CN114638538B - A controller fatigue warning method, device, system and warning threshold acquisition method - Google Patents

Abstract

The application provides a controller fatigue early warning method, a device, a system and an early warning threshold value acquisition method, which are used for acquiring fatigue early warning parameters in a preset time period of all control units so as to further acquire fatigue coefficients of the whole control industry and all control units, comparing the group fatigue degree acquired by fatigue test before a post with a group early warning threshold value acquired by predicting the post fatigue requirement of a post to be on duty, comparing the individual fatigue degree of a controller acquired by fatigue monitoring in the post with a controller individual fatigue early warning threshold value acquired by predicting the post on duty fatigue requirement in a future time period of the controller, and realizing fatigue monitoring, prediction and early warning on the whole industry, independent control units, groups and controller individuals in a strategy stage of the group layer, thereby being beneficial to managing and controlling fatigue risks from multiple directions so as to ensure aviation safety.

Description

A controller fatigue warning method, device, system and warning threshold acquisition method

Technical Field

The present application relates to the field of early warning, and specifically to a controller fatigue early warning method, device, system and early warning threshold acquisition method.

Background technique

Ensuring civil aviation transportation safety is the basic responsibility of civil aviation control units. Civil aviation air traffic controllers (also known as controllers) play a key role in civil aviation transportation safety. Therefore, fatigue risk control of controllers on duty has become an important and complex task. In recent years, some technologies have emerged to monitor the fatigue status of controllers in real time. However, these technologies focus more on real-time monitoring of the controller's own on-the-job status. If measures are taken when the controller is found to be fatigued, aviation safety is often at high risk. In addition, ICAO clearly pointed out in the "Manual on Oversight of Fatigue Management Practices" (Doc 9966, Second Edition, 2019 Revision) that fatigue is a physiological state in which the human body's mental or physical strength is reduced, which will impair a person's alertness and ability to perform safety-related operational duties. That is, fatigue judgment is closely related to job safety duties. The management goal of fatigue risk is to ensure that relevant personnel (controllers) maintain sufficient alertness (acceptable fatigue level) when performing their duties. It is obvious that the existing controller fatigue detection/monitoring technology cannot predict in advance whether the current fatigue level of the controller can cope with the needs of the upcoming work, and it is difficult to meet the actual needs of civil aviation safety management. In addition, current technology is unable to make advance predictions and overall assessments of the entire control industry, each control unit, and the teams in the control unit. Therefore, how to achieve fatigue detection, future job fatigue demand prediction, and fatigue warning for the industry as a whole, independent control units, teams, and individual controllers at the strategic stage at the control unit organization level, the pre-tactical (pre-job) stage at the team level, and the tactical (on-duty) stage at the individual level, and then conduct fatigue risk control for different needs from multiple perspectives to ensure aviation safety is a technical problem that the current existing technology urgently needs to solve.

Summary of the invention

In view of the above technical problems, the technical solution adopted in the present application is: a fatigue warning method for air traffic controllers, the method comprising the following steps: S100, judging whether the current demand meets the first fatigue determination demand, the second fatigue determination demand or the third fatigue determination demand; when the current demand meets the first fatigue determination demand, executing step S210, when the current demand meets the second fatigue determination demand, executing step S310, otherwise executing step S410; S210, obtaining fatigue warning parameters Air_P=(Air_P ₁ , Air_P ₂ , Air_P ₃ ,..., _{Air_PN} ) of all control units within a first preset time period, N is the number of all control units, wherein the fatigue warning parameters Air_P _i of the i-th control unit at least include unit type Air_type _i , altitude Air_height _i , control mode Air_manage _i , control area Air_area _i , total number of on-duty air traffic controllers Air_num _i , total number of air traffic controllers on duty at night Air_num_night _i , the total number of controllers on duty during peak traffic period Air_num_high _i , the duty time of controllers Air_time _i , the duty time of controllers at their posts Air_work_time _i , the duty time of controllers at their posts at night Air_night_time _i , the duty time of controllers at their posts during peak traffic period Air_high_time _i , the non-duty time of controllers Air_rwork_time _i , the duty time of controllers at their posts for other operations Air_other_time _i , 1≤i≤N; S220, based on the fatigue warning parameter Air_P, obtaining the overall fatigue coefficient IoF of all control units and/or the unit fatigue coefficient Air_IoF of all control units = [Air_IoF ₁ , Air_IoF ₂ , Air_IoF ₃ , ..., Air_IoF _N ], wherein, Unit fatigue coefficient of the i-th control unit _{Air_Fi1} = 1, which is the weighted coefficient of the air traffic controller's duty time Air_work_time _i of the ith control unit. When the ith control unit is a civil airport, _{Air_Fi2} = 0. Otherwise, _{Air_Fi2} = Air_fm _i × Air_other_time _i / Air_work_time _i . Air_fm _i is the weighted coefficient of other working time of the ith control unit. _{Air_Fi3} is a function of the altitude Air_height _i of the ith control unit. When the hourly aircraft support flight volume of the ith control unit is ≥ the preset flow threshold Air_FT _i , the ith control unit is determined to be in the peak flow period. When the ith control unit is in the non-peak flow period, _{Air_Fi4} = 0. Otherwise, _{Air_Fi4} = Air_num_high _i × Air_high_time _i / Air_work_time _i , _{Air_Fi5} = Air_num_night _i × Air_night_time _i / Air_work_time _i , is a function of the night duty time Air_night_time _i of the ith control unit; Air_F _i6 is a function of the control mode Air_manage _i of the ith control unit, Air_F _i7 is a function of the control area Air_area _i of the ith control unit, and Air_F _i8 is a weighted coefficient of the non-duty time of the ith control unit; S230, fatigue warning is performed on the entire control industry according to the overall fatigue coefficient IoF, and/or fatigue warning is performed on control units with higher fatigue coefficients in different control unit types based on the unit fatigue coefficient Air_IoF and the fatigue warning parameter Air_P; S310, obtain the test fatigue value Set_PreIoF＝[Set_PreIoF ₁ ,Set_PreIoF ₂ ,...,Set_PreIoF _M ] of the team controller of the ith control unit in the second preset time period before taking up the post, where M is the total number of the team controllers, and the test fatigue value Set_PreIoF of the kth controller is _k is obtained according to at least one of the video test data, alertness test data and subjective scale data of the k-th controller, 1≤k≤M; S320, calculating the warning value Set_AL=[Set_AL ₁ , Set_AL ₂ , ..., Set_AL _M ] of the team controller, and giving a warning based on the warning value Set_AL, wherein Set_AL _k =Set_PreIoF _k -Set_ALP _k , Set_ALP _k is the fatigue warning threshold of the k-th controller for the post to be on duty, and the team fatigue warning threshold Set_ALP=[Set_ALP ₁ , Set_ALP ₂ , ..., Set_ALP _M ], obtain the warning threshold parameters within the specified time interval of the team's upcoming duty time, the warning threshold parameters at least including the team's duty time, predicted flight volume, predicted duty duration, predicted weather conditions, predicted special conditions, and predicted attributes of the controller itself; S410, obtain a controller's comprehensive fatigue level value Person_IoF = Person_C ₁ ×F_pose+Person_C ₂ ×F_face+Person_C ₃ ×F_voice, where F_pose is the current controller fatigue level value based on behavior and posture, F_face is the current controller fatigue level value based on facial features, F_voice is the current controller fatigue level value based on air-ground communication, Person_C ₁ is the weight coefficient of F_pose, Person_C ₂ is the weight coefficient of F_face, and Person_C ₃ is the weight coefficient of F_voice; S420, calculate the warning value Person_AL=Person_IoF-Person_ALP of the controller Person, and give a warning based on the warning value Person_AL, wherein the controller's individual fatigue warning threshold Person_ALP is obtained according to the warning threshold judgment parameters set in the future time zone during the duty period of the controller Person on duty, and the warning threshold judgment parameters include at least the predicted flight volume related to the duty post, the predicted weather conditions, the predicted special situation, and the predicted attributes of the controller Person itself.

A fatigue warning device for a traffic controller comprises: a first data acquisition module, for obtaining fatigue warning parameters Air_P=(Air_P ₁ , Air_P ₂ , Air_P ₃ , ..., _{Air_PN} ) of all control units within a first preset time period when a current demand meets a first fatigue determination demand, wherein N is the number of all control units, wherein the fatigue warning parameter Air_P _i of the i-th control unit at least comprises unit type Air_type _i , altitude Air_height _i , control mode Air_manage _i , control area Air_area _i , total number of on-duty controllers Air_num _i , total number of on-duty controllers at night Air_num_night _i , total number of on-duty controllers during peak traffic flow period Air_num_high _i , duty time of controller Air_time _i , duty time of controller post Air_work_time _i , duty time of controller post at night Air_night_time _i , duty time of controller post during peak traffic flow period Air_high_time _i , non-job duty time of the controller Air_rwork_time _i , job duty time of the controller for other operations Air_other_time _i , 1≤i≤N; a first data processing module, used to obtain the overall fatigue coefficient IoF of all control units and/or the unit fatigue coefficient Air_IoF of all control units based on the fatigue warning parameter Air_P = [Air_IoF ₁ , Air_IoF ₂ , Air_IoF ₃ , ..., Air_IoF _N ], wherein, Unit fatigue coefficient of the i-th control unit _{Air_Fi1} = 1, which is the weighted coefficient of the air traffic controller's duty time Air_work_time _i of the ith control unit. When the ith control unit is a civil airport, _{Air_Fi2} = 0. Otherwise, _{Air_Fi2} = Air_fm _i × Air_other_time _i / Air_work_time _i . Air_fm _i is the weighted coefficient of other working time of the ith control unit. _{Air_Fi3} is a function of the altitude Air_height _i of the ith control unit. When the hourly aircraft support flight volume of the ith control unit is ≥ the preset flow threshold Air_FT _i , the ith control unit is determined to be in the peak flow period. When the ith control unit is in the non-peak flow period, _{Air_Fi4} = 0. Otherwise, _{Air_Fi4} = Air_num_high _i × Air_high_time _i / Air_work_time _i , _{Air_Fi5} = Air_num_night _i × Air_night_time _i / Air_work_time _i , is a function of the night duty time Air_night_time _i of the ith control unit; Air_F _i6 is a function of the control mode Air_manage _i of the ith control unit, Air_F _i7 is a function of the control area Air_area _i of the ith control unit, and Air_F _i8 is a weighted coefficient of the non-duty time of the ith control unit; a first fatigue warning module, used to perform fatigue warning on the entire control industry according to the overall fatigue coefficient IoF, and/or based on the unit fatigue coefficient Air_IoF and the fatigue warning parameter Air_P, perform fatigue warning on the control units of different control unit types with higher fatigue coefficients; a second data acquisition module, used to obtain the test fatigue value Set_PreIoF=[Set_PreIoF ₁ ,Set_PreIoF 2 ,...,Set_PreIoF _{M] of} the team controller of the ith control unit within the second preset time period before taking up the post when the current demand meets the second fatigue judgment demand ], M is the total number of controllers in the team, wherein the test fatigue value Set_PreIoF _k of the k-th controller is obtained according to at least one of the video test data, alertness test data and subjective scale data of the k-th controller, 1≤k≤M; a second data processing module is used to calculate the warning value Set_AL＝[Set_AL ₁ ,Set_AL ₂ ,...,Set_AL _M ] of the team controller, wherein Set_AL _k ＝Set_PreIoF _k -Set_ALP _k , Set_ALP _k is the fatigue warning threshold of the k-th controller about the post to be on duty, and the team fatigue warning threshold Set_ALP＝[Set_ALP ₁ ,Set_ALP ₂ ,...,Set_ALP _M ], based on the warning threshold parameters within the specified time interval of the team's upcoming duty period, the warning threshold parameters at least include the team's duty time, predicted flight volume, predicted duty duration, predicted weather conditions, predicted special situations, and predicted attributes of the controller itself; the second fatigue warning module is used to issue a warning based on the warning value Set_AL. The third data acquisition module is used to obtain a controller's comprehensive fatigue level value Person_IoF = Person_C ₁ ×F_pose+Person_C ₂ ×F_face+Person_C ₃ ×F_voice of a controller Person when the current demand meets the third fatigue judgment demand, where F_pose is the current controller fatigue level value based on behavioral posture, F_face is the current controller fatigue level value based on facial features, F_voice is the current controller fatigue level value based on land-air communication, Person_C ₁ is the weight coefficient of F_pose, Person_C ₂ is the weight coefficient of F_face, and Person_C ₃ is the weight coefficient of F_voice; the third data processing module is used to calculate the warning value Person_AL of the controller Person = Person_IoF-Person_ALP, wherein the individual warning threshold Person_ALP of the controller is obtained according to the warning threshold judgment parameters set in the future time area during the duty period of the controller Person on duty, and the warning threshold judgment parameters at least include the predicted flight volume related to the duty post, the predicted weather conditions, the predicted special situation, and the predicted attributes of the controller Person itself; the third fatigue warning module is used to issue a warning based on the warning value Person_AL.

