Virtual reality-based locomotive operator job fatigue evaluation system

Abstract

The application belongs to the technical field of locomotive driver ergonomics. The application provides a virtual reality-based evaluation system for the fatigue of a locomotive driver. The disclosed embodiment builds a highly simulated driving scene, simulates long-time and high-intensity driving tasks, reproduces various situations that may cause driver fatigue in a risk-free manner, and monitors and analyzes the physiological response, behavioral change and psychological state of the driver in the simulation process in real time through high-precision equipment, effectively solving the problems of lack of accuracy and real-time of the existing evaluation methods, and timely and accurately evaluating and finding the fatigue degree of the locomotive driver in the task execution process, thereby providing safety guarantee for locomotive driving.

Description

Technical Field

[0001] This disclosure relates to the field of locomotive driver ergonomics, and more particularly to a locomotive driver fatigue evaluation system based on virtual reality. Background Technology

[0002] During high-intensity, long-duration driving operations, driver fatigue is becoming increasingly prominent, posing a significant threat to traffic safety. Studies show that driver fatigue leads to decreased reaction speed, distraction, and impaired judgment, which can, in severe cases, cause traffic accidents. Especially in the complex railway environment, the sustained high efficiency of locomotive drivers is a key factor in ensuring traffic safety and maintaining transport order. Therefore, in-depth research into locomotive driver fatigue is crucial for improving the safety and efficiency of railway transportation.

[0003] Although the dangers of driver fatigue are widely recognized, current research still faces significant challenges and limitations. Existing fatigue monitoring technologies often rely on physiological indicators. For example, some researchers use cameras and other sensors to monitor drivers' facial expressions, blinking frequency, head posture, and other physiological indicators; others use surface electromyography (EMG) to collect electromyographic signals for real-time monitoring and evaluation of operational fatigue. These indicators are difficult to accurately capture in actual driving environments, and the assessments tend to be somewhat limited. While currently used methods can reveal driver fatigue characteristics to some extent, their accuracy and real-time performance remain to be improved due to technological limitations and the diversity of fatigue manifestations in complex driving environments. They have failed to effectively enhance locomotive safety and reduce the likelihood of accidents caused by driver fatigue. Furthermore, the objective state of locomotive drivers during operation is difficult for maintenance personnel to monitor, especially when drivers are fatigued and a driving accident is possible. Therefore, existing equipment and methods, such as subjective scale assessments and physiological characteristic-based operational fatigue assessments, have certain shortcomings in objectivity, comprehensiveness, and accuracy. Summary of the Invention

[0004] To avoid the shortcomings of existing technologies, this invention provides a virtual reality-based locomotive driver work fatigue evaluation system to address the problems of existing technologies, such as subjective scale assessment and work fatigue assessment based on physiological characteristics, which have certain deficiencies in objectivity, comprehensiveness, and accuracy.

[0005] According to a first aspect of the present disclosure, a locomotive driver fatigue assessment system based on virtual reality is provided, the system comprising:

[0006] Virtual human-machine driving control unit, used to provide simulated driving scenarios;

[0007] The human-machine data acquisition unit is used to collect the locomotive driver's work behavior and tasks during driving to obtain driving operation data; wherein, the driving operation data includes the locomotive driver's work data, reaction time, accuracy rate, eye movement data and limb skeletal data;

[0008] The data processing and feature extraction analysis unit is used to preprocess the driving operation data and extract indicator feature values; wherein, the indicator feature values ​​include workload information, task performance information, eye fatigue characteristics, and limb fatigue characteristics;

[0009] A cross-system data time synchronization and dimensionless processing unit for various indicator features is used to perform time alignment on the work data and normalize and dimensionless the indicator feature values ​​to obtain workload indicators, task performance indicators, eye movement feature indicators and limb feature indicators.

[0010] The locomotive driver's work fatigue fuzzy comprehensive evaluation unit is used to fuse the workload, task performance, eye movement features and limb features to obtain the final evaluation result;

[0011] The locomotive driver's work fatigue comprehensive evaluation result level classification unit is used to assess the locomotive driver's work fatigue status based on the final evaluation result.

[0012] Furthermore, the virtual human-machine driving operation unit includes:

[0013] Virtual machine vehicle cockpit, virtual driving missions, virtual reality head-mounted display, and motion capture system; among them,

[0014] The virtual vehicle cockpit includes a seat, control panel, joystick, and digital model of the locomotive cockpit, used to simulate the operating environment of the locomotive;

[0015] The virtual driving task includes tasks such as controlling the joystick, using hand gestures to describe driving conditions, observing road information, and viewing dashboard information, and is used to simulate the driving of a locomotive.

[0016] The virtual reality head-mounted display is used to track the locomotive driver's eye movements;

[0017] The motion capture device is used to capture the driving actions of the locomotive driver.

[0018] Furthermore, the human-machine data integrated acquisition unit includes:

[0019] The system includes a virtual machine vehicle driving record module, a driving task execution record module, an eye-tracking data acquisition module, and a driving motion acquisition module; among these...

[0020] The virtual machine vehicle driving record module is used to collect the operation data of the locomotive driver;

[0021] The driving task execution record module is used to collect the locomotive driver's reaction time and accuracy rate during task execution;

[0022] The eye-tracking data acquisition module is used to acquire the eye-tracking data of the locomotive driver;

[0023] The driving motion acquisition module is used to collect the limb skeletal data of the locomotive driver.

[0024] Furthermore, the data processing and feature extraction analysis unit includes:

[0025] The module includes a workload calculation module, a task performance calculation module, an eye-tracking feature extraction module, and a limb feature extraction module; among them,

[0026] The workload calculation module is used to calculate the locomotive driver's work behavior time and work task requirements to obtain workload information;

[0027] The task performance calculation module is used to quantify the locomotive driver's task performance information based on the average reaction time obtained from the reaction time and the accuracy rate.

[0028] The eye movement feature extraction module is used to preprocess the eye movement data and extract eye fatigue features from the preprocessed eye movement data;

[0029] The limb feature extraction module is used to process the limb skeletal data and extract limb fatigue features from the processed limb skeletal data.

