A high-speed rail driver fatigue detection method and device based on EEG

Abstract

The present invention provides an EEG-based high-speed rail driver fatigue detection method and device, which belongs to the field of signal recognition technology, and specifically includes: extracting the high-speed rail driver's pedaling frequency, and when the pedaling frequency is less than a first pedaling threshold, extracting the number of times the high-speed rail driver pedals beyond a specified time threshold within a recent first time threshold, and judging that the number is greater than the first threshold, obtaining EEG data based on an EEG sampling module, and performing feature extraction on the EEG data to obtain EEG feature quantities; based on the EEG feature quantities, constructing an EEG prediction model based on an intelligent algorithm to obtain EEG recognition results; obtaining the high-speed rail driver's fatigue based on the EEG recognition results and the number of times, thereby making the fatigue detection result more accurate.

Description

Technical Field

[0001] The present invention belongs to the technical field of signal recognition, and in particular provides a method and device for detecting fatigue of high-speed rail drivers based on EEG. Background Art

[0002] Traditional high-speed train driver warning systems require drivers to step on the pedal at least once every 30 seconds. Otherwise, the train's control system will determine that the driver is driving abnormally. If the driver still hasn't stepped on the pedal after seven seconds, the train will automatically brake for passenger safety. However, this system fails to accurately assess the driver's fatigue level.

[0003] To address the above issues, in his master's thesis "Research on EEG Fatigue Detection for High-Speed Rail Drivers Based on Convolutional Neural Networks," author Yao Di built improved ResNet and MobileNet networks using the Pytorch deep learning framework. He used the cross-entropy loss function and mini-batch gradient descent method to train the preprocessed EEG dataset. He used metrics such as accuracy, precision, and recall to verify the accuracy of the training results and the classification performance of the built networks. However, he did not perform feature extraction on the EEG dataset, which affected the model's prediction accuracy and efficiency. Furthermore, the system was not integrated with the traditional pedal-based high-speed rail driver alarm system, relying solely on the EEG device for judgment. Due to the varying activity levels of EEG datasets across different populations, the detection results were inaccurate. When the pedal-pressing frequency is less than a certain threshold or there is a large amount of over-the-threshold behavior within a certain period of time, it indicates that the high-speed rail driver is in a highly suspicious state of fatigue. Combining this with the detection results of the EEG device further improves the final detection accuracy.

[0004] Based on the above technical problems, it is necessary to design a high-speed rail driver fatigue detection method and device based on EEG. Summary of the Invention

[0005] To achieve the purpose of the present invention, the present invention adopts the following technical solutions:

[0006] In order to solve the above technical problems, the first aspect of the present invention provides a high-speed rail driver fatigue detection method based on EEG, which is characterized by specifically comprising:

[0007] S11 extracts the pedaling frequency of the high-speed train driver, and when the pedaling frequency is less than a first pedaling threshold, proceeds to step S12;

[0008] S12 extracts the number of times the high-speed rail driver pedals beyond the specified time threshold within the most recent first time threshold, and determines whether the number is greater than the first threshold. If so, proceeds to step S13; otherwise, proceeds to step S11;

[0009] S13 obtains EEG data based on the EEG sampling module, and performs feature extraction on the EEG data to obtain EEG feature quantities;

[0010] S14 constructs an EEG prediction model based on an intelligent algorithm based on the EEG feature quantity to obtain an EEG recognition result;

[0011] S15 obtains the fatigue level of the high-speed rail driver based on the EEG recognition result and the number of times.

[0012] By first confirming the pedaling frequency, when the pedaling frequency is greater than a certain threshold, it indicates that the state of the high-speed rail driver at this time is suspicious, and then the number of times the pedaling exceeds the specified time threshold within a certain time threshold is determined, so that the state of the high-speed rail driver within a period of time is determined, and the suspicious state of the high-speed rail driver is further confirmed. On this basis, it is determined that the state of the high-speed rail driver is very suspicious. Then, by extracting the EEG feature quantity and obtaining the EEG prediction result according to the EEG prediction model, the original prediction efficiency and accuracy problem caused by not extracting the EEG prediction result is solved, and the prediction efficiency and accuracy are further improved. On this basis, by obtaining the fatigue level of the high-speed rail driver based on the EEG recognition result and the number of times, the original technical problem of inaccuracy caused by relying solely on the EEG recognition result is solved, the accuracy of the evaluation is further improved, and the foundation for the safe operation of the high-speed rail is laid.