A controller fatigue warning system includes a processor and a non-transitory storage medium for storing at least one instruction or at least one program. The at least one instruction or at least one program is loaded and executed by the processor to implement the fatigue warning method described above.

A method for obtaining an overall fatigue coefficient warning threshold value Total_ALP of a regulated industry comprises the following steps: S231, obtaining the overall fatigue coefficient and the number of unsafe events of all regulated units in a plurality of different historical time periods; S232, performing data-driven iteration using the overall fatigue coefficient and the number of unsafe events in the plurality of different historical time periods to obtain the overall fatigue coefficient warning threshold value Total_ALP.

A method for obtaining an overall fatigue coefficient of all control units and a unit fatigue coefficient of each unit, the method comprising the following steps: S210, obtaining fatigue warning parameters Air_P=(Air_P ₁ , Air_P ₂ , Air_P ₃ , ..., _{Air_PN} ) of all control units in a first preset time period, where N is the number of all control units, wherein the fatigue warning parameters Air_P _i of the i-th control unit at least include unit type Air_type _i , altitude Air_height _i , control mode Air_manage _i , control area Air_area _i , total number of on-duty controllers Air_num _i , total number of on-duty controllers at night Air_num_night _i , total number of on-duty controllers during peak traffic period Air_num_high _i , controller duty time Air_time _i , controller post duty time Air_work_time _i , controller post duty time at night Air_night_time _i , controller post duty time during peak traffic period Air_high_time _i , the non-job duty time of the controller Air_rwork_time _i , the job duty time of the controller for other operations Air_other_time _i , 1≤i≤N; S220, based on the fatigue warning parameter Air_P, obtaining the overall fatigue coefficient IoF of all control units and/or the unit fatigue coefficient Air_IoF of all control units = [Air_IoF ₁ , Air_IoF ₂ , Air_IoF ₃ , ..., Air_IoF _N ], wherein, Unit fatigue coefficient of the i-th control unit _{Air_Fi1} = 1, which is the weighted coefficient of the air traffic controller's duty time Air_work_time _i of the ith control unit. When the ith control unit is a civil airport, _{Air_Fi2} = 0. Otherwise, _{Air_Fi2} = Air_fm _i × Air_other_time _i / Air_work_time _i . Air_fm _i is the weighted coefficient of other working time of the ith control unit. _{Air_Fi3} is a function of the altitude Air_height _i of the ith control unit. When the hourly aircraft support flight volume of the ith control unit is ≥ the preset flow threshold Air_FT _i , the ith control unit is determined to be in the peak flow period. When the ith control unit is in the non-peak flow period, _{Air_Fi4} = 0. Otherwise, _{Air_Fi4} = Air_num_high _i × Air_high_time _i / Air_work_time _i , _{Air_Fi5} = Air_num_night _i × Air_night_time _i / Air_work_time _i , is a function of the night duty time Air_night_time _i of the ith control unit, Air_F _i6 is a function of the control method Air_manage _i of the ith control unit, Air_F _i7 is a function of the control area Air_area _i of the ith control unit, and Air_F _i8 is the weighted coefficient of the non-duty time of the ith control unit.

A method for obtaining a team fatigue warning threshold Set_ALP comprises: S331, obtaining warning threshold parameters within a specified time interval in a time period during which the team is to be on duty, wherein the warning threshold parameters at least include the duty time of the team, predicted flight volume, predicted duty duration, predicted special conditions, predicted weather conditions, and predicted attributes of the controller itself, and the duration of the specified time interval has a value range of [20 minutes, 300 minutes]; S332, obtaining a team basic fatigue value Set_P=[Set_P ₁ , Set_P ₂ , ..., Set_P _M ] of the team according to the warning threshold parameters, wherein Set_P _k represents the basic fatigue value of the kth controller of the team; S333, obtaining a team fatigue warning threshold Set_ALP=[Set_ALP ₁ , Set_ALP ₂ , ..., Set_ALP _M ] of the team based on Set_P, wherein Set_ALP _k =Set_C×Set_P _k , and Set_C is a team post adjustment coefficient.

A method for obtaining an individual fatigue warning threshold Person_ALP of a controller, comprising: S431, obtaining a warning threshold determination parameter set in a future time zone in a post duty time period of a controller Person on duty, wherein the warning threshold determination parameter at least includes a predicted flight volume, a predicted weather condition, a predicted special situation, and a predicted attribute of the controller Person, and the duration of the set future time zone has a value range of [3 minutes, 20 minutes]; S432, obtaining a basic requirement value Person_F _b of fatigue degree of the controller on duty according to the warning threshold determination parameter; S433, obtaining the individual fatigue warning threshold Person_ALP of the controller according to the basic requirement value Person_F _b of fatigue degree of the controller on duty = (Person_A1+Person_A2×Person_A3)×Person_F _b Person_A1 is the circadian rhythm coefficient of the controller during the duty time; Person_A2 is the position coefficient, which is used to indicate the fatigue level required by the controller's position; Person_A3 is the position task requirement coefficient, which is used to indicate the fatigue level required by the expected flight plan traffic flow situation.

This application has at least the following technical effects: by obtaining fatigue warning parameters of all control units within a preset time period, the fatigue coefficient of the entire control industry or even each control unit is obtained, and based on the fatigue coefficient, fatigue monitoring, prediction and warning are performed on the entire control industry and high-risk control units, teams and individuals. In addition, by comparing the fatigue level of the team obtained by the pre-job fatigue test with the team warning threshold obtained by predicting the job fatigue requirements of the job that the team will be on duty, and by comparing the personal fatigue level of the controller obtained by the fatigue monitoring during the job with the individual fatigue warning threshold of the controller obtained by predicting the job fatigue requirements of the controller in the future, it is possible to detect whether the control personnel can cope with the work requirements of the job to be on duty before the job, and even after the controller takes up the job, it is possible to judge whether the personnel on the job will have fatigue abnormalities in the next 3-20 minutes, that is, to predict from multiple angles whether the controller can cope with future aviation safety needs. It realizes fatigue detection, prediction of future job fatigue demand and fatigue warning for the industry as a whole, independent control units, teams and individual controllers at the strategic stage at the organizational level of control units, the pre-tactical (pre-job) stage at the team level, and the tactical (on-duty) stage at the individual level. This is conducive to fatigue risk management for different needs from multiple angles and ensure aviation safety.

BRIEF DESCRIPTION OF THE DRAWINGS

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

FIG1 is a flow chart of a controller fatigue warning method provided by an embodiment of the present application;

FIG2 is a structural diagram of a controller fatigue warning device provided by an embodiment of the present application;

FIG3 is a method for obtaining a team fatigue warning threshold Set_ALP provided by another embodiment of the present application;

FIG4 is a method for obtaining an individual fatigue warning threshold Person_ALP of a controller provided by another embodiment of the present application;

FIG5 is a flow chart of a controller fatigue warning method provided in another embodiment of the present application.

Detailed ways

The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

FIG1 is a controller fatigue warning method provided by an embodiment of the present application, the method comprising the following steps:

S100, judging whether the current demand meets the first fatigue determination demand, the second fatigue determination demand or the third fatigue determination demand. When the current demand meets the first fatigue determination demand, executing step S210, when the current demand meets the second fatigue determination demand, executing step S310, otherwise executing step S410. The first fatigue determination demand is, for example, the fatigue determination demand at the strategic stage of the control unit organization level, the second fatigue determination demand is, for example, the fatigue determination demand at the pre-tactical stage of the team level, and the third fatigue determination demand is, for example, the fatigue determination demand at the tactical stage of the individual level. By distinguishing different stages, different demands can be met from three different aspects. For example, at the strategic stage of the control unit organization level, the fatigue coefficient of the entire control industry can be calculated, and the current status of the entire control industry can be judged as a whole, and the entire control industry can be regulated according to the current status. For example, if the number of control employees is insufficient, the control personnel can be increased. The fatigue coefficient of different control units can also be calculated, and the relevant situation of the control unit can be judged, and employee adjustments can be made accordingly.

S210, obtaining fatigue warning parameters Air_P=(Air_P ₁ , Air_P ₂ , Air_P ₃ , ..., _{Air_PN} ) of all control units in a first preset time period, where N is the number of all control units, wherein the fatigue warning parameters Air_P _i of the i-th control unit at least include unit type Air_type _i , altitude Air_height _i , control mode Air_manage _i , control area Air_area _i , total number of on-duty controllers Air_num _i , total number of on-duty controllers at night Air_num_night _i , total number of on-duty controllers during peak traffic period Air_num_high _i , duty time of controllers Air_time _i , duty time of controllers at work Air_work_time _i , duty time of controllers at work at night Air_night_time _i , duty time of controllers at work during peak traffic period Air_high_time _i , and non-duty time of controllers Air_rwork_time _i , the duty time of the controller for other operations Air_other_time _i , 1≤i≤N.

