 | A kind of patient with diabetic feet plantar pressure characteristic recognition method |
| CN111803065B (en) * | 2020-06-23 | 2023-12-26 | 北方工业大学 | Dangerous traffic scene identification method and system based on electroencephalogram data |
| CN111803065A (en) * | 2020-06-23 | 2020-10-23 | 北方工业大学 | Dangerous traffic scene identification method and system based on electroencephalogram data |
| CN112365705A (en) * | 2020-08-27 | 2021-02-12 | 招商局重庆交通科研设计院有限公司 | Method for determining road traffic volume |
| CN113633296A (en) * | 2021-10-15 | 2021-11-12 | 西南交通大学 | Reaction time prediction model construction method, device, equipment and readable storage medium |
| CN114081504A (en) * | 2021-11-23 | 2022-02-25 | 青岛理工大学 | Driving intention identification method and system based on electroencephalogram signals |
| CN114081504B (en) * | 2021-11-23 | 2024-03-01 | 青岛理工大学 | Driving intention recognition method and system based on electroencephalogram signals |
| CN114224359A (en) * | 2021-12-18 | 2022-03-25 | 西南交通大学 | Electroencephalogram fatigue index determination method for high-speed rail dispatcher |
| CN114343661A (en) * | 2022-03-07 | 2022-04-15 | 西南交通大学 | Method, device and equipment for estimating reaction time of high-speed rail driver and readable storage medium |
| CN114343661B (en) * | 2022-03-07 | 2022-05-27 | 西南交通大学 | Method, device and equipment for estimating reaction time of driver in high-speed rail and readable storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107095670A (en) | Time of driver's reaction Forecasting Methodology | |
| CN108052888B (en) | A kind of driver's replacement system | |
| CN106913333B (en) | Method for selecting sensitivity characteristic index of continuous attention level | |
| CN103035140B (en) | Automobile driver dangerous driving behavior correction and evaluation system | |
| CN105877766A (en) | Mental state detection system and method based on multiple physiological signal fusion | |
| CN106956591A (en) | A kind of system for being used to judge that human pilot drives authority | |
| Shaw et al. | Influence of parent discipline style on delinquent behaviour: The mediating role of control orientation | |
| CN113743471B (en) | Driving evaluation method and system | |
| CN107577343A (en) | It is a kind of based on the notice of haptic device and electroencephalogramsignal signal analyzing training and evaluating apparatus | |
| CN105405253B (en) | Method and apparatus for monitoring fatigue state of driver | |
| CN111311093B (en) | Road intersection risk assessment and early warning method based on driver physiological data | |
| CN108903936A (en) | The Intelligent mining helmet control method merged based on human body information and environmental information | |
| CN107095671A (en) | Driving fatigue recognition methods and system | |
| CN110123266B (en) | Maneuvering decision modeling method based on multi-modal physiological information | |
| CN113753059A (en) | Method for predicting takeover capacity of driver under automatic driving system | |
| Tran et al. | Predicting driver’s work performance in driving simulator based on physiological indices | |
| Wang et al. | Cognitive Load Identification of Pilots Based on Physiological‐Psychological Characteristics in Complex Environments | |
| CN111657923A (en) | Method and system for testing danger perception capability of driver | |
| Miyaji et al. | Driver’s cognitive distraction detection using AdaBoost on pattern recognition basis | |
| Rahman et al. | EEG based brain alertness monitoring by statistical and artificial neural network approach | |
| CN117290781A (en) | Driver KSS grade self-evaluation training method for DDAW system test | |
| CN115994844A (en) | Driving safety evaluation method under road tunnel disaster scene | |
| CN211094182U (en) | Driver alertness detection mechanism based on electroencephalogram signals | |
| Wang et al. | Physiological Characteristics and Nonparametric Test for Master‐Slave Driving Task’s Mental Workload Evaluation in Mountain Area Highway at Night | |
| Wang et al. | [Retracted] Prediction Model for Driver Reaction Time Based on PSO‐BP Neural Network Model |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170829 |
|
| RJ01 | Rejection of invention patent application after publication |