[0013] By first confirming the pedaling frequency, when the pedaling frequency is greater than a certain threshold, the number of times the high-speed rail driver pedals beyond the specified time threshold within the most recent first time threshold is judged. Through double judgment, the status of the high-speed rail driver can be confirmed from a short time dimension and a long time dimension, so that the suspicious status of the high-speed rail driver can be accurately confirmed, further ensuring the reliability and accuracy of the status confirmation.

[0014] By first confirming the suspicious status of the high-speed rail driver and then confirming it through the EEG sampling module, unnecessary sampling and testing are ensured. While reducing the amount of testing, it also ensures the reliability and accuracy of the driver's status detection.

[0015] By extracting EEG feature quantities, the amount of input that the EEG prediction model needs to process is further reduced, and the prediction model can more accurately grasp the characteristics of EEG data, thereby greatly improving the accuracy and reliability of EEG recognition results and improving processing efficiency to a certain extent.

[0016] By obtaining the fatigue level of the high-speed rail driver based on the EEG recognition result and the number of times, the control of the fatigue level of the high-speed rail driver becomes more accurate, avoiding the problems of inaccurate fatigue detection results and poor applicability caused by using a single method, and further improving the safety and stability of high-speed rail operation.

[0017] A further technical solution is that the step of determining the first stepping threshold is:

[0018] S21 uses the expert scoring method to obtain the first basic threshold for stepping on the high-speed rail based on its running speed and passenger capacity;

[0019] S22 determines whether the high-speed rail driver's fatigue level is greater than a first fatigue threshold. If so, proceed to step S23; if not, use the first stepping basic threshold as the first stepping threshold.

[0020] S23 corrects the first stepping basic threshold based on the fatigue level of the high-speed rail driver to obtain a first stepping threshold.

[0021] By using the high-speed rail driver's fatigue to correct the pedaling threshold, the setting of the pedaling threshold can be more accurate. When the driver's fatigue reaches a certain level, the pedaling threshold can be reduced to a certain extent, so that more stringent requirements can be placed on the high-speed rail driver's pedaling frequency, thereby further ensuring the safety of high-speed rail operation.

[0022] A further technical solution is that the first pedaling threshold is smaller than the first pedaling basic threshold.

[0023] A further technical solution is that the first time threshold is determined according to the traveling speed of the train and the driving experience of the high-speed rail driver.

[0024] A further technical solution is that the EEG feature quantities include power spectrum density, sample entropy, and P300 potential.

[0025] A further technical solution is that the specific steps of obtaining the EEG recognition result are:

[0026] S31 determines whether the P300 potential is less than a first potential threshold, and if so, proceeds to step S32;

[0027] S32 constructs an EEG feature based on the power spectrum density, sample entropy, and P300 potential;

[0028] S33 transmits the EEG feature value to the EEG prediction model based on the IGWO-RBF algorithm to obtain the EEG recognition result.

[0029] When the P300 potential is greater than the first potential threshold, it indicates that the high-speed rail driver may be in a relatively obvious state of fatigue. At this time, the EEG recognition results can be confirmed more quickly based on the EEG feature quantity, and the use of the RBF algorithm based on IGWO optimization can further improve the efficiency of the EEG prediction model.

[0030] A further technical solution is that the IGWO algorithm is an improved GWO algorithm, and the number of hidden layers of the RBF algorithm is optimized.

[0031] A further technical solution is that the calculation formula of the fatigue degree is:

[0032] Where G is the EEG recognition result, t is the number of times the high-speed rail driver steps beyond the specified time threshold within the most recent first time threshold, and K1 is the weight.

[0033] A further technical solution is to issue an alarm based on the fatigue level and the EEG recognition result, wherein the specific steps of determining the alarm are:

[0034] S41 determines whether the EEG recognition result is greater than the first recognition threshold, and if so, proceeds to step S43, if not, proceeds to step S42;

[0035] S42 is based on whether the fatigue is greater than the first fatigue threshold, if the process proceeds to step S43, if not, returns to step S41;

[0036] S43 determines whether the number of times the high-speed train driver has stepped on the accelerator beyond the prescribed time threshold in the last 3 minutes is greater than a second threshold, where the first time threshold is greater than 3 minutes. If so, proceed to step S44; if not, return to step S41;

[0037] S44 outputs an abnormal status alarm signal for the high-speed rail driver.