In the present application, the first preset time period can be a historical time period or a future time period, and the length of the time period can be set according to specific needs, for example, it can be a week, a month, etc. When the first preset time period is a future time period, the relevant fatigue warning parameters can be obtained according to the flight plan of the control unit and the working hours of the controller, and the safety risk issues of the control industry and/or control unit in the future time period can be judged based on the fatigue parameters, so as to make employee changes in advance.

In the present application, the type of control unit Air_type _i can be a plurality of pre-set types, for example, it can be an area control type control unit, an approach control type control unit, a tower control and a civil control unit with an annual throughput of 500,000 to 2 million, etc. The total number of control unit types can be divided according to actual conditions. The altitude Air_height _i of the control unit can be obtained through a variety of channels, such as querying geographic information, etc. The control method Air_manage _i can be divided into four major types: procedural control, radar control, radar surveillance control and others. The control area Air_area _i is measured according to actual conditions. The total number of on-duty controllers Air_num _i of the i-th control unit, the total number of controllers on duty at night Air_num_night _i of the i-th control unit, and the total number of controllers on duty during peak traffic periods Air_num_high _i of the i-th control unit can be obtained by collecting the unique identifier of the controller and the actual working time of the controller, and on the other hand, the various parameters can be obtained according to the flight plan arrangement. The same method can also be used to obtain the controller's duty time Air_time _i of the i-th control unit, the controller's post duty time Air_work_time _i of the i-th control unit, the controller's night post duty time Air_night_time _i of the i-th control unit, the controller's post duty time during peak traffic period Air_high_time _i of the i-th control unit, the controller's non-post duty time Air_rwork_time _i , and the controller's duty time for other operations of the control unit Air_other_time _i . In addition, in the present application, when the aircraft hourly support flight volume of the i-th control unit is ≥ the preset traffic threshold Air_FT _i , the i-th control unit is determined to be in the peak traffic period. The preset traffic threshold Air_FT _i is set according to the actual flight data or flight plan of the control unit.

S220, based on the fatigue warning parameter Air_P, obtaining the overall fatigue coefficient IoF of all control units and/or the unit fatigue coefficient Air_IoF of all control units = [Air_IoF ₁ , Air_IoF ₂ , Air_IoF ₃ , ..., Air_IoF _N ], wherein: Unit fatigue coefficient of the i-th control unit _{Air_Fi1} = 1, which is the weighted coefficient of the air traffic controller's duty time Air_work_time _i of the ith control unit. When the ith control unit is a civil airport, _{Air_Fi2} = 0, otherwise, _{Air_Fi2} = Air_fm _i × Air_other_time _i / Air_work_time _i , where Air_fm _i is the weighted coefficient of other working time of the ith control unit, and its value range is [1.00, 4.00], preferably 1.80; _{Air_Fi3} is a function of the altitude Air_height _i of the ith control unit, and its value range is [0.00, 8.00], preferably 4.21; in the present application, when the hourly aircraft support flights of the ith control unit ≥ the preset flow threshold Air_FT _i , the ith control unit is determined to be in the peak flow period, and when the ith control unit is in the non-peak flow period, _{Air_Fi4} = 0; otherwise, _{Air_Fi4} =Air_num_high _i ×Air_high_time _i /Air_work_time _i , Air_F _i5 =Air_num_night _i ×Air_night_time _i /Air_work_time _i , which is a function of the night duty time Air_num_night _i of the controller of the i-th control unit; Air_F _i6 is a function of the control mode Air_manage _i of the i-th control unit, and its value range is [0.00,6.00]; Air_F _i7 is a function of the control area Air_area _i of the i-th control unit, and its value range is [0.00,7.00]; Air_F _i8 is the weighted coefficient of the non-duty time of the i-th control unit, and its value range is [0.20,1.00], and is preferably 0.64.

According to the above content, this application first obtains the relevant objective parameters that affect the control units and control industries, including the unit type Air_type _i , altitude Air_height _i , control mode Air_manage _i , control area Air_area _i , total number of on-duty controllers Air_num _i , total number of on-duty controllers at night Air_num_night _i , total number of on-duty controllers during peak traffic Air_num_high _i , duty time of controllers Air_time _i , duty time of controllers at work Air_work_time _i , duty time of controllers at work at night Air_night_time _i , duty time of controllers at work during peak traffic Air_high_time _i , non-duty time of controllers at work Air_rwork_time _i , duty time of controllers at work for other operations Air_other_time _i , and on the basis of these parameters combined with specific calculation methods, the overall fatigue coefficient of the regulated industry and the fatigue coefficient of each regulated unit are calculated. By integrating multiple influencing factors, the actual fatigue degree of the entire industry and each regulated unit can be evaluated more accurately and objectively, making the fatigue coefficient acquisition method more applicable and generalizable.

S230, fatigue warning is performed on the entire regulated industry according to the overall fatigue coefficient IoF, and/or fatigue warning is performed on the regulated units of the group with higher fatigue coefficients among different regulated unit types based on the unit fatigue coefficient Air_IoF and the fatigue warning parameter Air_P. Specifically, in one embodiment, fatigue warning is performed on the regulated units of the group with higher fatigue coefficients among different regulated unit types based on the unit fatigue coefficient Air_IoF and the fatigue warning parameter Air_P as follows: different types of regulated units are sorted according to the size of fatigue coefficients, and then warning is performed on the regulated units with higher numerical ranking among different types of regulated units, wherein the regulated unit type data can be obtained from the fatigue warning parameter Air_P, and the corresponding fatigue-related data is obtained from the fatigue coefficient Air_IoF. For example, there are 4 regulated units of the first type, and the corresponding fatigue coefficients are 200, 100, 150, and 300, respectively. Then, after sorting by size, they are 300, 200, 150, and 100. When warning is required for the first two, warning is required for the two regulated units with fatigue coefficients of 300 and 200, respectively. Among them, there are 4 examples of the first type of control units for the convenience of explaining the warning method, which does not represent the actual number of control units. The actual data of the first type of control units shall be based on the actual number obtained. In this application, a warning alarm benchmark can be set. When the fatigue coefficient of the control unit is greater than or equal to the warning alarm benchmark, the control unit will be warned, otherwise no warning is required. The warning alarm benchmark can be obtained based on historical experience or in other ways. The method for obtaining the warning alarm benchmark described in this application is not subject to specific restrictions. In another embodiment, based on the unit fatigue coefficient Air_IoF and the fatigue warning parameter Air_P, fatigue warning is performed on control units with higher fatigue coefficients in different control unit types as follows: first, all control units are classified based on the fatigue warning parameter Air_P, and then, each type of control unit is sorted according to the size of the fatigue coefficient. Then, each type of control unit after sorting is graded using a clustering method and divided into groups with lower, medium, and higher fatigue coefficients. Finally, in each type of control unit, a warning is given to the control unit that enters a higher value group after clustering. Clustering the fatigue coefficients of each control unit by clustering can better mine the inherent distribution law between data. Specifically, the commonly used K-mean clustering method can be used to cluster the fatigue coefficients, and the clustering method used in this application can be any clustering method in the prior art.

Specifically, fatigue warning is performed on the regulated industry as a whole according to the overall fatigue coefficient IoF. In one embodiment, the overall fatigue coefficient IoF can be compared with the historical overall fatigue coefficient of the entire industry. When the overall fatigue coefficient IoF is larger than the historical overall fatigue coefficient, a warning is performed. In another embodiment, when the overall fatigue coefficient IoF increases to a certain degree compared with the historical overall fatigue coefficient, a warning is performed. Specifically, the certain degree can be combined with the unsafe events within the first preset time period of the regulated unit (when the first preset time period is a historical time period) to determine whether to issue a warning. In another embodiment, a warning is performed when the overall fatigue coefficient IoF ≥ the overall fatigue coefficient warning threshold Total_ALP, wherein the overall fatigue coefficient warning threshold Total_ALP is obtained by data-driven iteration based on the overall fatigue coefficients of all regulated units and the number of unsafe events of all regulated units obtained in multiple different historical time periods. For example, the overall fatigue coefficient IoF_1 of all control units in the first historical time period is a first value, and the number of unsafe events is 0; the overall fatigue coefficient IoF_2 in the second historical time period is a second value, and the number of unsafe events is 0. Although the second value IoF_2 is greater than the first value IoF_1, because the number of unsafe events in the second historical time period is 0, IoF_2 is not used as the overall fatigue coefficient warning threshold Total_ALP. If the overall fatigue coefficient IoF_3 in the third historical time period is a third value, and the number of unsafe events is 2, because the third value IoF_3 is greater than the second value IoF_2, and more unsafe events occurred in the third historical time period, IoF_3 can be used as the overall fatigue coefficient warning threshold Total_ALP. By analogy, the overall fatigue coefficient and unsafe events in each historical time period can be comprehensively considered to finally obtain the overall fatigue coefficient warning threshold Total_ALP. Furthermore, the duration of the first historical time period, the second historical time period, and the third historical time period may be one month, one week, etc., and appropriate selection of the statistical duration may enable the overall fatigue coefficient warning threshold value Total_ALP to be quickly and effectively obtained.

The embodiment of the present invention also discloses a method for obtaining an overall fatigue coefficient warning threshold value Total_ALP, comprising the following steps:

S231, obtaining the overall fatigue coefficient and the number of unsafe events of all the control units in multiple different historical time periods;

S232: Perform data-driven iteration using the overall fatigue coefficient and the number of unsafe events in multiple different historical time periods to obtain the overall fatigue coefficient warning threshold Total_ALP.

By obtaining the objective indicators of the control units, we can obtain the overall fatigue coefficient of the control industry and the unit fatigue coefficient of each control unit. The overall fatigue coefficient and the unit fatigue coefficient can effectively and objectively detect whether there are safety risks in the entire control industry and the control units. When there are greater safety risks, on the one hand, we can control the relevant situation of the control industry as a whole and make appropriate adjustments. On the other hand, we can also promptly remind the control units in the group with higher fatigue coefficients to make appropriate adjustments to reduce fatigue risks.