[0030] According to a second aspect of the present disclosure, a method for evaluating locomotive driver fatigue based on virtual reality is provided, the method comprising:

[0031] Simulated driving scenarios;

[0032] Based on the driving scenario, the locomotive driver's work behavior and tasks during driving are collected to obtain driving operation data; wherein, the driving operation data includes the locomotive driver's work data, reaction time, accuracy rate, eye movement data and limb skeletal data;

[0033] The driving operation data is preprocessed, and indicator feature values ​​are extracted; wherein, the indicator feature values ​​include workload information, task performance information, eye fatigue characteristics, and limb fatigue characteristics;

[0034] The work data is time-aligned, and the indicator feature values ​​are normalized and dimensionless to obtain workload indicators, task performance indicators, eye movement feature indicators, and limb feature indicators.

[0035] The workload, task performance, eye movement features, and limb features are fused together to obtain the final evaluation result;

[0036] The operational fatigue status of the locomotive driver is assessed based on the final evaluation results.

[0037] Furthermore, the step of performing time alignment on the task data and normalizing and dimensionlessizing the indicator feature values ​​to obtain workload indicators, task performance indicators, eye-tracking feature indicators, and limb feature indicators includes:

[0038] The Unix timestamp is converted into a regular date and time format by calling the system time and Unix timestamp;

[0039] The cross-system data time synchronization and dimensionless processing unit for each indicator feature uses the extreme value standardization method to normalize and dimensionless the workload information, the task performance information, the eye fatigue feature and the limb fatigue feature to obtain the workload indicator, the task performance indicator, the eye movement feature indicator and the limb feature indicator.

[0040] Furthermore, the workload index, the task performance index, the eye movement characteristic index, and the limb characteristic index are secondary indicators; the tertiary indicators under the workload index are the current workload and the cumulative workload, the tertiary indicators under the task performance index are the average reaction time, the tertiary indicators under the eye movement characteristic index are the average pupil diameter and the average blink time, and the tertiary indicators under the limb characteristic index are the current discomfort and the cumulative discomfort.

[0041] Furthermore, the step of fusing the workload, task performance, eye-tracking features, and limb features to obtain the final evaluation result includes:

[0042] Construct a membership function based on the triangular fuzzy function;

[0043] The confidence level of the membership function's judgment matrix is ​​adjusted and a consistency check is performed.

[0044] The judgment matrix after confidence correction and consistency test is processed using defuzzification calculation and root-finding method to obtain the maximum eigenvalue and corresponding eigenvector of the judgment matrix;

[0045] The maximum eigenvalue and the corresponding eigenvector are normalized to obtain the comprehensive weight of the secondary index;

[0046] Then, the secondary indicators are ranked according to their relative importance to the risk of driver fatigue, and the weight values ​​of all tertiary indicators are determined by combining the comprehensive weights of the secondary indicators.

[0047] The fatigue evaluation value is obtained by calculating and dimensionlessly processing the weight values ​​of the three-level indicators and summing them with the corresponding weights. The final evaluation result is then determined based on the fatigue evaluation value.

[0048] Furthermore, the final evaluation results include:

[0049] Fatigue levels: not fatigued, slightly fatigued, somewhat fatigued, noticeably fatigued, and very fatigued; among which,

[0050] The fatigue rating value is 0~0.2 indicating no fatigue; 0.2~0.4 indicating slight fatigue; 0.4~0.6 indicating moderate fatigue; 0.6~0.8 indicating significant fatigue; and 0.8~1.0 indicating severe fatigue.

[0051] The technical solutions provided by the embodiments of this disclosure may include the following beneficial effects:

[0052] In the embodiments of this disclosure, the aforementioned virtual reality-based locomotive driver fatigue evaluation system, on the one hand, constructs highly simulated driving scenarios to simulate long-duration, high-intensity driving tasks, reproducing various situations that may lead to driver fatigue in a risk-free manner. Through high-precision equipment, it monitors and analyzes the driver's physiological reactions, behavioral changes, and psychological state in real time during the simulation process, effectively solving the problems of inaccuracy and real-time performance of existing evaluation methods. It can promptly and accurately assess and detect the degree of operational fatigue of locomotive drivers during task execution, providing safety assurance for locomotive driving. On the other hand, this system provides a more objective, comprehensive, and scientific evaluation method for assessing ergonomic locomotive driver operational fatigue, promoting the innovative application of high-tech technologies such as virtual reality, motion capture, and eye tracking in the field of traffic safety. Attached Figure Description

[0053] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this disclosure and, together with the description, serve to explain the principles of this disclosure. It is obvious that the drawings described below are merely some embodiments of this disclosure, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0054] Figure 1 A schematic diagram of a virtual reality-based locomotive driver fatigue assessment system is shown in an exemplary embodiment of this disclosure;

[0055] Figure 2 This diagram illustrates a station operation logic flowchart designed in a virtual environment according to an exemplary embodiment of the present disclosure.

[0056] Figure 3 This diagram shows a top view of an experimental setup built in a real-world environment based on the NOKOV optical motion capture system, as illustrated in an exemplary embodiment of this disclosure.

[0057] Figure 4 This illustration shows a schematic diagram of the process of converting motion capture skeletal animation data into an animation usable by the SoErgo software in an exemplary embodiment of this disclosure;

[0058] Figure 5 A schematic diagram of a comprehensive evaluation index system for locomotive driver work fatigue based on virtual reality in an exemplary embodiment of this disclosure is shown.

[0059] Figure 6 This diagram illustrates a comparison of the consistency of driving fatigue value curves in exemplary embodiments of this disclosure.

[0060] Figure 7 A graph showing the consistency of the fatigue self-rating scale score curves of subjects in an exemplary embodiment of this disclosure is provided.

[0061] Figure 8 The diagram illustrates the steps of a virtual reality-based locomotive driver fatigue evaluation method in an exemplary embodiment of this disclosure. Detailed Implementation

[0062] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that this disclosure will be more comprehensive and complete, and will fully convey the concept of the exemplary embodiments to those skilled in the art. The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0063] Furthermore, the accompanying drawings are merely illustrative diagrams of embodiments of this disclosure and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities.

[0064] This example implementation first provides a virtual reality-based locomotive driver fatigue assessment system. (Reference) Figure 1 As shown, the locomotive driver fatigue evaluation system based on virtual reality may include: a virtual human-machine driving operation unit, a human-machine data comprehensive acquisition unit, a data processing and feature extraction and analysis unit, a cross-system data time synchronization and dimensionless feature processing unit, and a locomotive driver fatigue fuzzy comprehensive evaluation unit.

