CN114783552A - Method and system for detecting fatigue of operator before post - Google Patents
Method and system for detecting fatigue of operator before post Download PDFInfo
- Publication number
- CN114783552A CN114783552A CN202210453967.XA CN202210453967A CN114783552A CN 114783552 A CN114783552 A CN 114783552A CN 202210453967 A CN202210453967 A CN 202210453967A CN 114783552 A CN114783552 A CN 114783552A
- Authority
- CN
- China
- Prior art keywords
- data
- fatigue
- data set
- operator
- experiment
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/40—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for data related to laboratory analysis, e.g. patient specimen analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/06—Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling
- G06Q10/063—Operations research, analysis or management
- G06Q10/0639—Performance analysis of employees; Performance analysis of enterprise or organisation operations
- G06Q10/06393—Score-carding, benchmarking or key performance indicator [KPI] analysis
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Human Resources & Organizations (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Primary Health Care (AREA)
- General Health & Medical Sciences (AREA)
- Epidemiology (AREA)
- Educational Administration (AREA)
- Evolutionary Computation (AREA)
- Strategic Management (AREA)
- Entrepreneurship & Innovation (AREA)
- Economics (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Game Theory and Decision Science (AREA)
- Pathology (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- Marketing (AREA)
- Operations Research (AREA)
- Quality & Reliability (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
The invention provides a method and a system for detecting the fatigue of an operator before post, comprising the following steps: designing a target simulation experiment based on an experiment method, and acquiring an experiment data set in the process of the target simulation experiment; analyzing the heart rate variability data to obtain a data index related to the heart rate variability data; taking the physiological data of the eye movement as a fatigue label, checking the physiological data by using subjective scale data and mental movement warning task data, taking a data index as a fatigue characteristic, and extracting and identifying the characteristic of an experimental data set to obtain a preprocessed data set; and dividing the preprocessed data set into a training data set and a testing data set, training a preset fatigue detection model on the training data set by using a machine learning method, and testing on the testing data set to obtain the trained fatigue detection model. The invention solves the technical problems of inaccurate measurement and low measurement efficiency in the prior art.
Description
Technical Field
The invention relates to the technical field of fatigue detection, in particular to a method and a system for detecting the fatigue of an operator before post.
Background
Fatigue is considered to be an important factor in causing accidents, injuries and deaths to a greater extent, which means that tired people are more likely to make unsafe behaviors. Fatigue has a significant impact on many industries, such as the transportation industry, e.g., road, aviation, and navigation, and other professional environments (e.g., hospitals, machinery plants, etc.), particularly when it comes to shifts and unscheduled working hours. There are almost every time there is fatigue, both during working hours and during rest hours, and the risk of accidents and injuries is therefore increased. Fatigue can have negative effects on humans, such as bradykinesia, side-length response times, inattention, easier wrong decision making, etc., which have been observed in many compelling accidents.
Fatigue or lethargy detection is an important research topic, and many techniques for fatigue detection have been successfully developed: scale measurements, performance measurements, behavior measurements, physiological data measurements, and the like. The Scale measurement is that an operator evaluates self fatigue and fills in a karolinskyhook Scale (KSS) and a Stanford Sleepiness Scale (SSS), and the measurement mode depends on subjective feeling of the operator and ignores some slight changes of fatigue; the performance measurement refers to the evaluation of Task performance of operators, and common Psychomotor Vigilance Task (PVT) tests and the like; the behavior measurement means that a camera is used for capturing facial actions and expressions of an operator, and the facial expressions such as blinking, yawning and nodding are recognized based on a machine vision technology, the method needs to be capable of replenishing the behavior characteristics of the operator in real time, and if the operator intentionally inhibits the behavior characteristics, the efficiency of the method is greatly reduced; physiological data measurements are highly correlated with the mental state of the operator and are most sensitive to fatigue detection, but require the operator to wear necessary sensory equipment, such as electroencephalograms (EEG), Electrocardiograms (ECG), Electrooculograms (EOG), and the like. Therefore, the existing fatigue measurement technology has the problems of inaccurate measurement and low measurement efficiency.
Disclosure of Invention
In view of this, the present invention aims to provide a method and a system for detecting fatigue of an operator before post, so as to alleviate the technical problems of inaccurate measurement and low measurement efficiency in the prior art.