S310, obtain the test fatigue value Set_PreIoF＝[Set_PreIoF ₁ , Set_PreIoF ₂ ,..., Set_PreIoF _M ] of the team controller of the i-th control unit before taking up the post, where M is the total number of the team controllers. The length of the second preset time period can be set by oneself, such as 10 minutes, 5 minutes, 3 minutes, etc. Preferably, in the present application, the length of the second preset time period is 5 minutes, which can ensure the authenticity of the controller's detection during the test on the one hand, and avoid pseudo fatigue of the controller due to the long test time on the other hand, thereby affecting the authenticity of the test. Specifically, in the present application, the test fatigue value Set_PreIoF _k of the k-th controller is obtained based on at least one of the video test data, alertness test data and subjective scale data of the k-th controller, 1≤k≤M. In one embodiment, Set_PreIoF _k =A _k1 ×W _k +A _k2 ×U _k +A _k3 ×V _k , where A _k1 , A _k2 , and A _k3 are weight coefficients, the value range of A _k1 is [0, 3], preferably 2.4, the value range of A _k2 is [4, 8], preferably 6, and the value range of A _k3 is [2, 5], preferably 3. Wherein, the alertness test data V _k = (Vo+Ve)/Vt, where Vo is the number of ignored target letters, Ve is the number of selected incorrect letters, and Vt is the total number of target letters; the alertness test data U _k = t/t _T , t is the correct reaction time during the alertness test, and t _T = 500 ms, which is the reaction time threshold. Video test data Wherein FPs represents the frame rate per second, tv represents the length of the video, W(l) represents the coverage of the eyeball of the person being tested at the lth frame, and sign{} represents the indicator function, where sign{true}=1, sign{false}=0, 1≤l≤p, p=FPs×tv, which is the number of frames of the video.

S320, calculate the warning value Set_AL=[Set_AL ₁ , Set_AL ₂ ,..., Set_AL _M ] of the team controller, and issue a warning based on the warning value Set_AL, wherein Set_AL _k =Set_PreIoF _k -Set_ALP _k , Set_ALP _k is the fatigue warning threshold of the kth controller for the post to be on duty, and the team fatigue warning threshold Set_ALP=[Set_ALP ₁ , Set_ALP ₂ ,..., Set_ALP _M ] is obtained according to the warning threshold parameters in the specified time interval of the team's upcoming duty period, and the warning threshold parameters at least include the team's duty time, predicted flight volume, predicted duty duration, predicted weather conditions, predicted special conditions, and predicted attributes of the controller itself. Specifically, when Set_AL _k ≥Set_ALP _k , that is, Set_AL _k ≥0, the system issues a warning to the controller, and the team can promptly arrange other controllers that meet the management requirements to replace them.

In one embodiment of the present invention, a method for obtaining a team fatigue warning threshold Set_ALP is also disclosed. As shown in FIG3 , the method includes:

S331, obtain the warning threshold parameters within the specified time interval in the duty time period of the team, and the warning threshold parameters at least include the duty time of the team, the predicted number of flights, the predicted duty time, the predicted weather conditions, the predicted special situation, and the predicted attributes of the controller itself. The duration of the specified time interval ranges from [20 minutes to 300 minutes], preferably 120 minutes. For example, if the team starts work at 4 pm, then we set the specified time interval to 4 to 6 pm. The duty time can be, for example, morning, evening, etc. The number of flights is the predicted number of flights, which can be predicted based on the flight plan of the control unit. For example, the number of flights in the flight plan within the specified time zone is 3, but temporary landing flights are not excluded. At this time, it is necessary to predict the number of flights of the team in the specified time interval based on the above-mentioned flight plan and temporary factors. Similarly, the duty time is also the duty time predicted by the team, such as 1 hour, 2 hours, etc. The attributes of the controller itself include fatigue status, circadian rhythm type, recent sleep status, previous duty status, gender, age, length of service, etc.

S332, according to the warning threshold parameter, obtain the team basic fatigue value Set_P = [Set_P ₁ , Set_P ₂ , ..., _{Set_PM} ] of the team, wherein Set_P _k represents the basic fatigue value of the kth controller of the team. Specifically, in this embodiment, the basic fatigue value Set_P _k of the kth controller is a function of the warning threshold parameter of the controller, for example, the duty time assigned to the team, the predicted number of flights, the predicted duty time, the predicted weather conditions, and the predicted attributes of the kth controller itself are assigned different weight values and then summed to obtain the basic fatigue value Set_P _k .

S333, based on Set_P, obtain the team fatigue warning threshold Set_ALP = [Set_ALP ₁ , Set_ALP ₂ , ..., Set_ALP _M ] of the team, wherein Set_ALP _k = Set_C × Set_P _k , Set_C is the team post adjustment coefficient, and its value range is [1.0, 5.0], such as the post adjustment of the control seat and the coordination seat, and the post adjustment of different sectors of the control seat. The adjustment coefficient expresses the maximum internal post adjustment amount that the team can allow in order to cope with the job to be on duty. By using the team post adjustment coefficient, the warning standard can be made more in line with the actual situation, and thus the warning can be made more accurate. In addition, it can also satisfy the team's overall grasp of the job to be on duty, which is convenient for the controller to adjust the work deployment. Further, Set_P is continuously optimized according to the false alarm rate and missed alarm rate of the warning.

By conducting fatigue tests on the team's controllers before taking up their posts, the most realistic fatigue level of the controllers before taking up their posts can be obtained. Since the team's basic fatigue value is obtained based on the team's plans for the posts they will be on duty and the controller's own attributes, it can more objectively reflect the team's actual needs within a specified time in the future. By comparing the two, the team can more accurately and effectively obtain the fatigue status of the entire team's controllers, and make work adjustments accordingly, thereby controlling safety risks caused by controller fatigue.

S310, obtaining a comprehensive fatigue level value of a controller Person Person_IoF = Person_C ₁ ×F_pose+Person_C ₂ ×F_face+Person_C ₃ ×F_voice, wherein F_pose is the current fatigue level value of the controller based on behavior and posture, F_face is the current fatigue level value of the controller based on facial features, F_voice is the current fatigue level value of the controller based on ground-to-air communication, Person_C ₁ is the weight coefficient of F_pose, Person_C ₂ is the weight coefficient of F_face, and Person_C ₃ is the weight coefficient of F_voice. The value range of Person_C ₁ is [0.00, 0.80], preferably 0.65, the value range of Person_C ₂ is [0.00, 0.40], preferably 0.20, the value range of Person_C ₃ is [0.00, 0.25], preferably 0.15, and Person_C ₁ , Person_C ₂ and Person_C ₃ cannot be 0.00 at the same time.

In this application, the behavior and posture information of the controller on duty is collected in real time by linking with the control hall or tower surveillance camera, such as sleeping on duty, leaving the post, etc., and the facial feature information of the controller on duty is collected by placing a high-definition camera at an appropriate position on the control screen, and the controller's land-air conversation information is obtained by real-time collection. Specifically, the collected behavior and posture information of the controller is analyzed and processed, and the dynamic data of the controller's body and posture are analyzed to obtain the current controller fatigue level value F_pose based on the behavior and posture; the collected facial information of the controller is analyzed and processed, and the controller's yawning, blinking, eyelid closure, etc. are analyzed to obtain the current controller fatigue level value F_face based on facial features; the collected land-air conversation information of the controller is analyzed and processed, the controller's voice is recognized, and the controller's semantics and reaction time are recognized and analyzed to obtain the current controller fatigue level value F_voice based on land-air conversation. And F_pose, F_face, and F_voice can all be implemented by any method in the existing technology, which will not be repeated here.

S320, calculate the warning value Person_AL of the controller Person = Person_IoF-Person_ALP, and issue a warning based on the warning value Person_AL. Specifically, when Person_IoF≥Person_ALP, the system issues a warning, wherein the controller's individual fatigue warning threshold Person_ALP is obtained based on the warning threshold determination parameters set in the future time zone during the duty period of the controller Person on duty, and the warning threshold determination parameters at least include the predicted flight volume related to the duty post, the predicted weather conditions, the predicted special situation, and the predicted attributes of the controller Person itself.

In one embodiment of the present invention, a method for obtaining an individual fatigue warning threshold Person_ALP of a controller is also disclosed. As shown in FIG4 , the method includes:

S431, obtain the warning threshold determination parameters set in the future time zone in the duty time period of the controller Person on duty, the warning threshold determination parameters at least include the predicted flight volume, predicted weather conditions, predicted special conditions, predicted controller Person's own attributes, and the duration of the set future time zone has a value range of [3 minutes, 20 minutes], preferably 5 minutes. The controller's own attributes may include factors that have an impact on the controller's fatigue, such as fatigue status, circadian rhythm type, recent sleep status, previous duty situation, gender, age, length of service, etc.

S432, obtaining the duty post fatigue basic requirement value Person_F _b of the controller according to the warning threshold determination parameter. Specifically, in this embodiment, the duty post fatigue basic requirement value Person_F _b is a function of the warning threshold determination parameter of the controller, for example, obtained by summing up different weight values assigned to the predicted flight volume, predicted weather conditions, and predicted attributes of the controller Person.

S433, according to the basic requirement value of fatigue degree of the duty post Person_F _b, the individual fatigue warning threshold value of the controller Person_ALP = (Person_A1 + Person_A2 × Person_A3) × Person_F _b is obtained, where Person_A1 is the human circadian rhythm coefficient of the controller during the duty time, that is, a function of time, for example, the coefficient is high during the night shift and low during the day shift; Person_A2 is the post coefficient, which is used to indicate the need for fatigue degree of the controller's post, for example, when the controller is in the control seat, the coefficient is small, when the coordination seat, the coefficient is large, the busy sector coefficient is small, and the general sector coefficient is large; Person_A3 is the post task requirement coefficient, which is used to indicate the need for fatigue degree of the controller for the expected flight plan traffic flow situation, for example, when the traffic volume is large, the controller is required to be high, and the coefficient is small at this time, otherwise the coefficient is large. Further, Person_ALP is continuously optimized according to the false alarm rate and missed alarm rate of the warning.

By collecting real-time data of on-duty personnel and obtaining judgment thresholds based on parameters of the duty post within a specified time in the future, the working status of the controller on the post can be truly and accurately calculated, the real-time status of the controller can be monitored, and it can be judged whether the controller's current fatigue status can meet the safety fatigue requirements within a specified time in the future. When it is judged that the controller cannot meet the safety fatigue requirements, an alarm will be issued in time and personnel will be replaced, which can effectively avoid safety risks caused by controller fatigue while on duty.

Based on the above content, it can be known that by obtaining the fatigue warning parameters of all control units within the preset time period, the fatigue coefficient of the entire control industry or even each control unit is obtained, and based on the fatigue coefficient, fatigue monitoring and warning are carried out for the entire control industry and the control units in the group with higher fatigue coefficients. In addition, by comparing the fatigue level of the team obtained by the pre-job fatigue test with the team warning threshold obtained by predicting the job fatigue requirements of the job position that the team will be on duty, and by comparing the personal fatigue level of the controller obtained by the fatigue monitoring during the job and the individual fatigue warning threshold of the controller obtained by predicting the job fatigue requirements of the controller in the future, it is possible to detect whether the control personnel can cope with the work requirements of the job position to be on duty before the job, and even after the controller takes office, it is possible to judge whether the personnel on the job will have fatigue risks in the next 3-20 minutes, that is, to predict from multiple angles whether the controller can cope with future aviation safety needs. It realizes fatigue detection, prediction of future job fatigue demand and fatigue warning for the industry as a whole, independent control units, teams and individual controllers at the strategic stage at the organizational level of control units, the pre-tactical (pre-job) stage at the team level, and the tactical (on-duty) stage at the individual level. This is conducive to fatigue risk management for different needs from multiple angles and ensure aviation safety.