[0065] The system includes a virtual human-machine driving operation unit for providing simulated driving scenarios; a human-machine data acquisition unit for collecting the locomotive driver's work behavior and tasks during driving to obtain driving operation data, including the driver's work data, reaction time, accuracy rate, eye movement data, and limb skeletal data; a data processing and feature extraction analysis unit for preprocessing the driving operation data and extracting indicator feature values, including workload information, task performance information, eye fatigue characteristics, and limb fatigue characteristics; and cross-system data... The time synchronization and dimensionless processing unit is used to align the work data in time and normalize and dimensionless the indicator feature values ​​to obtain workload indicators, task performance indicators, eye movement feature indicators, and limb feature indicators. The locomotive driver's work fatigue fuzzy comprehensive evaluation unit is used to fuse the workload, task performance, eye movement features, and limb features to obtain the final evaluation result. The locomotive driver's work fatigue comprehensive evaluation result level classification unit is used to assess the work fatigue state of the locomotive driver based on the final evaluation result.

[0066] The aforementioned virtual reality-based locomotive driver fatigue assessment system addresses two key issues. First, by constructing highly realistic driving scenarios, it simulates prolonged, high-intensity driving tasks, reproducing various situations that could lead to driver fatigue in a risk-free manner. Second, through high-precision equipment, it monitors and analyzes the driver's physiological responses, behavioral changes, and psychological state in real time during the simulation. This effectively solves the problems of inaccuracy and real-time performance issues found in existing assessment methods, enabling timely and accurate evaluation and detection of the degree of operational fatigue among locomotive drivers during task execution, thus providing safety assurance for locomotive driving. Third, the system provides a more objective, comprehensive, and scientific evaluation method for assessing ergonomic locomotive driver fatigue, promoting the innovative application of advanced technologies such as virtual reality, motion capture, and eye tracking in the field of traffic safety.

[0067] Below, we will refer to Figures 1 to 7 The various parts of the virtual reality-based locomotive driver fatigue assessment system described in this example embodiment will be explained in more detail.

[0068] In one embodiment, the virtual human-machine driving operation unit provides a highly simulated experimental environment. The human-machine data acquisition unit collects accurate objective data and outputs it to the data processing and feature extraction analysis unit for data preprocessing and extraction of effective feature values. Through the cross-system data time synchronization and dimensionless feature processing unit, the data time is unified and the data feature values ​​are dimensionless. The data is then input into the locomotive driver's work fatigue fuzzy comprehensive evaluation unit to calculate and generate continuous and quantitative human-machine ergonomic work fatigue evaluation results. Finally, the results are input into the evaluation result level classification unit to output the human-machine work fatigue state analysis results.

[0069] More specifically, the virtual human-machine driving operation unit includes: a virtual machine driver's cockpit, a virtual driving task, a virtual reality head-mounted display, and a motion capture device. The virtual machine driver's cockpit is a digital model of a locomotive driver's cockpit created using digital modeling software, including the seat, control panel, and joystick. The virtual driving task simulates four typical locomotive driving tasks, including joystick control, driving gesture control, road information observation, and instrument panel information viewing. It utilizes a high-end, high-precision virtual reality head-mounted display. The motion capture device includes a hardware and software system. The hardware mainly consists of a PoE switch, a PoE splitter, a calibration rod, an infrared optical lens, and several 15mm diameter reflective markers.

[0070] The human-machine data acquisition unit includes: a virtual vehicle driving record module, a driving task execution record module, an eye-tracking data acquisition module, and a driving motion acquisition module. Specifically: the virtual vehicle driving record module collects the locomotive driver's operational data; the driving task execution record module collects the locomotive driver's reaction time and accuracy during task execution; the eye-tracking data acquisition module collects the locomotive driver's pupil size, blink count, and blink duration; and the driving motion acquisition module collects the locomotive driver's three-dimensional skeletal motion data.

[0071] The data processing and feature extraction analysis unit includes: a workload calculation module, a task performance calculation module, an eye-tracking feature extraction module, and a limb feature extraction module. The workload calculation module includes calculating the locomotive driver's operational time and task requirements; the task performance calculation module quantifies the locomotive driver's task performance using average reaction time obtained through reaction time and accuracy rate; the eye-tracking feature extraction module includes preprocessing of eye-tracking data and extraction of eye fatigue features; and the driving action acquisition module includes limb data repair preprocessing, data processing based on animation redirection, and limb fatigue feature extraction based on SoErgo.

[0072] The fuzzy comprehensive evaluation unit for locomotive driver fatigue includes the following secondary indicators: workload indicators, task performance indicators, eye movement characteristic indicators, and limb characteristic indicators. The tertiary indicators under the workload indicator are current workload and cumulative workload; the tertiary indicator under the task performance indicator is average reaction time; the tertiary indicators under the eye movement characteristic indicator are average pupil diameter and average blink time; and the tertiary indicators under the limb characteristic indicator are current discomfort and cumulative discomfort. All tertiary indicators are first processed to be dimensionless.

[0073] The comprehensive evaluation results of locomotive driver fatigue are divided into five levels: 0-0.2 indicates no fatigue; 0.2-0.4 indicates slight fatigue; 0.4-0.6 indicates moderate fatigue; 0.6-0.8 indicates significant fatigue; and 0.8-1.0 indicates severe fatigue.

[0074] In a specific embodiment, such as Figure 1 As shown, this application discloses a locomotive driver fatigue evaluation system based on virtual reality, which includes: a virtual human-machine driving operation unit, a human-machine data comprehensive acquisition unit, a data processing and feature extraction analysis unit, a locomotive driver fatigue fuzzy comprehensive evaluation unit, and a locomotive driver fatigue comprehensive evaluation result level classification unit.

[0075] This application utilizes the virtual reality-based locomotive driver fatigue evaluation method of the aforementioned system, including: driving records during driving tasks, task execution records, data preprocessing and feature extraction based on raw eye movement and limb data, workload calculation, calculation and dimensionless transformation of three-level indicators, calculation of the comprehensive evaluation result of locomotive driver fatigue based on virtual reality, and classification of the evaluation result level.

[0076] The virtual machine driver's cockpit in the virtual human-machine driving operation unit uses Unity Long Term Support version 2022.3.13f1c1 as the platform for building the virtual reality simulation environment to achieve a high degree of simulation of locomotive driving operation scenarios. Simultaneously, SteamVR plugin version 2.2.3 is integrated, and the corresponding Unity Package is imported to ensure compatibility and interactivity with VR devices. The specific software and systems used are shown in Table 1, the development environment.