In a first aspect, an embodiment of the present invention provides a method for detecting pre-post fatigue of an operator, including: designing a target simulation experiment based on an experiment method, and acquiring an experiment data set in the process of the target simulation experiment; the target simulation experiment is an operator fatigue experiment which is developed on a multi-attribute task platform at different time periods based on a simulation scene; the experimental data set includes: heart rate variability data, physiological data of eye movements, subjective scale data, and psychomotor alert task data; analyzing the heart rate variability data to obtain a data index related to the heart rate variability data; taking the physiological data of the eye movement as a fatigue label, checking by using the subjective scale data and the mental exercise warning task data, taking the data index as a fatigue feature, and extracting and identifying the feature of the experimental data set to obtain a preprocessed data set; dividing the preprocessed data set into a training data set and a testing data set, training a preset fatigue detection model on the training data set by using a machine learning method, and testing on the testing data set to obtain a trained fatigue detection model; and performing on-duty fatigue detection on the operator to be detected by using the trained fatigue detection model.
Further, analyzing the heart rate variability data to obtain a data index of the heart rate variability data, including: dividing the heart rate variability data into a plurality of time slices; and respectively carrying out time domain analysis, frequency domain analysis and nonlinear analysis on the data in each time slice to obtain a data index about the heart rate variability data.
Further, the physiological data of the eye movement comprises a total blink time and a unit blink time; performing a test using the subjective scale data and the psychomotor vigilance task data, comprising: performing correlation analysis on the total blink time and the subjective scale data, and verifying the correlation between the unit blink time and subjective fatigue; and carrying out correlation analysis on the total blink time and the mental exercise alert task data, and verifying the correlation between the unit blink time and the mental exercise alert task data.
Further, the subjective scale data comprises karlolin sca hypersomnia scale data and stanford hypersomnia scale data; the psychomotor alert task data includes a reaction time.
Further, the method further comprises: and carrying out correlation analysis on the unit blink time and the data indexes, and extracting the data indexes with the highest correlation coefficients as the fatigue features.
Further, the fatigue detection model after training is utilized to carry out on-duty fatigue detection on the operator to be detected, and the method comprises the following steps: acquiring heart rate variability data of the operator to be detected; and inputting the heart rate variability data of the operator to be detected into the fatigue detection model after training to obtain the pre-post fatigue detection result of the operator to be detected.
In a second aspect, an embodiment of the present invention further provides a system for detecting fatigue of an operator before post, including: the system comprises an experiment module, an analysis module, an extraction module, a training module and a detection module; the experimental module is used for designing a target simulation experiment based on an experimental method and acquiring an experimental data set in the target simulation experiment process; the target simulation experiment is an operator fatigue experiment which is developed on a multi-attribute task platform at different time periods based on a simulation scene; the experimental data set comprises: heart rate variability data, physiological data of eye movements, subjective scale data, and psychomotor alert task data; the analysis module is used for analyzing the heart rate variability data to obtain a data index related to the heart rate variability data; the extraction module is used for taking the physiological data of the eye movement as a fatigue label, checking the subjective scale data and the mental exercise warning task data, taking the data index as a fatigue feature, and extracting and identifying the feature of the experimental data set to obtain a preprocessed data set; the training module is used for dividing the preprocessed data set into a training data set and a testing data set, training a preset fatigue detection model on the training data set by using a machine learning method, and testing on the testing data set to obtain a trained fatigue detection model; and the detection module is used for performing on-duty fatigue detection on the operator to be detected by using the trained fatigue detection model.
Further, the detection module is further configured to: acquiring heart rate variability data of the operator to be detected; and inputting the heart rate variability data of the operator to be detected into the fatigue detection model after training to obtain the pre-post fatigue detection result of the operator to be detected.
In a third aspect, an embodiment of the present invention further provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, where the processor implements the steps of the method according to the first aspect when executing the computer program.
In a fourth aspect, an embodiment of the present invention further provides a computer-readable medium having a non-volatile program code executable by a processor, where the program code causes the processor to execute the method described in the first aspect.
The invention provides a method and a system for detecting the fatigue of an operator before post, which comprehensively use the experimental methods of scale measurement, behavior measurement and physiological data measurement to determine a model for detecting the fatigue of the operator before post, the model can accurately detect the fatigue state of the operator, the detection process is simple and efficient, and the technical problems of inaccurate measurement and low measurement efficiency in the prior art are solved.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the prior art descriptions will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and other drawings can be obtained by those skilled in the art without creative efforts.
Fig. 1 is a flowchart of a method for detecting fatigue of an operator before post according to an embodiment of the present invention;
FIG. 2 is a flowchart of another method for detecting fatigue of an operator before post according to an embodiment of the present invention;
FIG. 3 is a technical route diagram of data analysis of a method for detecting fatigue of an operator before post according to an embodiment of the present invention;
FIG. 4 is a general experimental flow chart for civil aircraft pilot design according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a system for detecting pre-post fatigue of an operator according to an embodiment of the present invention;
fig. 6 is a schematic diagram of a fatigue detection apparatus according to an embodiment of the present invention.