FIG2 is a controller fatigue warning device 1 provided in an embodiment of the present application, the device 1 comprising:

The first data acquisition module 11 is used to obtain fatigue warning parameters Air_P=(Air_P ₁ , Air_P ₂ , Air_P ₃ , ..., _{Air_PN} ) of all control units within a first preset time period when the current demand meets the first fatigue determination demand, where N is the number of all control units, wherein the fatigue warning parameters Air_P _i of the i-th control unit at least include unit type Air_type _i , altitude Air_height _i , control mode Air_manage _i , control area Air_area _i , total number of on-duty controllers Air_num _i , total number of on-duty controllers at night Air_num_night _i , total number of on-duty controllers during peak traffic period Air_num_high _i , duty time of controllers Air_time _i , duty time of controllers at work Air_work_time _i , duty time of controllers at work at night Air_night_time _i , duty time of controllers at work during peak traffic period Air_high_time _i , and non-duty time of controllers Air_rwork_time _i . The duty time of the controller for other operations Air_other_time _i , 1≤i≤N.

In the present application, the first fatigue determination requirement may be, for example, a fatigue determination requirement at the strategic stage of the control unit's organizational level, which is helpful in determining whether there are safety risks from the overall perspective of the control industry. The first preset time period may be a historical time period or a future time period, and the length of the time period may be set according to specific needs, for example, it may be a week, a month, and so on. When the first preset time period is a future time period, relevant fatigue warning parameters may be obtained based on the control unit's flight plan and the controller's working hours, and the safety risk issues of the control industry and/or control unit in the future time period may be determined based on the fatigue parameters, so as to make employee changes in advance.

In the present application, the type of control unit Air_type _i can be a plurality of pre-set types, for example, it can be an area control type control unit, an approach control type control unit, a tower control and a civil control unit with an annual throughput of 500,000 to 2 million, etc. The total number of control unit types can be divided according to actual conditions. The altitude Air_height _i of the control unit can be obtained through a variety of channels, such as obtaining by querying geographic data, etc. The control method Air_manage _i can be divided into four major types: procedural control, radar control, radar surveillance control and others. The control area Air_area _i is measured according to actual conditions. The total number of on-duty controllers Air_num _i of the i-th control unit, the total number of controllers on duty at night Air_num_night _i of the i-th control unit, and the total number of controllers on duty during peak traffic periods Air_num_high _i of the i-th control unit can be obtained by collecting the unique identifier of the controller and the actual working time of the controller, and on the other hand, the various parameters can be obtained according to the flight plan arrangement. The same method can also be used to obtain the controller's duty time Air_time _i of the i-th control unit, the controller's post duty time Air_work_time _i of the i-th control unit, the controller's night post duty time Air_night_time _i of the i-th control unit, the controller's post duty time during peak traffic period Air_high_time _i of the i-th control unit, the controller's non-post duty time Air_rwork_time _i , and the controller's duty time for other operations of the control unit Air_other_time _i . In addition, in the present application, when the aircraft hourly support flight volume of the i-th control unit is ≥ the preset traffic threshold Air_FT _i , the i-th control unit is determined to be in the peak traffic period. The preset traffic threshold Air_FT _i is set according to the actual flight data or flight plan of the control unit.

The first data processing module 12 is used to obtain the overall fatigue coefficient IoF of all control units and/or the unit fatigue coefficient Air_IoF of all control units based on the fatigue warning parameter Air_P = [Air_IoF ₁ , Air_IoF ₂ , Air_IoF ₃ , ..., Air_IoF _N ], wherein: Unit fatigue coefficient of the i-th control unit _{Air_Fi1} = 1, which is the weighted coefficient of the air traffic controller's duty time Air_work_time _i of the ith control unit. When the ith control unit is a civil airport, _{Air_Fi2} = 0, otherwise, _{Air_Fi2} = Air_fm _i × Air_other_time _i / Air_work_time _i , where Air_fm _i is the weighted coefficient of other working time of the ith control unit, and its value range is [1.00, 4.00], preferably 1.80; _{Air_Fi3} is a function of the altitude Air_height _i of the ith control unit, and its value range is [0.00, 8.00], preferably 4.21; in the present application, when the hourly aircraft support flights of the ith control unit ≥ the preset flow threshold Air_FT _i , the ith control unit is determined to be in the peak flow period, and when the ith control unit is in the non-peak flow period, _{Air_Fi4} = 0; otherwise, _{Air_Fi4} =Air_num_high _i ×Air_high_time _i /Air_work_time _i , Air_F _i5 =Air_num_night _i ×Air_night_time _i /Air_work_time _i , which is a function of the night duty time Air_night_time _i of the i-th control unit; Air_F _i6 is a function of the control mode Air_manage _i of the i-th control unit, and its value range is [0.00,6.00]; Air_F _i7 is a function of the control area Air_area _i of the i-th control unit, and its value range is [0.00,7.00]; Air_F _i8 is the weighted coefficient of the non-duty time of the i-th control unit, and its value range is [0.20,1.00], and is preferably 0.64.

The first fatigue warning module 13 is used to perform fatigue warning on the entire regulated industry based on the overall fatigue coefficient IoF, and/or to perform fatigue warning on regulated units in a group with a higher fatigue coefficient among different regulated unit types based on the unit fatigue coefficient Air_IoF and the fatigue warning parameter Air_P. Specifically, in one embodiment, based on the unit fatigue coefficient Air_IoF and the fatigue warning parameter Air_P, fatigue warning is performed on regulated units in a group with a higher fatigue coefficient among different regulated unit types as follows: rank different types of regulated units according to the size of their fatigue coefficients, and then issue warning prompts to regulated units with higher numerical rankings among different types of regulated units, wherein the regulated unit type data can be obtained from the fatigue warning parameter Air_P, and the corresponding fatigue-related data is obtained from the fatigue coefficient Air_IoF. For example, there are 4 first-category control units, and the corresponding fatigue coefficients are 200, 100, 150, and 300 respectively. Then, they are sorted by size to 300, 200, 150, and 100. When it is necessary to issue an early warning for the first two, it is necessary to issue an early warning for the two control units with fatigue coefficients of 300 and 200 respectively. Among them, the fact that there are 4 first-category control units is an example for the convenience of explaining the early warning method, and does not represent the actual number of control units. The actual data of the first-category control units shall be based on the actual number obtained. In this application, an early warning alarm benchmark can be set. When the fatigue coefficient of a control unit is greater than or equal to the early warning alarm benchmark, an early warning is issued to the control unit, otherwise no early warning is required. The early warning alarm benchmark can be obtained based on historical experience, or it can be obtained by other methods. The method for obtaining the early warning alarm benchmark described in this application is not subject to specific restrictions. In another embodiment, based on the unit fatigue coefficient Air_IoF and the fatigue warning parameter Air_P, fatigue warning is performed on control units with higher fatigue coefficients in different control unit types as follows: first, all control units are classified based on the fatigue warning parameter Air_P, and then, each type of control unit is sorted according to the size of the fatigue coefficient. Then, each type of control unit after sorting is graded using a clustering method and divided into groups with lower, medium, and higher fatigue coefficients. Finally, in each type of control unit, a warning is given to the control unit that enters a higher value group after clustering. Clustering the fatigue coefficients of each control unit by clustering can better mine the inherent distribution law between data. Specifically, the commonly used K-mean clustering method can be used to cluster the fatigue coefficients, and the clustering method used in this application can be any clustering method in the prior art.

Specifically, fatigue warning is performed for the entire regulated industry based on the overall fatigue coefficient IoF. In one embodiment, the overall fatigue coefficient IoF can be compared with the historical overall fatigue coefficient of the entire industry. When the overall fatigue coefficient IoF is larger than the historical overall fatigue coefficient, a warning is performed. In another embodiment, when the overall fatigue coefficient IoF increases to a certain extent compared with the historical overall fatigue coefficient, a warning is performed. Specifically, the certain extent can be combined with the unsafe events within the first preset time period of the regulated unit to determine whether to issue a warning. In another embodiment, a warning is performed when the overall fatigue coefficient IoF ≥ the overall fatigue coefficient warning threshold Total_ALP, wherein the overall fatigue coefficient warning threshold Total_ALP is obtained by data-driven iteration based on the overall fatigue coefficients of all regulated units and the number of unsafe events of all regulated units obtained in multiple different historical time periods. For example, the overall fatigue coefficient IoF_1 of all control units in the first historical time period is a first value, and the number of unsafe events is 0; the overall fatigue coefficient IoF_2 in the second historical time period is a second value, and the number of unsafe events is 0. Although the second value IoF_2 is greater than the first value IoF_1, because the number of unsafe events in the second historical time period is 0, IoF_2 is not used as the overall fatigue coefficient warning threshold Total_ALP. If the overall fatigue coefficient IoF_3 in the third historical time period is a third value, and the number of unsafe events is 2, because the third value IoF_3 is greater than the second value IoF_2, and more unsafe events occurred in the third historical time period, IoF_3 can be used as the overall fatigue coefficient warning threshold Total_ALP. By analogy, the overall fatigue coefficient and unsafe events in each historical time period can be comprehensively considered to finally obtain the overall fatigue coefficient warning threshold Total_ALP. Furthermore, the duration of the first historical time period, the second historical time period, and the third historical time period may be one month, one week, etc., and appropriate selection of the statistical duration may enable the overall fatigue coefficient warning threshold value Total_ALP to be quickly and effectively obtained.

The second data acquisition module 21 is used to obtain the test fatigue value Set_PreIoF=[Set_PreIoF ₁ , Set_PreIoF ₂ ,..., Set_PreIoF _M ] of the team controller of the i-th control unit in the second preset time period before taking up the post when the current demand meets the second fatigue determination demand, where M is the total number of the team controllers. The length of the second preset time period can be set by itself, such as 10 minutes, 5 minutes, 3 minutes, etc. Preferably, in the present application, the length of the second preset time period is 5 minutes, which can ensure the authenticity of the controller's detection during the test on the one hand, and avoid pseudo fatigue of the controller due to the long test time on the other hand, thereby affecting the authenticity of the test. Specifically, in the present application, the test fatigue value Set_PreIoF _k of the k-th controller is obtained based on at least one of the video test data, alertness test data and subjective scale data of the k-th controller, 1≤k≤M. In one embodiment, Set_PreIoF _k =A _k1 ×W _k +A _k2 ×U _k +A _k3 ×V _k , where A _k1 , A _k2 , and A _k3 are weight coefficients, the value range of A _k1 is [0, 3], preferably 2.4, the value range of A _k2 is [4, 8], preferably 6, and the value range of A _k3 is [2, 5], preferably 3. Wherein, the alertness test data V _k = (Vo+Ve)/Vt, where Vo is the number of ignored target letters, Ve is the number of selected incorrect letters, and Vt is the total number of target letters; the alertness test data U _k = t/t _T , t is the correct reaction time during the alertness test, and t _T = 500 ms, which is the reaction time threshold. Video test data Wherein FPs represents the frame rate per second, tv represents the length of the video, W(l) represents the coverage of the eyeball of the person being tested at the lth frame, and sign{} represents the indicator function, where sign{true}＝1,sign{false}＝0, 1≤l≤p, p＝FPs×tv, which is the number of frames of the video.