[0038] By step-by-step analysis of EEG recognition results and fatigue levels, and finally combining the number of times the high-speed rail driver has stepped on the pedals exceeding the specified time threshold in the last three minutes, the system can accurately judge the abnormal state of the high-speed rail driver and ensure the reliable and safe operation of the high-speed rail.

[0039] On the other hand, the present invention provides an EEG-based high-speed rail driver fatigue detection device, which adopts the above-mentioned EEG-based high-speed rail driver fatigue detection method, specifically comprising:

[0040] Stepping frequency monitoring module, stepping behavior monitoring module, EEG sampling module, EEG feature extraction module, EEG result recognition module, fatigue level generation module;

[0041] The pedaling frequency monitoring module is responsible for extracting the pedaling frequency of the high-speed rail driver;

[0042] The pedaling behavior monitoring module is responsible for extracting the number of pedaling times exceeding the specified time threshold by the high-speed rail driver within the most recent first time threshold, and transmitting the number to the fatigue generation module;

[0043] The EEG sampling module is responsible for extracting EEG data and transmitting the EEG data to the EEG feature extraction module;

[0044] The EEG feature extraction module is responsible for extracting features based on the EEG data to obtain EEG feature quantities, and transmitting the EEG feature quantities to the EEG result recognition module;

[0045] The EEG result recognition module is responsible for constructing an EEG prediction model based on an intelligent algorithm based on the EEG feature quantity to obtain an EEG recognition result;

[0046] The fatigue generation module is responsible for obtaining the fatigue level of the high-speed rail driver based on the EEG recognition result and the number of times.

[0047] Other features and advantages will be described in the following description, and in part will become apparent from the description, or understood by practicing the invention. The purpose and other advantages of the invention are realized and obtained by the structures particularly pointed out in the description and the drawings.

[0048] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] The above and other features and advantages of the present invention will become more apparent by describing in detail exemplary embodiments thereof with reference to the accompanying drawings.

[0050] Figure 1 This is a flowchart of a high-speed rail driver fatigue detection method based on EEG according to Example 1.

[0051] Figure 2 4 is a flowchart of the steps for determining the first pedaling threshold value according to the first embodiment.

[0052] Figure 3 This is a flowchart of specific steps for obtaining EEG recognition results according to Example 1.

[0053] Figure 4 This is a flowchart of specific steps for determining an alarm according to Example 1.

[0054] Figure 5 This is a diagram of the structure of an EEG-based high-speed rail driver fatigue detection device according to Example 2. DETAILED DESCRIPTION

[0055] Example embodiments will now be described more fully with reference to the accompanying drawings. However, example embodiments can be embodied in many forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete and will fully convey the concepts of the example embodiments to those skilled in the art. Like reference numerals in the figures represent like or similar structures, and thus their detailed description will be omitted.

[0056] The terms "a", "an", "the", and "said" are used to indicate the presence of one or more elements / components / etc.; the terms "including" and "having" are used to express an open-ended inclusive meaning and mean that additional elements / components / etc. may be present in addition to the listed elements / components / etc.

[0057] The original EEG monitoring system for high-speed rail drivers did not perform feature extraction on the EEG dataset, which affected the model's prediction accuracy and efficiency to a certain extent. Moreover, it was not integrated with the traditional pedal-type high-speed rail driver alarm system, and only used the EEG device for judgment. This not only caused a certain form of waste, but also the use of only the EEG device resulted in the final fatigue detection accuracy being not high enough.

[0058] Example 1

[0059] To solve the above problems, according to one aspect of the present invention, Figure 1 As shown, a high-speed rail driver fatigue detection method based on EEG is provided, which is characterized by specifically including:

[0060] S11 extracts the pedaling frequency of the high-speed train driver, and when the pedaling frequency is less than a first pedaling threshold, proceeds to step S12;

[0061] For example, the pedaling frequency is 2 times per minute, and the first pedaling threshold is 3 times per minute. At this time, it is less than the first pedaling threshold, and the process goes to step S12;

[0062] S12 extracts the number of times the high-speed rail driver pedals beyond the specified time threshold within the most recent first time threshold, and determines whether the number is greater than the first threshold. If so, proceeds to step S13; otherwise, proceeds to step S11;

[0063] For example, the first time threshold is 30 minutes, the specified time threshold is 25 seconds, and the first threshold is 10 times. If the number of pedaling times exceeds 25 seconds within 30 minutes is 15 times, and 15 times is greater than 10 times, then the process goes to S13;

[0064] S13 obtains EEG data based on the EEG sampling module, and performs feature extraction on the EEG data to obtain EEG feature quantities;

[0065] S14 constructs an EEG prediction model based on an intelligent algorithm based on the EEG feature quantity to obtain an EEG recognition result;

[0066] For example, the EEG recognition result is between 0 and 1. The larger the EEG recognition result, the higher the driver's fatigue level.