[0077] Table 1 Development Environment

[0078]

[0079] The virtual environment created in the experiment included a locomotive cab, platform, railway tracks, and road surface elements. The virtual locomotive cab was kept stationary, while the platform, railway tracks, and road surface moved continuously and dynamically along the negative x-axis. The scene contained 10 prefab instances, each containing a section of railway track and its corresponding road surface. Each prefab instance was loaded with the same script, using the principle of "parallax scrolling" to synchronously adjust the positions of background elements. The key logic of the locomotive's operation was as follows: Figure 2 As shown, collision detection points are precisely positioned on the platform. When the virtual train comes into contact with these collision points, it will trigger a braking action, play a braking sound effect, and start a coroutine to stop on the platform for 30 seconds.

[0080] The virtual driving task in the virtual human-machine driving control unit consists of four sub-tasks: controlling the joystick, observing road information and providing feedback, viewing dashboard information and providing feedback, and making corresponding gestures. These tasks are primarily implemented using scripts in Unity. Task rounds are controlled using Unity's coroutine functions and the `StartCoroutine()` method to schedule the start, progress, and end of tasks according to time. The joystick control task is implemented using Unity's Canvas and related functions. The road information observation and dashboard information viewing tasks are implemented through interaction between the Unity virtual environment and VR controllers. Each task round lasts 5 minutes, with a 1-minute break between rounds, for a total of ten rounds.

[0081] The virtual reality head-mounted display in the virtual human-machine driving control unit integrates an eye-tracking module.

[0082] The motion capture system in the virtual human-machine driving control unit employs an optical motion capture system. The experimental environment established is as follows: Figure 3 As shown, the overall dimensions of the space are approximately 2 meters × 2 meters × 1.6 meters. Obstruction between the camera and the subject is avoided, and there is no significant noise within the space. Eight infrared cameras (numbered 1 to 8 in the figure) are evenly distributed around the experimental area to ensure coverage of the subject's actions from all angles. Two SteamVR spatial locators (numbered 9 and 10 in the figure) are located on either side in front of the subject. Their height is adjusted to be roughly the same as the VR controllers and headset intended for use, minimizing controller and headset drift caused by positioning errors and ensuring that the subject's movements are accurately reflected in the virtual environment. In the core area directly in front of the subject, there is a stand for the camera to capture video (numbered 11 in the figure). Opposite the video shooting stand is a screen (marked 12 in the figure) placed behind, mainly serving as background isolation. This screen is used to synchronously record live video data of the experimental process. The main subject (wearing a VR headset, numbered 13 in the figure) performs a pre-set task in the center of the space, while the experiment facilitator (numbered 14 in the figure) monitors and guides the entire experimental process from outside the space.

[0083] The virtual vehicle driving record module in the human-machine data acquisition unit mainly records the work behavior and tasks during the driving process, providing a basis for subsequent workload calculation.

[0084] The driving task execution recording module in the human-machine data integration acquisition unit acquires and records data to a CSV file via Unity task scripts. Specifically, at each sampling moment, the system captures the current system time, corresponding reaction time information, and accuracy rate, and writes this data to the CSV file. All data recording is only active during task execution.

[0085] The eye-tracking data acquisition module in the human-machine data integration unit captures eye-tracking data based on the eye-tracking module of the virtual reality head-mounted display. This module can capture and record timestamps, gaze origin, gaze direction, pupil position, pupil size, and eye opening degree during task execution.

[0086] The driving motion acquisition module in the human-machine data integration unit is based on the Nokov series of passive optical motion capture systems with markers. Using a built-in 53-point Helen Hayes model, marker reflective points are affixed to various joints of the human body. By capturing the three-dimensional coordinate information of these reflective points through this system, the driver's limb skeletal data can be calculated and generated.

[0087] The workload calculation module in the data processing and feature extraction analysis unit includes the calculation of locomotive driver operation time and the calculation of task requirements. The operation time includes perception time.T v Cognitive Time T c and operation time T m Three items.

[0088] First, information perception time T v Using the "Method-Time Measurement" (MTM) method, the estimated information perception time is approximately 4-11 TMU (146-396 ms) based on the needs of the locomotive driver in this embodiment, with an average of 7.3 TMU (263 ms).

[0089] Cognitive time for decision-making T c Using Hick's law for estimation, the calculation formula is as follows:

[0090]

[0091] In the formula, n is the number of optional items; The cognitive time for the selected item can be derived from statistical results of most people. The empirical value is 155ms; For the first i The frequency (or probability) of each choice occurring.

[0092] Calculate the execution time of an operation using Fitts' Law. T m The calculation formula is as follows:

[0093]

[0094] In the formula, The time for the operation is taken as 100ms based on experience; D represents the manual distance required for the operation, and W represents the diameter of the object being operated on.

[0095] Therefore, the average operation time characteristic value of the operation behavior is calculated as follows:

[0096]

[0097] If a locomotive driver's task does not require any operational actions, then the time for that action is 0, i.e. T v , T c and T m It can be 0.

[0098] The workload requirement also includes three components: perceptual, cognitive, and operational requirements, calculated using the following formula:

[0099]

[0100] In the formula, This represents the required amount of sensory tasks at that moment; similarly, This represents the cognitive task requirement. Let B(t) be the task requirement. B(t) is the persistence symbol for the i-th behavioral element. When the behavior exists at time t, B(t) = 1; otherwise, B(t) = 0. In this implementation, the task requirement range for each basic behavioral element is adjusted to a 7-point scale (the higher the score, the higher the corresponding task requirement value). The specific values ​​of the task requirement for each behavioral element are shown in Table 2.

[0101] Table 2. Description of each basic visual, cognitive, and operational behavioral element and their task requirements.