Detailed Description
The technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings, and it should be understood that the described embodiments are some, but not all embodiments of the present invention. All other embodiments, which can be obtained by a person skilled in the art without making any creative effort based on the embodiments in the present invention, belong to the protection scope of the present invention.
The first embodiment is as follows:
fig. 1 is a flowchart of a method for detecting fatigue of an operator before post according to an embodiment of the present invention. As shown in fig. 1, the method specifically includes the following steps:
step S102, designing a target simulation experiment based on an experimental method, and acquiring an experiment data set in the process of the target simulation experiment; the target simulation experiment is an operator fatigue experiment which is developed on a multi-attribute task platform at different time periods based on a simulation scene; the experimental data set includes: heart Rate Variability (HRV), physiological data of eye movements, subjective scale data, and Psychomotor Vigilance Task data (PVT).
In the embodiment of the invention, firstly, operators participating in a target simulation experiment are deprived of sleep, and then fatigue experiments are carried out on a multi-attribute task platform in different time periods based on simulation scenes, wherein the main purpose is to induce different fatigue levels of the operators.
Optionally, in an embodiment of the present invention, the physiological data of eye movement comprises total blink time and unit blink time; subjective Scale data included karolinsca Sleepiness Scale data (Karolinska Sleepiness Scale, KSS) and Stanford Sleepiness Scale data (SSS); the psychomotor alert task data includes a reaction time.
And step S104, analyzing the heart rate variability data to obtain a data index related to the heart rate variability data.
Specifically, the heart rate variability data is divided into a plurality of time slices; and respectively carrying out time domain analysis, frequency domain analysis and nonlinear analysis on the data in each time slice to obtain a data index about the heart rate variability data.
And S106, taking the physiological data of the eye movement as a fatigue label, checking by using the subjective scale data and the mental and physical vigilance task data, taking the data index as a fatigue characteristic, and extracting and identifying the characteristic of the experimental data set to obtain a preprocessed data set.
Wherein the checking with the subjective scale data and the psychomotor alert task data comprises:
performing correlation analysis on the total blink time and the subjective scale data, and verifying the correlation between unit blink time and subjective fatigue;
and carrying out correlation analysis on the total blink time and the mental movement warning task data, and verifying the correlation between the unit blink time and the mental movement warning task data.
Optionally, in the embodiment of the present invention, the unit blinking time and the data index are subjected to correlation analysis, and the data index with the highest correlation coefficient is extracted as the fatigue feature.
And S108, dividing the preprocessed data set into a training data set and a testing data set, training a preset fatigue detection model on the training data set by using a machine learning method, and testing on the testing data set to obtain the trained fatigue detection model.
Optionally, in an embodiment of the present invention, the preprocessed data set is proportionally divided into a training data set and a testing data set, for example, the proportion may be 7: and 3, ensuring the accuracy of the model in the model training process.
And S110, performing on-duty fatigue detection on the operator to be detected by using the trained fatigue detection model.
Specifically, heart rate variability data of an operator to be detected is obtained; and inputting the heart rate variability data of the operator to be detected into the fatigue detection model after training to obtain the pre-post fatigue detection result of the operator to be detected. The detection result comprises the fatigue grade of the operator to be detected.
The invention provides a method for detecting the fatigue of an operator before post, which comprehensively utilizes the experimental methods of scale measurement, behavior measurement and physiological data measurement to determine a model for detecting the fatigue of the operator before post, wherein the model can accurately detect the fatigue state of the operator, has simple and efficient detection process, and relieves the technical problems of inaccurate measurement and low measurement efficiency in the prior art.
Optionally, fig. 2 is a flowchart of another method for detecting fatigue of an operator before post according to an embodiment of the present invention. As shown in fig. 2, the method specifically includes the following steps:
and step S1, designing an experiment according to the experimental method, and developing the experiment on the multi-attribute task platform at different time periods according to the simulation scene.
Specifically, step S1 further includes two substeps:
and a substep S11 of designing different experimental schemes for different posts, the time of sleep deprivation and the time interval for carrying out experiments in the simulated scene, which need to be considered.
Specifically, the purpose of the sleep deprivation experiment on the tested object is to enable the tested object to reach different levels of drowsiness fatigue states as far as possible, and the experiment under different time intervals can record relevant experimental data under different drowsiness fatigue levels of the tested object.
In particular, the multi-attribute task platform should be capable of including the corresponding meta-tasks for the operation of the operator, i.e. the operation mode of the simulation experiment should include a summary of the operations of the operator. The multi-attribute task platform should at least include: and the system monitors, manages resources, tracks, communicates, plans progress and other meta tasks. The main purpose of the simulation task is to establish a set of standardized operation procedures, so that the subsequent fatigue detection model can be more conveniently used.