[0067] S15 obtains the fatigue level of the high-speed rail driver based on the EEG recognition result and the number of times.

[0068] By first confirming the pedaling frequency, when the pedaling frequency is greater than a certain threshold, it indicates that the state of the high-speed rail driver at this time is suspicious, and then the number of times the pedaling exceeds the specified time threshold within a certain time threshold is determined, so that the state of the high-speed rail driver within a period of time is determined, and the suspicious state of the high-speed rail driver is further confirmed. On this basis, it is determined that the state of the high-speed rail driver is very suspicious. Then, by extracting the EEG feature quantity and obtaining the EEG prediction result according to the EEG prediction model, the original prediction efficiency and accuracy problem caused by not extracting the EEG prediction result is solved, and the prediction efficiency and accuracy are further improved. On this basis, by obtaining the fatigue level of the high-speed rail driver based on the EEG recognition result and the number of times, the original technical problem of inaccuracy caused by relying solely on the EEG recognition result is solved, the accuracy of the evaluation is further improved, and the foundation for the safe operation of the high-speed rail is laid.

[0069] By first confirming the pedaling frequency, when the pedaling frequency is greater than a certain threshold, the number of times the high-speed rail driver pedals beyond the specified time threshold within the most recent first time threshold is judged. Through double judgment, the status of the high-speed rail driver can be confirmed from a short time dimension and a long time dimension, so that the suspicious status of the high-speed rail driver can be accurately confirmed, further ensuring the reliability and accuracy of the status confirmation.

[0070] By first confirming the suspicious status of the high-speed rail driver and then confirming it through the EEG sampling module, unnecessary sampling and testing are ensured. While reducing the amount of testing, it also ensures the reliability and accuracy of the driver's status detection.

[0071] By extracting EEG feature quantities, the amount of input that the EEG prediction model needs to process is further reduced, and the prediction model can more accurately grasp the characteristics of EEG data, thereby greatly improving the accuracy and reliability of EEG recognition results and improving processing efficiency to a certain extent.

[0072] By obtaining the fatigue level of the high-speed rail driver based on the EEG recognition result and the number of times, the control of the fatigue level of the high-speed rail driver becomes more accurate, avoiding the problems of inaccurate fatigue detection results and poor applicability caused by using a single method, and further improving the safety and stability of high-speed rail operation.

[0073] In another possible embodiment, Figure 2 As shown, the steps for determining the first stepping threshold are:

[0074] S21 uses the expert scoring method to obtain the first basic threshold for stepping on the high-speed rail based on its running speed and passenger capacity;

[0075] To give a specific example, the higher the speed of the high-speed rail and the more passengers it carries, the smaller the basic threshold for the first step.

[0076] S22 determines whether the high-speed rail driver's fatigue level is greater than a first fatigue threshold. If so, proceed to step S23; if not, use the first stepping basic threshold as the first stepping threshold.

[0077] S23 corrects the first stepping basic threshold based on the fatigue level of the high-speed rail driver to obtain a first stepping threshold.

[0078] By using the high-speed rail driver's fatigue to correct the pedaling threshold, the setting of the pedaling threshold can be more accurate. When the driver's fatigue reaches a certain level, the pedaling threshold can be reduced to a certain extent, so that more stringent requirements can be placed on the high-speed rail driver's pedaling frequency, thereby further ensuring the safety of high-speed rail operation.

[0079] In another possible embodiment, the first pedaling threshold is less than the first pedaling basic threshold.

[0080] In another possible embodiment, the first time threshold is determined according to the traveling speed of the train and the driving experience of the high-speed rail driver.

[0081] In another possible embodiment, the EEG feature quantities include power spectrum density, sample entropy, and P300 potential.

[0082] In another possible embodiment, Figure 3 As shown in Figure 2, the specific steps to obtain EEG recognition results are:

[0083] S31 determines whether the P300 potential is less than a first potential threshold, and if so, proceeds to step S32;

[0084] S32 constructs an EEG feature based on the power spectrum density, sample entropy, and P300 potential;

[0085] S33 transmits the EEG feature value to the EEG prediction model based on the IGWO-RBF algorithm to obtain the EEG recognition result.