[0102] Task requirements Description of basic behavioral elements <![CDATA[ ρ v ]]> Visual behavior 1 Visual Discovery 3.7 Visual recognition 4 Visual inspection 5 Visual positioning 5.4 Visual tracking 5.9 Visual reading 7 Visual search and surveillance <![CDATA[ ρ c ]]> Cognitive Behavior 1 Automatic response 1.2 Solution Selection 3.7 Signal recognition 4.6 Assessment, judgment 5.3 Encoding, Decoding <![CDATA[ ρ m ]]> Operational behavior 2.2 Discrete operation (buttons, trigger switches, etc.) 2.6 Continuous adjustment (steering wheel, speed lever, joystick, etc.) 4 Coherent hand movements 5.8 Discrete adjustment (rotation, vertical wheel, lever position)

[0103] The formula for calculating the average workload of each task is as follows:

[0104]

[0105] This application defines the four sub-tasks mentioned above (controlling the joystick, observing and responding to road information, viewing and responding to dashboard information, and making corresponding gestures) as Task 1, Task 2, Task 3, and Task 4. Therefore, after calculating the operation time and workload of each sub-task, the workload of this round of driving tasks can be calculated using the following formula:

[0106]

[0107] This represents the average workload of a certain task, where This represents the time taken to perform a specific task. The task performance calculation module in the data processing and feature extraction analysis unit selects the reaction time captured by the system and written to a CSV file from the experimental data, divides it by the accuracy rate, and uses this average reaction time as the quantitative characteristic value of task performance. Therefore, the task performance is calculated as follows:

[0108]

[0109] In the formula, Indicates task performance, Indicates reaction time. Indicates accuracy rate.

[0110] The eye-tracking feature extraction module in the data processing and feature extraction analysis unit is divided into two parts: data processing and feature extraction. In eye-tracking data processing, firstly, the pupil diameters of both eyes are selected. If neither pupil diameter is -1, the average of the left and right pupil diameters is calculated as the pupil diameter; if the data for a single eye is -1, the data for the other eye is retained as the pupil diameter. Secondly, for missing pupil diameter data, linear interpolation is used to fill in the gaps. The missing data interval is located, and five consecutive valid data points are selected before and after this interval. The average of the first five valid data points is used as the interpolation starting point, denoted as . The mean of the last five valid data points is used as the interpolation endpoint, denoted as . To fill in the missing data points, substitute the two points into the following formula for calculation:

[0111]

[0112] In the formula, This is an approximate function for this data segment. This is the original function of the data segment. This represents the value at that point.

[0113] An FIR filter is used to eliminate outliers and noise in pupil diameter data. This embodiment designs and implements an FIR filter using the scipy.signal library in Python. By calling the firwin function and setting the appropriate filter order and a normalized cutoff frequency of 12.5Hz, the required filter coefficients are systematically calculated. Then, the interpolated pupil diameter data is low-pass filtered, and a Hanning window is used to filter out high-frequency signals above 12.5Hz.

[0114] In eye-tracking feature extraction, two main features are extracted: mean pupil diameter (PDM) and mean blink time (BTM). The mean pupil diameter (PDM) reflects the average size of the driver's pupil, and its calculation formula is shown below:

[0115]

[0116] Where N is the number of sampling points in the sample. The pupil diameter (PDM) was used to examine the differences in mean pupil diameter (PDM) under different fatigue levels of the same driver. Data from the first five minutes and the last five minutes of the experiment were selected for each participant, and a paired-samples t-test was used for quantitative analysis. With a significance level of 0.001, the t-test showed a significant difference in pupil diameter between normal driving and fatigued driving conditions for the same driver.

[0117] If the pupil size is less than 0.4 in several consecutive frames of data, this time period is marked as a complete blink. The duration of this time period and the timestamp of the starting frame are recorded. The average blink time (BTM) reflects the average time it takes for a driver to blink once, and its calculation formula is as follows:

[0118]

[0119] Where N is the number of blinks, The time taken to blink is considered as one blink. To examine the differences in average blink time (BTM) under different levels of fatigue for the same driver, data from the first five minutes and the last five minutes of the experiment were selected for each participant. A paired-samples t-test was used to quantitatively study the differences in average blink time. With a significance level of 0.001, it was found that there was a significant difference in average blink time between normal driving and fatigued driving for the same driver.

[0120] The limb feature extraction module in the data processing and feature extraction analysis unit is divided into limb data repair preprocessing, animation-based retargeting data processing, and SoErgo-based limb fatigue feature extraction. The limb data repair preprocessing uses the motion capture system's accompanying software to repair the data, automatically calculating interpolation based on the data before and after the missing area to complete the data.

[0121] Data processing based on animation redirection requires exporting all data as FBX files, and then redirecting it in MotionBuilder. This involves using the skeletal model built into MotionBuilder to redirect the animation data exported from the motion capture software to animation data usable by SoErgo. The process is as follows: Figure 4 As shown.

[0122] SoErgo-based limb fatigue feature extraction uses the software's built-in RULA comfort analysis score as the feature value. Specifically, current-time comfort is used as the feature value for current discomfort, and time-based cumulative comfort is used as the feature value for cumulative discomfort.

[0123] In the cross-system data time synchronization and dimensionless characteristic processing unit, because different hardware devices and software tools may use different methods to record data time information, the system time and Unix timestamps are called to convert the Unix timestamps into an easy-to-read date and time format, ensuring that most data achieves millisecond-level time accuracy during the synchronization process, thereby effectively supporting experimental data analysis and integration. The integration formula is as follows:

[0124]

[0125] In the formula, AX represents the Unix timestamp value to be converted, and Date represents the time when the conversion is completed.

[0126] For the extracted indicator feature values, this implementation uses the extreme value standardization method to normalize and dimensionlessly transform the quantitative indicator values. The specific operation steps are as follows:

[0127] Positive indicators that contribute more to the objective with larger values ​​include:

[0128]

[0129] For inverse indicators where larger values ​​contribute less to the objective, the following are examples:

[0130]

[0131] In the formula: This represents the i-th observation of the j-th indicator. and These represent the maximum and minimum values ​​that the j-th indicator can take, respectively, based on the upper and lower limits of the locomotive driver's working time and working conditions standards.

[0132] The fuzzy comprehensive evaluation unit for locomotive driver fatigue integrates dimensionless workload, task performance, eye-tracking characteristics, and limb characteristics to obtain the final evaluation result. First, this implementation introduces an expert-based method using triangular fuzzy numbers to construct the membership function; second, it performs confidence correction and consistency checks on the judgment matrix; then, it uses defuzzification calculation and root-finding methods to obtain the largest eigenvalue and corresponding eigenvector of the judgment matrix. Normalizing the finally obtained eigenvector yields the comprehensive weight of that level of indicator. Next, the secondary indicators are ranked according to their relative importance to driver fatigue risk to determine the weight values ​​of all tertiary indicators. Finally, the calculated and dimensionless eigenvalues ​​of the tertiary indicators are weighted and summed with their corresponding weights to obtain the final evaluation result. Figure 5 As shown, this is a comprehensive evaluation index system for locomotive driver fatigue based on virtual reality.