The second data processing module 22 is used to calculate the warning value Set_AL=[Set_AL ₁ , Set_AL ₂ , ..., Set_AL _M ] of the team controller, wherein Set_AL _k =Set_PreIoF _k -Set_ALP _k , Set_ALP _k is the fatigue warning threshold of the kth controller regarding the post to be on duty, and the team fatigue warning threshold Set_ALP=[Set_ALP ₁ , Set_ALP ₂ , ..., Set_ALP _M ] is obtained according to the warning threshold parameters in the specified time interval of the team's upcoming duty time period, and the warning threshold parameters at least include the team's duty time, predicted flight volume, predicted duty duration, predicted weather conditions, predicted special conditions, and predicted attributes of the controller itself.

The second fatigue warning module 23 is used to issue a warning based on the warning value Set_AL. Specifically, when Set_AL _k ≥ Set_ALP _k , that is, Set_AL _k ≥ 0, the system issues a warning to the controller, and the team can promptly arrange other controllers that meet the management requirements to replace them.

In one embodiment of the present invention, a method for obtaining a team fatigue warning threshold Set_ALP is also disclosed, the method comprising:

S331, obtain the warning threshold parameters within the specified time interval in the duty time period of the team, and the warning threshold parameters at least include the duty time of the team, the predicted number of flights, the predicted duty time, the predicted weather conditions, the predicted special situation, and the predicted attributes of the controller itself. The duration of the specified time interval ranges from [20 minutes to 300 minutes], preferably 120 minutes. For example, if the team starts work at 4 pm, then we set the specified time interval to 4 to 6 pm. The duty time can be, for example, morning, evening, etc. The number of flights is the predicted number of flights, which can be predicted based on the flight plan of the control unit. For example, the number of flights in the flight plan within the specified time zone is 3, but temporary landing flights are not excluded. At this time, it is necessary to predict the number of flights of the team in the specified time interval based on the above-mentioned flight plan and temporary factors. Similarly, the duty time is also the duty time predicted by the team, such as 1 hour, 2 hours, etc. The attributes of the controller itself include fatigue status, circadian rhythm type, recent sleep status, previous duty status, gender, age, length of service, etc.

S332, according to the warning threshold parameter, obtain the team basic fatigue value Set_P = [Set_P ₁ , Set_P ₂ , ..., _{Set_PM} ] of the team, wherein Set_P _k represents the basic fatigue value of the kth controller of the team. Specifically, in this embodiment, the basic fatigue value Set_P _k of the kth controller is a function of the warning threshold parameter of the controller, for example, the duty time assigned to the team, the predicted number of flights, the predicted duty time, the predicted weather conditions, the predicted special conditions, and the predicted attributes of the kth controller itself are assigned different weight values and then summed to obtain the basic fatigue value Set_P _k .

S333, based on Set_P, obtain the team fatigue warning threshold Set_ALP = [Set_ALP ₁ , Set_ALP ₂ , ..., Set_ALP _M ] of the team, wherein Set_ALP _k = Set_C × Set_P _k , Set_C is the team post adjustment coefficient, and its value range is [1.0, 5.0], such as the post adjustment of the control seat and the coordination seat, and the post adjustment of different sectors of the control seat. The adjustment coefficient expresses the maximum internal post adjustment amount that the team can allow in order to cope with the job to be on duty. By using the team post adjustment coefficient, the warning standard can be made more in line with the actual situation, and thus the warning can be made more accurate. In addition, it can also satisfy the team's overall grasp of the job to be on duty, which is convenient for the controller to adjust and deploy the work. Further, Set_P is continuously optimized according to the false alarm rate and missed alarm rate of the warning.

By conducting fatigue tests on the team's controllers before taking up their posts, the most realistic fatigue level of the controllers before taking up their posts can be obtained. Since the team's basic fatigue value is obtained based on the team's plans for the posts they will be on duty and the controller's own attributes, it can more objectively reflect the team's actual needs within a specified time in the future. By comparing the two, the team can more accurately and effectively obtain the fatigue status of the entire team's controllers, and make work adjustments accordingly, thereby controlling safety risks caused by controller fatigue.

The third data acquisition module 31 is used to obtain a comprehensive fatigue level value of a controller Person Person_IoF = Person_C ₁ ×F_pose+Person_C ₂ ×F_face+Person_C ₃ ×F_voice of a controller Person when the current demand meets the third fatigue judgment demand, wherein F_pose is the current fatigue level value of the controller based on behavior posture, F_face is the current fatigue level value of the controller based on facial features, F_voice is the current fatigue level value of the controller based on land-air communication, Person_C ₁ is the weight coefficient of F_pose, Person_C ₂ is the weight coefficient of F_face, and Person_C ₃ is the weight coefficient of F_voice. The value range of Person_C ₁ is [0.00, 0.80], preferably 0.65, the value range of Person_C ₂ is [0.00, 0.40], preferably 0.20, the value range of Person_C ₃ is [0.00, 0.25], preferably 0.15, and Person_C ₁ , Person_C ₂ and Person_C ₃ cannot be 0.00 at the same time.

In this application, the behavior and posture information of the controller on duty is collected in real time by linking with the control hall or tower surveillance camera, such as the controller sleeping on duty, leaving the post, etc., and the facial feature information of the controller on duty is collected by placing a high-definition camera at an appropriate position on the control screen, and the controller's land-air conversation information is obtained by real-time collection. Specifically, the collected behavior and posture information of the controller is analyzed and processed, and the dynamic data of the controller's body and posture are analyzed to obtain the current controller fatigue level value F_pose based on the behavior and posture; the collected facial information of the controller is analyzed and processed, and the controller's yawning, blinking, eyelid closure, etc. are analyzed to obtain the current controller fatigue level value F_face based on facial features; the collected land-air conversation information of the controller is analyzed and processed, the controller's voice is recognized, and the controller's semantics and reaction time are recognized and analyzed to obtain the current controller fatigue level value F_voice based on land-air conversation. And F_pose, F_face, and F_voice can all be implemented by any method in the existing technology, which will not be repeated here.

The third data processing module 32 is used to calculate the warning value Person_AL of the controller Person = Person_IoF-Person_ALP, wherein the controller's individual fatigue warning threshold Person_ALP is obtained according to the warning threshold judgment parameters set in the future time area during the duty period of the controller Person on duty, and the warning threshold judgment parameters include at least the predicted flight volume related to the duty post, the predicted weather conditions, the predicted special situation, and the predicted attributes of the controller Person itself.

The third fatigue warning module 33 is used to issue a warning based on the warning value Person_AL. Specifically, when Person_IoF≥Person_ALP, the system issues a warning. Among them, Person_ALP＝(Person_A1+Person_A2×Person_A3)×Person_F _b , Person_A1 is the human circadian rhythm coefficient of the controller during the duty time, that is, it is a function of time. For example, the coefficient is high during the night shift and low during the day shift; Person_A2 is the position coefficient, which is used to indicate the need for fatigue level of the controller's position. For example, when the controller is at the control seat, the coefficient is small, and when the coordination seat is large, the coefficient is large. The busy sector coefficient is small, and the general sector coefficient is large; Person_A3 is the position task requirement coefficient, which is used to indicate the need for controller fatigue level of the expected flight plan traffic flow situation. For example, when the traffic volume is large, the requirements for the controller are high, and the coefficient is small at this time, otherwise the coefficient is large; Person_F _b is the basic requirement value of fatigue level of the duty position, which is obtained according to the flight plan and historical experience.

In one embodiment of the present invention, a method for obtaining an individual fatigue warning threshold Person_ALP of a controller is also disclosed. The method comprises:

S431, obtain the warning threshold determination parameters set in the future time zone in the duty time period of the controller Person on duty, the warning threshold determination parameters at least include the predicted flight volume, predicted weather conditions, predicted special conditions, predicted controller Person's own attributes, and the duration of the set future time zone has a value range of [3 minutes, 20 minutes], preferably 5 minutes. The controller's own attributes may include factors that have an impact on the controller's fatigue, such as fatigue status, circadian rhythm type, recent sleep status, previous duty situation, gender, age, length of service, etc.

S432, obtaining the duty post fatigue basic requirement value Person_F _b of the controller according to the warning threshold determination parameter. Specifically, in this embodiment, the duty post fatigue basic requirement value Person_F _b is a function of the warning threshold determination parameter of the controller, for example, obtained by respectively assigning different weight values to the predicted flight volume, predicted weather conditions, predicted special conditions, and predicted attributes of the controller Person, and then summing them up.

S433, according to the basic requirement value of fatigue degree of the duty post Person_F _b, the individual fatigue warning threshold value of the controller Person_ALP = (Person_A1 + Person_A2 × Person_A3) × Person_F _b is obtained, where Person_A1 is the human circadian rhythm coefficient of the controller during the duty time, that is, a function of time, for example, the coefficient is high during the night shift and low during the day shift; Person_A2 is the post coefficient, which is used to indicate the need for fatigue degree of the controller's post, for example, when the controller is in the control seat, the coefficient is small, when the coordination seat, the coefficient is large, the busy sector coefficient is small, and the general sector coefficient is large; Person_A3 is the post task requirement coefficient, which is used to indicate the need for fatigue degree of the controller for the expected flight plan traffic flow situation, for example, when the traffic volume is large, the controller is required to be high, and the coefficient is small at this time, otherwise the coefficient is large. Further, Person_ALP is continuously optimized according to the false alarm rate and missed alarm rate of the warning.

By collecting real-time data of on-duty personnel and obtaining judgment thresholds based on parameters of the duty post within a specified time in the future, the working status of the controller on the post can be truly and accurately calculated, the real-time status of the controller can be monitored, and it can be judged whether the controller's current fatigue status can meet the safety fatigue requirements within a specified time in the future. When it is judged that the controller cannot meet the safety fatigue requirements, an alarm will be issued in time and personnel will be replaced, which can effectively avoid safety risks caused by controller fatigue while on duty.

Based on the above content, it can be known that by obtaining the fatigue warning parameters of all control units within the preset time period, the fatigue coefficient of the entire control industry or even each control unit is obtained, and based on the fatigue coefficient, fatigue monitoring and warning are carried out for the entire control industry and the control units in the group with higher fatigue coefficients. In addition, by comparing the fatigue level of the team obtained by the pre-job fatigue test with the team warning threshold obtained by predicting the job fatigue requirements of the job that the team will be on duty, and by comparing the personal fatigue level of the controller obtained by the fatigue monitoring during the job and the individual fatigue warning threshold of the controller obtained by predicting the job fatigue requirements of the controller in the future, it is possible to detect whether the controller can cope with the work requirements of the job to be on duty before the job, and even after the controller takes up the job, it is possible to judge whether the personnel on the job will have abnormal fatigue in the next 3-20 minutes, that is, to predict from multiple angles whether the controller can cope with future aviation safety needs. It realizes fatigue detection, prediction of future job fatigue demand and fatigue warning for the industry as a whole, independent control units, teams and individual controllers at the strategic stage at the organizational level of control units, the pre-tactical (pre-job) stage at the team level, and the tactical (on-duty) stage at the individual level. This is conducive to fatigue risk management for different needs from multiple angles and ensure aviation safety.