In substep S12, the experimental design method includes: the experimental equipment, experimenters, experimental procedures, cautions and the like are selected, and meanwhile, proper experimental equipment is selected, wherein the proper experimental equipment at least comprises electrocardio acquisition equipment and eye movement acquisition equipment, and the used equipment at least can collect related information of HRV data and blink time.
Specifically, for different posts, a worker at the post is selected as an experimenter, and a simulation experiment adaptive to the work of the post is designed for the work of the post. The factors to be considered in the simulation experiment include task duration, task difficulty, task sequence and the like. And the operator needs to perform a number of pre-experiments to eliminate the training effect before the experiment is formally started.
Specifically, the electrocardio acquisition equipment selected by the experiment should adopt equipment with higher precision as much as possible, such as medical-grade related instruments; the eye movement device is selected to accurately record the data related to the blinking of the subject, such as an eye movement apparatus.
In step S2, HRV data, physiological data of eye movements, subjective scale data, and PVT data are collected during the experiment.
Specifically, step S2 further includes three sub-steps:
and a substep S21, performing time domain analysis, frequency domain analysis and nonlinear analysis on the extracted HRV data, dividing time slices, and extracting data indexes in the slices.
Specifically, time slices are divided for the acquired HRV data, and time domain analysis, frequency domain analysis, and nonlinear analysis are performed on each time slice data.
And a substep S22 of extracting blink related data in the physiological data of the eye movement and calculating unit blink time: UBT is BT/AT. Where UBT is the unit blink time, BT represents the sum of the times the eyes are closed over a period of time, and AT represents the total time involved in the calculation.
In particular, during the experiment, the time windows in which the physiological data of the eye movement are acquired and the HRV data are synchronized, and therefore, should correspond to the HRV data when time slicing the physiological data of the eye movement.
In the substep S23, before performing the multi-attribute task platform simulation task each time, filling a Carolina Pasteur sleepiness scale and a Stanford sleepiness scale; PVT test is carried out after each experiment, and the reaction time MRT is obtained, so that reference is provided for the effectiveness of subsequent verification labels.
Alternatively, in an embodiment of the present invention, the reaction time MRT is calculated by the following equation: MRT ═ RT/SN; wherein RT is total reaction time, and SN is stimulation times.
In the embodiment of the present invention, the influence factors of the physiological state of the operator may generate different inducing effects under the same inducing condition due to individual differences of the operator, for example, different experimenters may generate different fatigue changes under the same condition of depriving sleep for 12 hours, and therefore, the experimental results need to be checked every time the experiment is performed. For example, the subjective fatigue level of the test subject is tested by using a somnolence scale before each experiment, and the response level of the test subject is tested by using PVT after each experiment, and the two indexes are also important bases for testing the unit blinking time as a fatigue label in the follow-up test of the embodiment of the invention.
And step S3, performing descriptive statistical analysis on the preprocessed data, taking the physiological data of the eye movement as a label, checking by using the subjective scale data and the PVT data, taking the index of the HRV data after analysis as a feature, and performing feature extraction and recognition to obtain a data set for training.
Specifically, step S3 further includes three sub-steps:
and a substep S31, carrying out correlation analysis on the total blinking time of each experiment and KSS and SSS scales, and verifying the correlation between unit blinking time and subjective fatigue.
Alternatively, the pearson correlation test was used to test the correlation between total blink time for each experiment and the subjective scale, removing experimental data for which correlation was not significant. The subjective fatigue data is an index for classifying the fatigue level in the embodiment of the present invention.
And a substep S32, carrying out correlation analysis on the total blink time of each experiment and the MRT, and verifying the correlation between the unit blink time and the MRT.
Alternatively, pearson's correlation test was used to check the correlation between total blink time and MRT for each experiment, removing experimental data for which correlation was not significant.
And a substep S33, determining unit blinking time as a fatigue label, performing correlation analysis between the unit blinking time and the HRV index, and extracting the index with obvious correlation and high correlation coefficient as fatigue characteristics.
Specifically, correlation analysis is performed on the extracted HRV indexes in the same time slice and unit blink time, and indexes which are significantly correlated and have high correlation coefficients are extracted as fatigue features.
And step S4, training and verifying the data set by using a machine learning method to obtain a pre-post fatigue detection model of the operator.
Specifically, step S4 further includes two substeps:
substep S41, the data set is updated according to the ratio of 7: 3, because the fatigue and non-fatigue data in the embodiment of the invention are unbalanced, an oversampling method is adopted to balance the data before dividing the data set so as to improve the effect of the model. And then training the model on a training data set by using different machine learning algorithms to obtain a plurality of before-post operator fatigue detection models, and testing on a testing data set to ensure the accuracy of the model.