[0133] The locomotive driver's fatigue assessment results are categorized into five levels: 0-0.2 indicates no fatigue; 0.2-0.4 indicates slight fatigue; 0.4-0.6 indicates moderate fatigue; 0.6-0.8 indicates significant fatigue; and 0.8-1.0 indicates severe fatigue. This grading system allows for a clear analysis of the locomotive driver's fatigue status.

[0134] The system and method constructed in this embodiment have been further validated in the laboratory, such as... Figure 6 and Figure 7 As shown, the fatigue value curve of driving operation obtained in this embodiment and the fatigue self-assessment scale score curve of the subjects show the same trend, and the correlation analysis shows that the two have good consistency.

[0135] Furthermore, this example embodiment also provides a method for evaluating locomotive driver fatigue based on virtual reality. (Reference) Figure 8 As shown, the method includes steps S101 to S105.

[0136] Step S101: Simulate a driving scenario;

[0137] Step S102: Based on the driving scenario, collect the locomotive driver's work behavior and tasks during driving to obtain driving operation data; wherein, the driving operation data includes the locomotive driver's work data, reaction time, accuracy rate, eye movement data and limb skeletal data;

[0138] Step S103: Preprocess the driving operation data and extract indicator feature values; wherein, the indicator feature values ​​include workload information, task performance information, eye fatigue characteristics, and limb fatigue characteristics;

[0139] Step S104: Time-align the work data and normalize and dimensionless the indicator feature values ​​to obtain workload indicators, task performance indicators, eye movement feature indicators, and limb feature indicators.

[0140] Step S105: The workload, task performance, eye movement features, and limb features are fused to obtain the final evaluation result;

[0141] Step S106: Assess the operational fatigue status of the locomotive driver based on the final evaluation results.

[0142] Specifically, this includes driving records, task execution records, data preprocessing and feature extraction based on raw eye-tracking and limb data, workload calculation, calculation and dimensionless transformation of three-level indicators, calculation of comprehensive evaluation results of locomotive driver work fatigue based on virtual reality, and classification of evaluation results.

[0143] The virtual machine driver's cockpit in the virtual human-machine driving operation unit uses Unity Long Term Support version 2022.3.13f1c1 as the platform for building the virtual reality simulation environment to achieve a high degree of simulation of locomotive driving operation scenarios. Simultaneously, SteamVR plugin version 2.2.3 is integrated, and the corresponding Unity Package is imported to ensure compatibility and interactivity with VR devices.

[0144] The virtual driving tasks in the virtual human-machine driving control unit are implemented using scripts in Unity. Task rounds are initiated using Unity's coroutine functions and the `StartCoroutine()` method, allowing tasks to start, proceed, and end according to set times. Joystick control tasks are implemented using Unity's Canvas and related functions. Road information observation and dashboard information viewing tasks are implemented through interaction between the Unity virtual environment and VR controllers. Each task round lasts 5 minutes, with a 1-minute break between rounds, for a total of ten rounds.

[0145] The virtual reality head-mounted display in the virtual human-machine driving control unit integrates an eye-tracking module.

[0146] The motion capture system in the virtual human-machine driving operation unit adopts an optical motion capture system.

[0147] The virtual vehicle driving record module in the human-machine data acquisition unit mainly records the work behavior and tasks during the driving process, providing a basis for subsequent workload calculation.

[0148] The driving task execution recording module in the human-machine data integration acquisition unit acquires and records data to a CSV file via Unity task scripts. Specifically, at each sampling moment, the system captures the current system time, corresponding reaction time information, and accuracy rate, and writes this data to the CSV file. All data recording is only active during task execution.

[0149] The eye-tracking data acquisition module in the human-machine data integration unit captures eye-tracking data based on the eye-tracking module of the virtual reality head-mounted display. This module can capture and record timestamps, gaze origin, gaze direction, pupil position, pupil size, and eye opening degree during task execution.

[0150] The driving motion acquisition module in the human-machine data integration unit is based on the Nokov series of passive optical motion capture systems with markers. The system works as follows: First, markers are attached to the driver according to the Helen Hayes model. A set of high-performance lenses captures the spatial position information of each marker by detecting the infrared light reflected back from the pre-set reflective markers on their surfaces. The system is equipped with infrared LEDs of specific wavelengths around the lenses, emitting invisible infrared beams into the capture space. Subsequently, the high-reflectivity markers attached to the object reflect this light back to the lenses, where a high-speed image sensor integrated into the lens captures two-dimensional image information. After image analysis algorithms process the data, the coordinates of each marker in the two-dimensional image are converted into three-dimensional spatial coordinates, and the corresponding motion parameters such as velocity and acceleration are further calculated. Based on the built-in Helen Hayes model, the driver's limb skeletal data can be calculated and generated.

[0151] The workload calculation module in the data processing and feature extraction analysis unit includes the calculation of locomotive driver's work behavior time and the calculation of work task requirements. Specifically, work behavior time includes three components: perception, cognition, and operation time. The "Method-Time Measurement" (MTM) method is used to estimate information perception time. T v Using Hick's Law to estimate the cognitive time for decision-making T c Use Fitts' Law to calculate the execution time of the operation. T m Therefore, the average task time characteristic value of the task behavior is calculated as follows:

[0152]

[0153] If a locomotive driver's task does not require any operational actions, then the time for that action is 0, i.e. T v , T c and T m It can be 0.

[0154] The task requirements also include three aspects: perception, cognition, and operation. This implementation adjusts the task requirement range for each basic behavioral element to a 7-point scale (a higher score indicates a higher corresponding task requirement). The specific values ​​for the task requirements of each behavioral element are shown in Table 1. The characteristic value of the average workload is the arithmetic square root of the sum of the squares of the three task requirements. Therefore, the characteristic value of the current workload is obtained by multiplying the average working time by the average workload. The characteristic value of the cumulative workload is obtained by summing the characteristic values ​​of the current workload.

[0155] The task performance calculation module in the data processing and feature extraction analysis unit selects the reaction time captured by the system and written to the CSV file in the experimental data and divides it by the accuracy rate to determine the average reaction time. The average reaction time is then used as a quantitative feature value of task performance.