Another embodiment of the present application discloses a flow chart of a controller fatigue warning method, as shown in FIG5 . The above-mentioned related contents are also applicable to this embodiment and will not be repeated here. The method comprises the following steps:

S100, determine whether the current demand meets the first fatigue determination requirement, the second fatigue determination requirement or the third fatigue determination requirement; when the current demand meets the first fatigue determination requirement, execute step S210, when the current demand meets the second fatigue determination requirement, execute step S310, otherwise execute step S410.

S210, obtaining fatigue warning parameters Air_P=(Air_P ₁ , Air_P ₂ , Air_P ₃ , ..., _{Air_PN} ) of all control units in a first preset time period, where N is the number of all control units, wherein the fatigue warning parameters Air_P _i of the i-th control unit at least include unit type Air_type _i , altitude Air_height _i , control mode Air_manage _i , control area Air_area _i , total number of on-duty controllers Air_num _i , total number of on-duty controllers at night Air_num_night _i , total number of on-duty controllers during peak traffic period Air_num_high _i , duty time of controllers Air_time _i , duty time of controllers at work Air_work_time _i , duty time of controllers at work at night Air_night_time _i , duty time of controllers at work during peak traffic period Air_high_time _i , and non-duty time of controllers Air_rwork_time _i , the duty time of the controller for other operations Air_other_time _i , 1≤i≤N.

S220, based on the fatigue warning parameter Air_P, obtaining the overall fatigue coefficient IoF of all control units and/or the unit fatigue coefficient Air_IoF of all control units = [Air_IoF ₁ , Air_IoF ₂ , Air_IoF ₃ , ..., Air_IoF _N ], wherein: Unit fatigue coefficient of the i-th control unit _{Air_Fi1} = 1, which is the weighted coefficient of the air traffic controller's duty time Air_work_time _i of the ith control unit. When the ith control unit is a civil airport, _{Air_Fi2} = 0. Otherwise, _{Air_Fi2} = Air_fm _i × Air_other_time _i / Air_work_time _i . Air_fm _i is the weighted coefficient of other working time of the ith control unit. _{Air_Fi3} is a function of the altitude Air_height _i of the ith control unit. When the hourly aircraft support flight volume of the ith control unit is ≥ the preset flow threshold Air_FT _i , the ith control unit is determined to be in the peak flow period. When the ith control unit is in the non-peak flow period, _{Air_Fi4} = 0. Otherwise, _{Air_Fi4} = Air_num_high _i × Air_high_time _i / Air_work_time _i , _{Air_Fi5} = Air_num_night _i × Air_night_time _i / Air_work_time _i , is a function of the night duty time Air_night_time _i of the ith control unit; Air_F _i6 is a function of the control method Air_manage _i of the ith control unit; Air_F _i7 is a function of the control area Air_area _i of the ith control unit; Air_F _i8 is the weighted coefficient of the non-duty time of the ith control unit.

S230, obtaining the overall fatigue coefficient warning threshold value Total_ALP of all control units, and classifying all control units based on the fatigue warning parameter Air_P, and then sorting each type of control units according to the size of the fatigue coefficient according to the unit fatigue coefficient Air_IoF, and then, using a clustering method to grade each type of control units after sorting, and divide them into groups with low, medium and high fatigue coefficients, wherein the overall fatigue coefficient warning threshold value Total_ALP is obtained by data-driven iteration based on the overall fatigue coefficients of all control units obtained in multiple different historical time periods and the number of unsafe events of all control units.

S240, obtaining fatigue warning values Total_AL of all control units = the overall fatigue coefficient IoF - the overall fatigue coefficient warning threshold Total_ALP.

S250, based on the fatigue warning value Total_AL of all the control units, a fatigue warning is issued for the control industry as a whole, that is, a warning is issued when the fatigue warning value Total_AL of all the control units is ≥0, and/or, in each type of control unit, a warning is issued for the control units that enter the group with higher values after clustering.

S310, obtaining the test fatigue parameters of the team controller of the i-th control unit within the second preset time period before taking up the post, wherein the test fatigue parameters are obtained through at least one of the video test, the alertness test and the subjective scale test; and obtaining the warning threshold parameters within the specified time interval in the time period when the team will be on duty, wherein the warning threshold parameters at least include the team's duty time, the predicted flight volume, the predicted duty duration, the predicted weather conditions, the predicted special situation, and the predicted attributes of the controller itself.

S320, obtaining the test fatigue value Set_PreIoF=[Set_PreIoF ₁ , Set_PreIoF ₂ , ..., Set_PreIoF _M ] of the team controller according to the test fatigue parameter, where M is the total number of team controllers; obtaining the team basic fatigue value Set_P=[Set_P ₁ , Set_P ₂ , ..., Set_P _M ] of the team according to the warning threshold parameter, where Set_P _k represents the basic fatigue value of the kth controller of the team.

S330, based on the team basic fatigue value Set_P, obtain the team fatigue warning threshold Set_ALP = [Set_ALP ₁ , Set_ALP ₂ , ..., Set_ALP _M ] of the team, wherein the fatigue warning threshold Set_ALP _k = Set_C×Set_P _k of the kth controller regarding the post to be on duty, and Set_C is the team post adjustment coefficient.

S340, calculating the warning value Set_AL = [Set_AL ₁ , Set_AL ₂ , ..., Set_AL _M ] of the team controller according to the team fatigue warning threshold Set_ALP and the test fatigue value Set_PreIoF, wherein Set_AL _k = Set_PreIoF _k - Set_ALP _k .

S350: issuing an early warning based on the early warning value Set_AL, that is, issuing an early warning when Set_AL _k ≥0.

S410, obtaining fatigue level parameters F_pose, F_face, and F_voice of a controller Person, wherein F_pose is the current controller fatigue level value based on behavioral posture, F_face is the current controller fatigue level value based on facial features, and F_voice is the current controller fatigue level value based on air-ground communication, as well as obtaining early warning threshold determination parameters set within a future time zone within the duty period of the controller Person on duty, wherein the early warning threshold determination parameters include at least predicted flight volume, predicted weather conditions, predicted special conditions, and predicted attributes of the controller Person itself.

S420, obtaining the controller comprehensive fatigue level value Person_IoF of the controller Person: _{Person_C1} ×F_pose+ _{Person_C2} ×F_face+ _{Person_C3} ×F_voice; _{Person_C1} is the weight coefficient of F_pose, _{Person_C2} is the weight coefficient of F_face, and _{Person_C3} is the weight coefficient of F_voice; obtaining the duty post fatigue level basic requirement value _{Person_Fb} of the controller according to the warning threshold determination parameter.

S430, according to the basic requirement value Person_F _b of fatigue degree of the duty post, obtain the individual fatigue warning threshold value Person_ALP of the controller = (Person_A1 + Person_A2 × Person_A3) × Person_F _b , where Person_A1 is the human circadian rhythm coefficient of the individual controller during the duty time, that is, it is a function of time, for example, the coefficient is high during the night shift and low during the day shift; Person_A2 is the post coefficient, which is used to indicate the need for fatigue degree of the controller's post, for example, when the controller is in the control seat, the coefficient is small, when the coordination seat, the coefficient is large, the busy sector coefficient is small, and the general sector coefficient is large; Person_A3 is the post task requirement coefficient, which is used to indicate the need for fatigue degree of the controller due to the expected flight plan traffic flow situation, for example, when the traffic flow is large, the requirement for the controller is high, and the coefficient is small at this time, otherwise the coefficient is large.

S440, calculating the warning value Person_AL of the controller Person = Person_IoF - Person_ALP.

S450: Issue an early warning based on the early warning value Person_AL. That is, when Person_AL≥0, issue an early warning.

A controller fatigue warning system includes a processor and a non-transitory storage medium for storing at least one instruction or at least one program. The at least one instruction or the at least one program is loaded and executed by the processor to implement the fatigue warning method provided in the above embodiment.

An embodiment of the present application also provides a non-transitory computer-readable storage medium, which can be set in an electronic device to store at least one instruction or at least one program related to implementing a method in a method embodiment. The at least one instruction or the at least one program is loaded and executed by the processor to implement the method provided in the above embodiment.

An embodiment of the present application also provides an electronic device, including a processor and the aforementioned non-transitory computer-readable storage medium.

An embodiment of the present application further provides a computer program product, which includes a program code. When the program product is run on an electronic device, the program code is used to enable the electronic device to execute the steps of the method according to various exemplary embodiments of the present application described above in this specification.

Although some specific embodiments of the present application have been described in detail by way of example, it will be appreciated by those skilled in the art that the above examples are only for illustration and not for limiting the scope of the present application. It will also be appreciated by those skilled in the art that various modifications may be made to the embodiments without departing from the scope and spirit of the present application. The scope of the present application is defined by the appended claims.