Optionally, the machine learning algorithm comprises: support vector machines, K nearest neighbor classification, decision trees, naive Bayes, random forests, neural networks and the like.
And a substep S42 of checking the generalization ability of the model on the test set by using indexes such as accuracy, F1 value, precision and the like, and selecting the model with the highest generalization ability as a final model. A technical route map of specific data analysis is shown in fig. 3, and fig. 3 is a technical route map of data analysis of a method for detecting fatigue of an operator before post according to an embodiment of the present invention.
Alternatively, the trained training model may be evaluated by different methods, and common evaluation indexes include: accuracy, recall, sensitivity, specificity, F1 values, ROC curves, and the like. The F1 value is an important index for balancing accuracy and recall, and is also an applied index for evaluating the generalization capability of the model in the following.
From the above description, the embodiment of the invention designs a set of sleep deprivation experimental process, provides experimental tasks through a multi-attribute task platform, and measures multidimensional data including subjective fatigue data, HRV data, eye movement data, behavior data and the like in each experiment; then, the method provided by the embodiment of the invention performs correlation analysis on the subjective data, the behavior data and the unit blinking time data, and verifies the validity of the unit blinking time as a fatigue label; finally, the method provided by the embodiment of the invention utilizes the related indexes of HRV as fatigue characteristics, and can train out the operator post fatigue detection model based on the data of the experimenter collected by the sleep deprivation experiment. In the application process, an operator only needs to wear data equipment for recording electrocardio to perform a simple multi-attribute platform operation task for several minutes, and the fatigue state can be detected. The method has low invasiveness and high efficiency, and can provide guidance for decision makers.
The second embodiment:
since the method for detecting fatigue of an operator before the shift can be applied to various scenes, the embodiment of the invention describes a specific embodiment in combination with a specific scene of fatigue detection and screening before the shift of a civil aircraft pilot.
People nowadays select civil aviation as a vehicle more and more, and as a rapidly developing industry, the safety of civil aviation flight has been highly concerned by people all the time. Fatigue is an important factor influencing flight safety of pilots, and rapid fatigue detection and screening are carried out in a short time before the pilots are on duty, so that manual intervention is carried out, and the method plays a key role in reducing the risk of accidents. Therefore, aiming at the working scene and the operation task of the pilot, the embodiment of the invention designs a simulation experiment by using the multi-attribute task platform and designs a sleep deprivation experiment, wherein the time interval of each experiment is 4 hours. The total duration of each experiment was 15 minutes from 7 am the first day to 7 am the second day.
The KSS and SSS scales are filled before each experiment to record the subjective fatigue level of the tested eye, the eye movement data and the electrocardio data of the tested eye are synchronously recorded by using experimental equipment in the process of carrying out the simulation experiment, the tested eye movement data and the electrocardio data are subjected to PVT test after the simulation experiment is finished, and the average reaction time in the test process is recorded. Fig. 4 is a general experimental flowchart designed for civil aircraft pilots according to the embodiment of the present invention.
In the embodiment of the application, time slices are divided by taking one minute as a unit for the acquired HRV data, and time domain, frequency domain and nonlinear analysis are performed on the HRV data in the slices to obtain related indexes. Some of the relevant indices are shown in table 1.
TABLE 1 HRV indices
In the method provided by the embodiment of the application, the eye tracker is adopted to collect the tested eye movement data, extract the tested blink related data and calculate the unit blink time, and the calculation formula is as follows:
preferably, the time integration involved in the calculation in the present embodiment, i.e. the length of the slice, is 1 minute.
The evaluation data before and after the test are collected. The subjective fatigue before the experiment is checked by using a Carolina sleepiness scale and a Stanford sleepiness scale; the tested reaction capability after the experiment records the tested reaction time through the psychomotor warning test, and the collected data is stored in the storage unit.
Pearson correlation test was performed on the total blink time per experiment with KSS and SSS scale, removing data that were not significantly correlated, and dividing fatigue into three grades according to the grade of the subjective scale: no fatigue, general fatigue and severe fatigue, as shown in table 2.
TABLE 2 fatigue rating Scale Table
| KSS score | SSS score | Grade of fatigue |
| 1-5 | 1-3 | Without fatigue |
| 6-7 | 4-5 | General fatigue |
| 8-9 | 6-7 | Severe fatigue |
Pearson correlation test was performed on total blink time and MRT data for each experiment to remove data with insignificant correlation.