[0086] When the P300 potential is greater than the first potential threshold, it indicates that the high-speed rail driver may be in a relatively obvious state of fatigue. At this time, the EEG recognition results can be confirmed more quickly based on the EEG feature quantity, and the use of the RBF algorithm based on IGWO optimization can further improve the efficiency of the EEG prediction model.

[0087] In another possible embodiment, the IGWO algorithm is an improved GWO algorithm, which optimizes the number of hidden layers of the RBF algorithm.

[0088] In another possible embodiment, the calculation formula for the fatigue degree is:

[0089]

[0090] Where G is the EEG recognition result, t is the number of times the high-speed rail driver steps beyond the specified time threshold within the most recent first time threshold, and K1 is the weight.

[0091] In another possible embodiment, Figure 4 As shown, an alarm is issued according to the fatigue level and the EEG recognition result, wherein the specific steps of determining the alarm are:

[0092] S41 determines whether the EEG recognition result is greater than the first recognition threshold, and if so, proceeds to step S43, if not, proceeds to step S42;

[0093] S42 is based on whether the fatigue is greater than the first fatigue threshold, if the process proceeds to step S43, if not, returns to step S41;

[0094] S43 determines whether the number of times the high-speed train driver has stepped on the accelerator beyond the prescribed time threshold in the last 3 minutes is greater than a second threshold, where the first time threshold is greater than 3 minutes. If so, proceed to step S44; if not, return to step S41;

[0095] S44 outputs an abnormal status alarm signal for the high-speed rail driver.

[0096] By step-by-step analysis of EEG recognition results and fatigue levels, and finally combining the number of times the high-speed rail driver has stepped on the pedals exceeding the specified time threshold in the last three minutes, the system can accurately judge the abnormal state of the high-speed rail driver and ensure the reliable and safe operation of the high-speed rail.

[0097] Example 2

[0098] like Figure 5 The device for detecting fatigue of a high-speed railway driver based on EEG adopts the above-mentioned method for detecting fatigue of a high-speed railway driver based on EEG, specifically comprising:

[0099] Stepping frequency monitoring module, stepping behavior monitoring module, EEG sampling module, EEG feature extraction module, EEG result recognition module, fatigue level generation module;

[0100] The pedaling frequency monitoring module is responsible for extracting the pedaling frequency of the high-speed rail driver;

[0101] The pedaling behavior monitoring module is responsible for extracting the number of pedaling times exceeding the specified time threshold by the high-speed rail driver within the most recent first time threshold, and transmitting the number to the fatigue generation module;

[0102] The EEG sampling module is responsible for extracting EEG data and transmitting the EEG data to the EEG feature extraction module;

[0103] The EEG feature extraction module is responsible for extracting features based on the EEG data to obtain EEG feature quantities, and transmitting the EEG feature quantities to the EEG result recognition module;

[0104] The EEG result recognition module is responsible for constructing an EEG prediction model based on an intelligent algorithm based on the EEG feature quantity to obtain an EEG recognition result;

[0105] The fatigue generation module is responsible for obtaining the fatigue level of the high-speed rail driver based on the EEG recognition result and the number of times.

[0106] In the embodiments of the present invention, the term "plurality" refers to two or more, unless otherwise specified. Terms such as "installed," "connected," and "fixed" should be interpreted broadly. For example, "connected" can refer to a fixed connection, a detachable connection, or an integral connection. Those skilled in the art will understand the specific meanings of these terms in the embodiments of the present invention based on specific circumstances.

[0107] In the description of the embodiments of the present invention, it should be understood that the terms "upper" and "lower" and the like indicate directions or positional relationships based on the directions or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the embodiments of the present invention and simplifying the description, rather than indicating or implying that the device or unit referred to must have a specific direction, be constructed and operated in a specific direction. Therefore, they cannot be understood as limitations on the embodiments of the present invention.

[0108] Throughout this specification, terms such as "one embodiment" and "a preferred embodiment" mean that the specific features, structures, materials, or characteristics described in conjunction with that embodiment or example are included in at least one embodiment or example of the present invention. In this specification, schematic representations of these 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 any one or more embodiments or examples.

[0109] The above is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Those skilled in the art will readily appreciate that various modifications and variations are possible in the present invention. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present invention are intended to be within the scope of protection of the present invention.