[0156] The eye-tracking feature extraction module in the data processing and feature extraction analysis unit is divided into two parts: data processing and feature extraction. In eye-tracking data processing, firstly, the pupil diameters of both eyes are selected. If neither pupil diameter is -1, the average of the two pupil diameters is calculated as the pupil diameter; if the data for a single eye is -1, the data for the other eye is retained as the pupil diameter. Secondly, missing pupil diameter data is filled using linear interpolation. Finally, an FIR filter is used to eliminate outliers and noise in the pupil diameter data. In eye-tracking feature extraction, the sum of the pupil diameters at sampling points within the sampling time is divided by the number of sampling points to obtain the average pupil diameter feature value. In the time series of pupil data, if the pupil diameter is less than 0.4 in several consecutive frames, this time period is marked as a complete blink process, and this time period is called the blink time. The sum of the blink times during the sampling period is divided by the number of blinks to obtain the average blink time feature value.

[0157] The limb feature extraction module in the data processing and feature extraction analysis unit is divided into limb data repair preprocessing, animation-redirected data processing, and SoErgo-based limb fatigue feature extraction. Limb data repair preprocessing uses the motion capture system's software to repair the data, automatically calculating interpolation based on the data before and after the missing area to complete the data. Animation-redirected data processing requires exporting all data to FBX files, then redirecting it in MotionBuilder. This involves using the skeletal model built into MotionBuilder to redirect the animation data exported from the motion capture software to animation data usable by SoErgo. SoErgo-based limb fatigue feature extraction uses the software's built-in RULA comfort analysis score as feature values. The current time comfort level is used as the feature value for current discomfort, and the cumulative comfort level over time is used as the feature value for cumulative discomfort.

[0158] In the cross-system data time synchronization and dimensionless characteristic processing unit, because different hardware devices and software tools may use different methods to record data time information, the system time and Unix timestamp are called to convert the Unix timestamp into an easy-to-read date and time format, ensuring that most data achieves millisecond-level time accuracy during the synchronization process, thereby effectively supporting experimental data analysis and integration. For the extracted indicator characteristic values, this implementation adopts the extreme value standardization method to normalize and dimensionless the quantitative indicator values.

[0159] The fuzzy comprehensive evaluation unit for locomotive driver fatigue integrates dimensionless workload, task performance, eye-tracking characteristics, and limb characteristics to obtain the final evaluation result. First, this implementation introduces an expert-based method using triangular fuzzy numbers to construct the membership function. Second, the judgment matrix undergoes confidence correction and consistency checks. Then, defuzzification calculation and root-finding methods are used to obtain the maximum eigenvalue and corresponding eigenvector of the judgment matrix. Normalizing the finally obtained eigenvector yields the comprehensive weight of this level of indicator. Next, the secondary indicators are ranked according to their relative importance to driver fatigue risk to determine the weight values ​​of all tertiary indicators. Finally, the calculated and dimensionless eigenvalues ​​of the tertiary indicators are weighted and summed with their corresponding weights to obtain the final evaluation result.

[0160] Compared to existing methods, this embodiment provides a more objective evaluation method. All data is collected from devices and systems and is not affected by the subjective consciousness of the evaluators. The data collected in this embodiment is comprehensive, including limb movement data that is difficult to measure and evaluate, making it more comprehensive and effective than existing methods. This embodiment incorporates high-tech technologies such as virtual reality and motion capture systems, which not only obtains high-precision objective data, but also brings convenience to the implementation process and greatly improves the evaluation efficiency of the experiment.

[0161] The aforementioned virtual reality-based locomotive driver fatigue assessment system addresses two key issues. First, by constructing highly realistic driving scenarios, it simulates prolonged, high-intensity driving tasks, reproducing various situations that could lead to driver fatigue in a risk-free manner. Second, through high-precision equipment, it monitors and analyzes the driver's physiological responses, behavioral changes, and psychological state in real time during the simulation. This effectively solves the problems of inaccuracy and real-time performance issues found in existing assessment methods, enabling timely and accurate evaluation and detection of the degree of operational fatigue among locomotive drivers during task execution, thus providing safety assurance for locomotive driving. Third, the system provides a more objective, comprehensive, and scientific evaluation method for assessing ergonomic locomotive driver fatigue, promoting the innovative application of advanced technologies such as virtual reality, motion capture, and eye tracking in the field of traffic safety.

[0162] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this disclosure. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. In addition, those skilled in the art can combine and integrate the different embodiments or examples described in this specification.

[0163] Other embodiments of this disclosure will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this disclosure that follow the general principles of this disclosure and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this disclosure are indicated by the appended claims.