Finally, the unit blinking time is used as a fatigue label through the steps, and the fatigue grade of the fatigue label corresponds to the scores of the KSS and SSS subjective fatigue scale. And (3) carrying out correlation analysis on the indexes of the HRV data in the same time slice after screening and unit blink time, and selecting the indexes with obvious correlation and high correlation coefficient as fatigue characteristics.
Due to the unbalanced data of fatigue, general fatigue and severe fatigue in the embodiment of the invention, the class balance is carried out by an oversampling method to improve the accuracy of a subsequent model before dividing the data set. And (3) processing the processed data set according to the following steps of 7: and 3, dividing the training data set and the testing data set, and selecting different machine learning algorithms to train the model. Machine learning algorithms that may be selected are: support vector machines, K nearest neighbor classification, decision trees, naive Bayes, random forests, neural networks and the like.
The generalization ability of the model was examined on the test data set with the F1 value, and the model with the highest F1 value was selected as the final model. Finally, the fatigue rapid detection model before the post of the operator can be obtained through the steps, and the fatigue state of the operator can be detected only by carrying out 10-minute simulation experiment and collecting HRV data before the post, so that corresponding preventive measures are implemented to reduce risks.
Example three:
fig. 5 is a schematic diagram of a system for detecting pre-post fatigue of an operator according to an embodiment of the present invention. As shown in fig. 5, the system includes: experiment module 10, analysis module 20, extraction module 30, training module 40 and detection module 50.
Specifically, the experiment module 10 is configured to design a target simulation experiment based on an experiment method, and acquire an experiment data set in a target simulation experiment process; the target simulation experiment is an operator fatigue experiment which is developed on a multi-attribute task platform at different time periods based on a simulation scene; the experimental data set included: heart rate variability data, physiological data of eye movements, subjective scale data, and psychomotor alert task data.
And the analysis module 20 is configured to analyze the heart rate variability data to obtain a data index related to the heart rate variability data.
Optionally, an analysis module 20, further configured to divide the heart rate variability data into a plurality of time slices; and respectively carrying out time domain analysis, frequency domain analysis and nonlinear analysis on the data in each time slice to obtain a data index about the heart rate variability data.
The extraction module 30 is configured to extract and identify features of the experimental data set by using the physiological data of the eye movement as a fatigue tag, using the subjective scale data and the mental exercise warning task data for inspection, and using the data index as a fatigue feature, so as to obtain a preprocessed data set.
Optionally, the extracting module 30 is further configured to perform correlation analysis on the total blinking time and the subjective scale data, and verify the correlation between the unit blinking time and the subjective fatigue; and carrying out correlation analysis on the total blink time and the mental movement warning task data, and verifying the correlation between the unit blink time and the mental movement warning task data.
And the training module 40 is configured to divide the preprocessed data set into a training data set and a test data set, train a preset fatigue detection model on the training data set by using a machine learning method, and test the preset fatigue detection model on the test data set to obtain a trained fatigue detection model.
And the detection module 50 is used for performing on-duty fatigue detection on the operator to be detected by using the trained fatigue detection model.
Specifically, the detection module 50 is further configured to acquire heart rate variability data of an operator to be detected; and inputting the heart rate variability data of the operator to be detected into the fatigue detection model after training to obtain the pre-post fatigue detection result of the operator to be detected.
The invention provides an operator off-duty fatigue detection system, which comprehensively utilizes the experimental methods of scale measurement, behavior measurement and physiological data measurement to determine an operator off-duty fatigue detection model, the model can accurately detect the fatigue state of an operator, the detection process is simple and efficient, and the technical problems of inaccurate measurement and low measurement efficiency in the prior art are solved.
An embodiment of the present invention further provides an electronic device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, the steps of the methods in the first and second embodiments are implemented.
Embodiments of the present invention further provide a computer readable medium having a non-volatile program code executable by a processor, where the program code causes the processor to execute the method in the first and second embodiments.
Example four:
fig. 6 is a schematic diagram of a fatigue detection apparatus according to an embodiment of the present invention. As shown in fig. 6, the apparatus includes: an experiment unit 410, a storage unit 420 and a calculation unit 430.
Specifically, the experiment unit 410 is configured to formulate a simulation experiment under a scene and a task, and simultaneously includes that the electrocardiograph device can collect tested HRV data in an experiment process, a preset scale is provided for the tested to collect subjective fatigue state before the experiment, and a program of a PVT experiment is provided for the tested to collect behavior data after the experiment, wherein the used scale and program are embedded into the experiment unit, and the experiment can be called every time.
A storage unit 420 configured to store the collected subjective data, physiological data, and behavioral data; and the calculating unit is configured to calculate various indexes of the HRV according to the electrocardio data, train the HRV by using a machine learning algorithm by combining unit blink time data as a label to obtain an operator before-post fatigue detection model, and perform efficiency evaluation.