Claims

1. A high-speed rail driver fatigue detection method based on EEG, characterized in that: Specifically include: S11 extracts the pedaling frequency of the high-speed train driver, and when the pedaling frequency is less than a first pedaling threshold, proceeds to step S12; S12 extracts the number of times the high-speed rail driver pedals beyond the specified time threshold within the most recent first time threshold, and determines whether the number is greater than the first threshold. If so, proceeds to step S13; otherwise, proceeds to step S11; S13 obtains EEG data based on the EEG sampling module, and performs feature extraction on the EEG data to obtain EEG feature quantities; S14 constructs an EEG prediction model based on an intelligent algorithm based on the EEG feature quantity to obtain an EEG recognition result; S15 obtains the fatigue level of the high-speed rail driver based on the EEG recognition result and the number of times; The steps for determining the first stepping threshold are: S21 uses the expert scoring method to obtain the first basic threshold for stepping on the high-speed rail based on its running speed and passenger capacity; S22 determines whether the high-speed rail driver's fatigue level is greater than a first fatigue threshold. If so, proceed to step S23; if not, use the first stepping basic threshold as the first stepping threshold. S23: modifying the first stepping basic threshold based on the fatigue level of the high-speed rail driver to obtain a first stepping threshold; The calculation formula of the fatigue degree is: ； Where G is the EEG recognition result, t is the number of times the high-speed rail driver steps on the vehicle beyond the specified time threshold within the most recent first time threshold, and K1 is the weight; An alarm is issued based on the fatigue level and the EEG recognition result, wherein the specific steps of determining the alarm are: S41 determines whether the EEG recognition result is greater than the first recognition threshold, and if so, proceeds to step S43, if not, proceeds to step S42; S42 is based on whether the fatigue is greater than the first fatigue threshold, if the process proceeds to step S43, if not, returns to step S41; S43 determines whether the number of times the high-speed train driver has stepped on the accelerator beyond the prescribed time threshold in the last 3 minutes is greater than a second threshold, where the first time threshold is greater than 3 minutes. If so, proceed to step S44; if not, return to step S41; S44 outputs an abnormal status alarm signal for the high-speed rail driver.

2. The high-speed rail driver fatigue detection method according to claim 1, characterized in that: The first pedaling threshold is less than the first pedaling base threshold.

3. The high-speed rail driver fatigue detection method according to claim 1, characterized in that: The first time threshold is determined according to the travel speed of the train and the driving experience of the high-speed rail driver.

4. The high-speed rail driver fatigue detection method according to claim 1, characterized in that: The EEG feature quantities include power spectrum density, sample entropy, and P300 potential.

5. The high-speed rail driver fatigue detection method according to claim 4, characterized in that: The specific steps to obtain EEG recognition results are: S31 determines whether the P300 potential is less than a first potential threshold, and if so, proceeds to step S32; S32 constructs an EEG feature based on the power spectrum density, sample entropy, and P300 potential; S33 transmits the EEG feature value to the EEG prediction model based on the IGWO-RBF algorithm to obtain the EEG recognition result.

6. The high-speed rail driver fatigue detection method according to claim 5, characterized in that: The IGWO is an improved GWO algorithm that optimizes the number of hidden layers of the RBF.

7. An EEG-based high-speed rail driver fatigue detection device, using the EEG-based high-speed rail driver fatigue detection method according to any one of claims 1 to 6, specifically comprising: Stepping frequency monitoring module, stepping behavior monitoring module, EEG sampling module, EEG feature extraction module, EEG result recognition module, fatigue level generation module; The pedaling frequency monitoring module is responsible for extracting the pedaling frequency of the high-speed rail driver; The pedaling behavior monitoring module is responsible for extracting the number of pedaling times exceeding the specified time threshold by the high-speed rail driver within the most recent first time threshold, and transmitting the number to the fatigue generation module; The EEG sampling module is responsible for extracting EEG data and transmitting the EEG data to the EEG feature extraction module; The EEG feature extraction module is responsible for extracting features based on the EEG data to obtain EEG feature quantities, and transmitting the EEG feature quantities to the EEG result recognition module; The EEG result recognition module is responsible for constructing an EEG prediction model based on an intelligent algorithm based on the EEG feature quantity to obtain an EEG recognition result; The fatigue generation module is responsible for obtaining the fatigue level of the high-speed rail driver based on the EEG recognition result and the number of times.

Images