Claims

1. A locomotive driver fatigue assessment system based on virtual reality, characterized in that, The system includes: Virtual human-machine driving control unit, used to provide simulated driving scenarios; The human-machine data acquisition unit is used to collect the locomotive driver's work behavior and tasks during driving to obtain driving operation data; wherein, the driving operation data includes the locomotive driver's work data, reaction time, accuracy rate, eye movement data and limb skeletal data; The data processing and feature extraction analysis unit is used to preprocess the driving operation data and extract indicator feature values; wherein, the indicator feature values ​​include workload information, task performance information, eye fatigue characteristics, and limb fatigue characteristics; A cross-system data time synchronization and dimensionless processing unit for various indicator features is used to perform time alignment on the work data and normalize and dimensionless the indicator feature values ​​to obtain workload indicators, task performance indicators, eye movement feature indicators and limb feature indicators. The locomotive driver's work fatigue fuzzy comprehensive evaluation unit is used to fuse the workload, task performance, eye movement features and limb features to obtain the final evaluation result; The locomotive driver's work fatigue comprehensive evaluation result level classification unit is used to assess the work fatigue status of the locomotive driver based on the final evaluation result; The human-machine data acquisition unit includes: The system includes a virtual machine vehicle driving record module, a driving task execution record module, an eye-tracking data acquisition module, and a driving motion acquisition module; among these... The virtual machine vehicle driving record module is used to collect the operation data of the locomotive driver; The driving task execution record module is used to collect the locomotive driver's reaction time and accuracy rate during task execution; The eye-tracking data acquisition module is used to acquire the eye-tracking data of the locomotive driver; The driving motion acquisition module is used to collect the limb skeletal data of the locomotive driver; The data processing and feature extraction analysis unit includes: The module includes a workload calculation module, a task performance calculation module, an eye-tracking feature extraction module, and a limb feature extraction module; among them, The workload calculation module is used to calculate the locomotive driver's work behavior time and work task requirements to obtain workload information; The task performance calculation module is used to quantify the locomotive driver's task performance information based on the average reaction time obtained from the reaction time and the accuracy rate. The eye movement feature extraction module is used to preprocess the eye movement data and extract eye fatigue features from the preprocessed eye movement data; The limb feature extraction module is used to process the limb skeletal data and extract limb fatigue features from the processed limb skeletal data. The workload calculation module includes calculation of locomotive driver operation time and calculation of task requirements. Operation time includes perception time. T v Cognitive Time T c and operation time T m : Perceiving Time T v Method used - Time measurement method for calculation; Cognitive Time T c The expression is: In the formula, n is the number of optional items; For the time required to recognize the selected items; For the first i The frequency of each choice; Operation time T m The expression is: In the formula, D is the operation execution time; W is the manual distance required for the operation; D is the diameter of the object being operated on. The time allotted for the task is: The required workload is as follows: in, This represents the required amount of sensing tasks at time t. This represents the cognitive task requirement at time t. The required number of tasks to be performed at time t; Let i be the continuous symbol for the i-th behavioral element. When this behavior exists at time t, we have: =1, otherwise =0; The number of types of behavioral elements; Let be the perceptual task requirement of the i-th behavioral element at time t; Let be the cognitive task requirement of the i-th behavioral element at time t; Let i be the operational task requirement of the i-th behavioral element at time t; The average workload for each task is: Each round of driving tasks consists of 4 tasks. After calculating the operation time and task requirements for each task, the workload of this round of driving tasks is calculated: in, Indicates the first Average workload of each task Indicates the first The time required for each task's execution; Task performance is: in, Indicates task performance, Indicates reaction time. Indicates accuracy rate; In eye-tracking feature extraction, the average pupil diameter and average blink time are extracted: The average pupil diameter is: in, The number of sampling points in the sample. The diameter of the pupil; The average blink time is: in, The number of blinks, The time it takes to blink; In the cross-system data time synchronization and dimensionless processing unit for various indicator characteristics, the system time and Unix timestamp are called to convert the Unix timestamp into an easy-to-read date and time format, ensuring that most data achieves millisecond-level time accuracy during the synchronization process, thereby effectively supporting experimental data analysis and integration; the integration formula is as follows: Where AX represents the Unix timestamp value to be converted, and Date represents the time when the conversion is completed; For the extracted indicator feature values, the extreme value standardization method is used to normalize and dimensionlessly transform the quantitative indicator values; the specific operation steps are as follows: Positive indicators that contribute more to the objective with larger values ​​include: For inverse indicators where larger values ​​contribute less to the objective, the following are examples: in, This represents the i-th observation of the j-th indicator. and These represent the maximum and minimum values ​​that the j-th indicator can take, respectively, based on the upper and lower limits of the locomotive driver's working time and working conditions standards.

2. The locomotive driver fatigue assessment system based on virtual reality according to claim 1, characterized in that, The virtual human-machine driving operation unit includes: Virtual machine vehicle cockpit, virtual driving missions, virtual reality head-mounted display, and motion capture system; among them, The virtual vehicle cockpit includes a seat, control panel, joystick, and digital model of the locomotive cockpit, used to simulate the operating environment of the locomotive; The virtual driving task includes tasks such as controlling the joystick, using hand gestures to describe driving conditions, observing road information, and viewing dashboard information, and is used to simulate the driving of a locomotive. The virtual reality head-mounted display is used to track the locomotive driver's eye movements; The motion capture device is used to capture the driving actions of the locomotive driver.

3. A method for evaluating locomotive driver fatigue based on virtual reality, characterized in that, Applied to the system as described in any one of claims 1 to 2, the method comprises: Simulated driving scenarios; Based on the driving scenario, the locomotive driver's work behavior and tasks during driving are collected to obtain driving operation data; wherein, the driving operation data includes the locomotive driver's work data, reaction time, accuracy rate, eye movement data and limb skeletal data; The driving operation data is preprocessed, and indicator feature values ​​are extracted; wherein, the indicator feature values ​​include workload information, task performance information, eye fatigue characteristics, and limb fatigue characteristics; The task data is time-aligned, and the indicator feature values ​​are normalized and dimensionless to obtain workload indicators, task performance indicators, eye movement feature indicators, and limb feature indicators. Specifically, this includes: converting the Unix timestamp into a regular date and time format by calling the system time and Unix timestamp; the cross-system data time synchronization and dimensionless feature processing unit uses extreme value normalization to normalize and dimensionless the workload information, task performance information, eye fatigue features, and limb fatigue features to obtain the workload indicators, task performance indicators, eye movement feature indicators, and limb feature indicators. The workload, task performance, eye movement features, and limb features are fused to obtain the final evaluation result. The workload, task performance, eye movement features, and limb features are secondary indicators. The tertiary indicators under the workload indicator are current workload and cumulative workload; the tertiary indicators under the task performance indicator are average reaction time; the tertiary indicators under the eye movement features indicator are average pupil diameter and average blink time; and the tertiary indicators under the limb features indicator are current discomfort and cumulative discomfort. Specifically, this includes: constructing a membership function based on a triangular fuzzy function; and determining the judgment matrix of the membership function. Confidence correction and consistency verification are performed. The judgment matrix after confidence correction and consistency verification is processed using defuzzification calculation and root-finding method to obtain the maximum eigenvalue and corresponding eigenvector of the judgment matrix. The maximum eigenvalue and corresponding eigenvector are normalized to obtain the comprehensive weight of the secondary indicators. The secondary indicators are ranked according to their relative importance to driving fatigue risk, and the comprehensive weight of the secondary indicators is used to determine the weight values ​​of all tertiary indicators. The calculated and dimensionless weight values ​​of the tertiary indicators are weighted and summed with their corresponding weights to obtain the fatigue evaluation value. The final evaluation result is determined based on the fatigue evaluation value. The operational fatigue status of the locomotive driver is assessed based on the final evaluation results.

4. The method for evaluating locomotive driver fatigue based on virtual reality according to claim 3, characterized in that, The final evaluation results include: Fatigue levels: not fatigued, slightly fatigued, somewhat fatigued, noticeably fatigued, and very fatigued; among which, The fatigue rating value is 0~0.2 indicating no fatigue; 0.2~0.4 indicating slight fatigue; 0.4~0.6 indicating moderate fatigue; 0.6~0.8 indicating significant fatigue; and 0.8~1.0 indicating severe fatigue.

Images