The experiment unit 410 can be divided into three subunits, an experiment platform subunit, an electrocardiogram acquisition device subunit and an evaluation data subunit. The experiment platform subunit can provide an experiment scene for a tested person, and comprises software meeting simulation experiments, an operating rod for experiment operation, a mouse, a keyboard, audio input and output equipment and the like. The electrocardio acquisition equipment subunit comprises equipment for acquiring tested electrocardio data, such as a heart rate belt, a bracelet, an electrode plate and the like. The evaluation data subunit is mainly used for collecting evaluation data in the experimental process, comprises a scale for testing subjective fatigue of a tested object and a PVT program for testing reaction time of the tested object, and can call the scale and the program in an interface interaction mode and store the data after evaluation is finished.
Specifically, the storage unit 420 should include data import and export functions and be capable of operating on an interactive interface.
In particular, the means of the calculation unit 430 should be able to support all the calculation requirements of the proposed method, including at least the calculation of the evaluation data, the extraction and statistical analysis of the electrocardiogram data and the eye movement data, and the training of the physiological data with a machine learning algorithm to obtain the fatigue classification model.
In the embodiment of the present invention, the experiment unit 410, the storage unit 420, and the calculation unit 430 shown in fig. 6 respectively implement corresponding processes in the embodiments of the methods in fig. 1 to fig. 4. The above description of the embodiments of the method is specifically provided, and the detailed description is omitted here as appropriate to avoid redundancy.
Finally, it should be noted that: the above embodiments are only used to illustrate the technical solution of the present invention, and not to limit the same; while the invention has been described in detail and with reference to the foregoing embodiments, it will be understood by those skilled in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some or all of the technical features may be equivalently replaced; and the modifications or the substitutions do not make the essence of the corresponding technical solutions depart from the scope of the technical solutions of the embodiments of the present invention.
Claims (10)
1. A method for detecting the fatigue of an operator before post is characterized by comprising the following steps:
designing a target simulation experiment based on an experimental method, and acquiring an experiment data set in the target simulation experiment process; the target simulation experiment is an operator fatigue experiment which is developed on a multi-attribute task platform at different time periods based on a simulation scene; the experimental data set includes: heart rate variability data, physiological data of eye movements, subjective scale data, and psychomotor alert task data;
analyzing the heart rate variability data to obtain a data index related to the heart rate variability data;
taking the physiological data of the eye movement as a fatigue label, checking by using the subjective scale data and the mental exercise warning task data, taking the data index as a fatigue feature, and extracting and identifying the feature of the experimental data set to obtain a preprocessed data set;
dividing the preprocessed data set into a training data set and a testing data set, training a preset fatigue detection model on the training data set by using a machine learning method, and testing on the testing data set to obtain a trained fatigue detection model;
and performing on-duty fatigue detection on the operator to be detected by using the trained fatigue detection model.
2. The method of claim 1, wherein analyzing the heart rate variability data to obtain a data index of the heart rate variability data comprises:
dividing the heart rate variability data into a plurality of time slices;
and respectively carrying out time domain analysis, frequency domain analysis and nonlinear analysis on the data in each time slice to obtain a data index about the heart rate variability data.
3. The method of claim 1, wherein the physiological data of the eye movement comprises total blink time and unit blink time; performing a test using the subjective scale data and the psychomotor vigilance task data, comprising:
performing correlation analysis on the total blink time and the subjective scale data, and verifying the correlation between the unit blink time and subjective fatigue;
and carrying out correlation analysis on the total blink time and the mental exercise alert task data, and verifying the correlation between the unit blink time and the mental exercise alert task data.
4. The method of claim 3, wherein the subjective scale data comprises Carolina Pasteur Scale data and Stanford Pasteur Scale data; the psychomotor alert task data includes a reaction time.
5. The method of claim 3, further comprising: and carrying out correlation analysis on the unit blinking time and the data indexes, and extracting the data indexes with the highest correlation coefficient as the fatigue features.
6. The method of claim 1, wherein the fatigue detection before post-job for the operator to be tested by using the fatigue detection model after training comprises:
acquiring heart rate variability data of the operator to be detected;
and inputting the heart rate variability data of the operator to be detected into the fatigue detection model after training to obtain the pre-post fatigue detection result of the operator to be detected.
7. An operator pre-post fatigue detection system, comprising: the system comprises an experiment module, an analysis module, an extraction module, a training module and a detection module; wherein, the first and the second end of the pipe are connected with each other,
the experiment module is used for designing a target simulation experiment based on an experiment method and acquiring an experiment data set in the process of the target simulation experiment; the target simulation experiment is an operator fatigue experiment based on a simulation scene on a multi-attribute task platform at different time periods; the experimental data set comprises: heart rate variability data, physiological data of eye movements, subjective scale data, and psychomotor alert task data;
the analysis module is used for analyzing the heart rate variability data to obtain a data index related to the heart rate variability data;
the extraction module is used for taking the physiological data of the eye movement as a fatigue label, checking the subjective scale data and the mental and physical vigilance task data, taking the data index as a fatigue feature, and extracting and identifying the feature of the experimental data set to obtain a preprocessed data set;
the training module is used for dividing the preprocessed data set into a training data set and a testing data set, training a preset fatigue detection model on the training data set by using a machine learning method, and testing on the testing data set to obtain a trained fatigue detection model;
and the detection module is used for performing on-duty fatigue detection on the operator to be detected by using the trained fatigue detection model.
8. The system of claim 7, wherein the detection module is further configured to:
acquiring heart rate variability data of the operator to be detected;
and inputting the heart rate variability data of the operator to be detected into the fatigue detection model after training to obtain the pre-post fatigue detection result of the operator to be detected.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor implements the steps of the method of any one of claims 1 to 6 when executing the computer program.
10. A computer-readable medium having non-volatile program code executable by a processor, wherein the program code causes the processor to perform the method of any of claims 1-6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210453967.XA CN114783552A (en) | 2022-04-24 | 2022-04-24 | Method and system for detecting fatigue of operator before post |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210453967.XA CN114783552A (en) | 2022-04-24 | 2022-04-24 | Method and system for detecting fatigue of operator before post |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114783552A true CN114783552A (en) | 2022-07-22 |
Family
ID=82432154
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210453967.XA Pending CN114783552A (en) | 2022-04-24 | 2022-04-24 | Method and system for detecting fatigue of operator before post |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114783552A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117290781A (en) * | 2023-10-24 | 2023-12-26 | 中汽研汽车检验中心(宁波)有限公司 | Driver KSS grade self-evaluation training method for DDAW system test |
-
2022
- 2022-04-24 CN CN202210453967.XA patent/CN114783552A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117290781A (en) * | 2023-10-24 | 2023-12-26 | 中汽研汽车检验中心(宁波)有限公司 | Driver KSS grade self-evaluation training method for DDAW system test |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20210000426A1 (en) | Classification system of epileptic eeg signals based on non-linear dynamics features | |
| KR102129730B1 (en) | Patient condition predicting apparatus based on artificial intelligence and predicting method using the same | |
| KR102008196B1 (en) | Apparatus for generating of Potassium level determination model using ECG and method thereof | |
| CN111513730B (en) | Psychological stress prediction method and system based on multichannel physiological data | |
| Jeon et al. | Multi-class classification of construction hazards via cognitive states assessment using wearable EEG | |
| CN109741007A (en) | A kind of subject personnel's screening technique of aviation cockpit workload test | |
| CN114693170A (en) | Multi-source parameter fusion-based psychological assessment method for parachuting training of pilots | |
| CN114154592A (en) | Flight training quality and efficiency evaluation method fusing physiological data and flight parameter data | |
| Nath et al. | Machine learning-based anxiety detection in older adults using wristband sensors and context feature | |
| US20030153817A1 (en) | Pattern analysis system and method | |
| CN114783552A (en) | Method and system for detecting fatigue of operator before post | |
| Kaur et al. | Using virtual reality to examine the neural and physiological anxiety-related responses to balance-demanding target-reaching leaning tasks | |
| CN112617772B (en) | Driving fatigue identification method and system based on pulse wave signals | |
| Yu et al. | Air traffic controllers' mental fatigue recognition: A multi-sensor information fusion-based deep learning approach | |
| Dehzangi et al. | Driver distraction detection using MEL cepstrum representation of galvanic skin responses and convolutional neural networks | |
| CN115017996A (en) | Mental load prediction method and system based on multiple physiological parameters | |
| Javel et al. | Epileptic seizure detection via EEG using tree-based pipeline optimization tool | |
| JP2018078982A (en) | Method for evaluating stress reaction during training | |
| Eskandar et al. | Using Deep Learning for Assessment of Workers' Stress and Overload | |
| CN112351728A (en) | Improved method for quantization balancing | |
| CN112656372B (en) | Brain network topological rule-based mental fatigue detection method and system | |
| Vianne et al. | A Promising Approach to Detect Signs of Apathy from a Single Tri-axial Accelerometer | |
| Sun et al. | Research on recognizable physiological signals of workers working at heights | |
| CN112656373B (en) | Physical fatigue detection method and system based on brain network topological law | |
| JP7507984B1 (en) | Physical condition detection device, physical condition detection system, physical condition detection method, and physical condition detection program |
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 |