Claims (13)

1. A controller fatigue early warning method, characterized in that the method comprises the following steps: S100, determining whether the current demand meets the first fatigue determination requirement, the second fatigue determination requirement or the third fatigue determination requirement; when the current demand meets the first fatigue determination requirement, executing step S210, when the current demand meets the second fatigue determination requirement, executing step S310, otherwise executing step S410; S210, obtaining fatigue warning parameters Air_P=(Air_P ₁ , Air_P ₂ , Air_P ₃ , ..., _{Air_PN} ) of all control units in a first preset time period, where N is the number of all control units, wherein the fatigue warning parameters Air_P _i of the i-th control unit at least include unit type Air_type _i , altitude Air_height _i , control mode Air_manage _i , control area Air_area _i , total number of on-duty controllers Air_num _i , total number of on-duty controllers at night Air_num_night _i , total number of on-duty controllers during peak traffic period Air_num_high _i , duty time of controllers Air_time _i , duty time of controllers at work Air_work_time _i , duty time of controllers at work at night Air_night_time _i , duty time of controllers at work during peak traffic period Air_high_time _i , and non-duty time of controllers Air_rwork_time _i , the duty time of the controller for other operations Air_other_time _i , 1≤i≤N; S220, based on the fatigue warning parameter Air_P, obtaining the overall fatigue coefficient IoF of all control units and/or the unit fatigue coefficient Air_IoF of all control units = [Air_IoF ₁ , Air_IoF ₂ , Air_IoF ₃ , ..., Air_IoF _N ], wherein: Unit fatigue coefficient of the i-th control unit is the weighted coefficient of the controller's post duty time Air_work_time _i of the ith control unit. When the ith control unit is a civil airport, Air_F _i2 = 0. Otherwise, Air_F _i2 = Air_fm _i × Air_other_time _i / Air_work_time _i . Air_fm _i is the weighted coefficient of other working time of the ith control unit. Air_F _i3 is a function of the altitude Air_height _i of the ith control unit. When the hourly aircraft support flight volume of the ith control unit is ≥ the preset flow threshold Air_FT _i , the ith control unit is determined to be in the peak flow period. When the ith control unit is in the non-peak flow period, Air_F _i4 = 0. Otherwise, Air_F _i4 = Air_num_high _i × Air_high_time _i / Air_work_time _i . Air_F _i5 = Air_num_night _i × Air_night_time _i / Air_work_time _i , is a function of the night duty time Air_night_time _i of the ith control unit, Air_F _i6 is a function of the control method Air_manage _i of the ith control unit, Air_F _i7 is a function of the control area Air_area _i of the ith control unit, and Air_F _i8 is a weighted coefficient of the non-duty time of the ith control unit; S230, performing fatigue warning for the entire regulated industry according to the overall fatigue coefficient IoF, and/or performing fatigue warning for regulated units of a group with a higher fatigue coefficient among different regulated unit types based on the unit fatigue coefficient Air_IoF and the fatigue warning parameter Air_P; S310, obtaining a test fatigue value Set_PreIoF＝[Set_PreIoF ₁ , Set_PreIoF ₂ , ..., Set_PreIoF _M ] of a team controller of the i-th control unit in a second preset time period before taking up the post, where M is the total number of the team controllers, wherein the test fatigue value Set_PreIoF _k of the k-th controller is obtained according to at least one of the video test data, the alertness test data and the subjective scale data of the k-th controller, and 1≤k≤M; S320, calculating the warning value Set_AL=[Set_AL ₁ , Set_AL ₂ , ..., Set_AL _M ] of the team controller, and issuing a warning based on the warning value Set_AL, wherein Set_AL _k =Set_PreIoF _k -Set_ALP _k , Set_ALP _k is the fatigue warning threshold of the k-th controller regarding the post to be on duty, and the team fatigue warning threshold Set_ALP=[Set_ALP ₁ , Set_ALP ₂ , ..., Set_ALP _M ] is obtained according to the warning threshold parameters in the specified time interval of the team's upcoming duty time period, and the warning threshold parameters at least include the team's duty time, predicted flight volume, predicted duty duration, predicted weather conditions, predicted special conditions, and predicted attributes of the controller itself; S410, obtaining a controller comprehensive fatigue value Person_IoF of a controller Person, _{Person_C1} ×F_pose+ _{Person_C2} ×F_face+ _{Person_C3} ×F_voice, wherein F_pose is a current controller fatigue value based on behavior and posture, F_face is a current controller fatigue value based on facial features, F_voice is a current controller fatigue value based on ground-air communication, _{Person_C1} is a weight coefficient of F_pose, _{Person_C2} is a weight coefficient of F_face, and _{Person_C3} is a weight coefficient of F_voice; S420, calculating the warning value Person_AL=Person_IoF-Person_ALP of the controller Person, and issuing a warning based on the warning value Person_AL, wherein the controller's individual fatigue warning threshold Person_ALP is obtained according to the warning threshold determination parameters set in the future time zone during the duty period of the controller Person on duty, and the warning threshold determination parameters at least include the predicted flight volume, predicted weather conditions, predicted special conditions, and predicted attributes of the controller Person itself related to the duty post; Based on the unit fatigue coefficient Air_IoF and the fatigue warning parameter Air_P, fatigue warning is performed on control units with higher fatigue coefficients in different control unit types as follows: first, all control units are classified based on the fatigue warning parameter Air_P; second, each type of control unit is sorted according to the size of the fatigue coefficient; then, each type of control unit after sorting is graded using a clustering method to be divided into groups with lower, medium, and higher fatigue coefficients; finally, in each type of control unit, a warning prompt is given to the control unit that enters a group with a higher value after clustering; Performing fatigue warning on the entire regulated industry according to the overall fatigue coefficient IoF includes: issuing a warning when the overall fatigue coefficient IoF ≥ the overall fatigue coefficient warning threshold Total_ALP, wherein the overall fatigue coefficient warning threshold Total_ALP is obtained by data-driven iteration based on the overall fatigue coefficients of all regulated units and the number of unsafe events of all regulated units obtained in multiple different historical time periods; The acquisition of the team fatigue warning threshold Set_ALP includes the following steps: S331, obtaining the warning threshold parameters within the specified time interval of the team's upcoming duty time, wherein the warning threshold parameters at least include the team's duty time, predicted flight volume, predicted duty duration, predicted special conditions, predicted weather conditions, and predicted attributes of the controller itself, and the duration of the specified time interval has a value range of [20 minutes, 300 minutes]; S332, obtaining a team basic fatigue value Set_P=[Set_P ₁ , Set_P ₂ , ..., Set_P _M ] of the team according to the warning threshold parameter, wherein Set_P _k represents the basic fatigue value of the kth controller of the team; S333, based on Set_P, obtaining the team fatigue warning threshold Set_ALP = [Set_ALP ₁ , Set_ALP ₂ , ..., Set_ALP _M ] of the team, wherein Set_ALP _k = Set_C × Set_P _k , and Set_C is the team position adjustment coefficient; The acquisition of the controller's individual fatigue warning threshold Person_ALP includes the following steps: S431, obtaining the warning threshold determination parameters set in the future time zone in the duty time period of the controller Person on duty, wherein the warning threshold determination parameters at least include the predicted flight volume, the predicted weather conditions, the predicted special conditions, and the predicted attributes of the controller Person himself, and the duration of the set future time zone has a value range of [3 minutes, 20 minutes]; S432, obtaining the basic required value of fatigue level of the controller's duty post Person_F _b according to the warning threshold determination parameter; S433, according to the basic requirement value of fatigue degree of the duty post Person_F _b, obtain the individual fatigue warning threshold value of the controller Person_ALP = (Person_A1 + Person_A2 × Person_A3) × Person_F _b , where Person_A1 is the human circadian rhythm coefficient of the controller during the duty time, Person_A2 is the post coefficient, which is used to indicate the requirement of fatigue degree of the controller's post; Person_A3 is the post task requirement coefficient, which is used to indicate the requirement of fatigue degree of the controller due to the expected traffic flow situation of the flight plan. 2. The method according to claim 1 is characterized in that the value range of Air_F _i3 is [0.00, 8.00]; the value range of Air_F _i6 is [0.00, 6.00]; the value range of Air_F _i7 is [0.00, 7.00], and the value range of Air_F _i8 is [0.20, 1.00]. 3. A method for obtaining a team fatigue warning threshold Set_ALP based on the method of claim 1, characterized in that it comprises the following steps: S331, obtaining the warning threshold parameters within the specified time interval of the team's upcoming duty period, wherein the warning threshold parameters at least include the team's duty time, predicted flight volume, predicted duty duration, predicted weather conditions, predicted special conditions, and predicted attributes of the controller itself, and the duration of the specified time interval has a value range of [20 minutes, 300 minutes]; S332, obtaining a team basic fatigue value Set_P=[Set_P ₁ , Set_P ₂ , ..., Set_P _M ] of the team according to the warning threshold parameter, wherein Set_P _k represents the basic fatigue value of the kth controller of the team; S333, based on Set_P, obtain the team fatigue warning threshold Set_ALP = [Set_ALP ₁ , Set_ALP ₂ , ..., Set_ALP _M ] of the team, wherein Set_ALP _k = Set_C × Set_P _k , and Set_C is the team position adjustment coefficient. 4. A method for obtaining the controller's individual fatigue warning threshold Person_ALP based on the method of claim 1, characterized in that it comprises the following steps: S431, obtaining the warning threshold determination parameters set in the future time zone in the duty time period of the controller Person on duty, wherein the warning threshold determination parameters at least include the predicted flight volume, the predicted weather conditions, the predicted special conditions, and the predicted attributes of the controller Person himself, and the duration of the set future time zone has a value range of [3 minutes, 20 minutes]; S432, obtaining the basic required value of fatigue level of the controller's duty post Person_F _b according to the warning threshold determination parameter; S433, according to the basic requirement value of fatigue degree of the duty post Person_F _b, obtain the individual fatigue warning threshold value of the controller Person_ALP = (Person_A1 + Person_A2 × Person_A3) × Person_F _b , where Person_A1 is the human circadian rhythm coefficient of the controller during the duty time, Person_A2 is the post coefficient, which is used to indicate the requirement of fatigue degree of the controller's post; Person_A3 is the post task requirement coefficient, which is used to indicate the requirement of fatigue degree of the controller due to the expected traffic flow situation of the flight plan. 5. A method for obtaining the overall fatigue coefficient warning threshold value Total_ALP based on the method of claim 1, characterized in that it comprises the following steps: S231, obtaining the overall fatigue coefficient and the number of unsafe events of all the control units in multiple different historical time periods; S232: Perform data-driven iteration using the overall fatigue coefficients and the number of unsafe events in the multiple different historical time periods to obtain the overall fatigue coefficient warning threshold Total_ALP. 6. A method for obtaining the overall fatigue coefficient of all control units and the unit fatigue coefficient of each unit based on the method of claim 1, characterized in that the method comprises the following steps: S210, obtaining fatigue warning parameters Air_P=(Air_P ₁ , Air_P ₂ , Air_P ₃ , ..., _{Air_PN} ) of all control units in a first preset time period, where N is the number of all control units, wherein the fatigue warning parameters Air_P _i of the i-th control unit at least include unit type Air_type _i , altitude Air_height _i , control mode Air_manage _i , control area Air_area _i , total number of on-duty controllers Air_num _i , total number of on-duty controllers at night Air_num_night _i , total number of on-duty controllers during peak traffic period Air_num_high _i , duty time of controllers Air_time _i , duty time of controllers at work Air_work_time _i , duty time of controllers at work at night Air_night_time _i , duty time of controllers at work during peak traffic period Air_high_time _i , and non-duty time of controllers Air_rwork_time _i , the duty time of the controller for other operations Air_other_time _i , 1≤i≤N; S220, based on the fatigue warning parameter Air_P, obtaining the overall fatigue coefficient IoF of all control units and/or the unit fatigue coefficient Air_IoF of all control units = [Air_IoF ₁ , Air_IoF ₂ , Air_IoF ₃ , ..., Air_IoF _N ], wherein: Unit fatigue coefficient of the i-th control unit is the weighted coefficient of the controller's post duty time Air_work_time _i of the ith control unit. When the ith control unit is a civil airport, Air_F _i2 = 0. Otherwise, Air_F _i2 = Air_fm _i × Air_other_time _i / Air_work_time _i . Air_fm _i is the weighted coefficient of other working time of the ith control unit. Air_F _i3 is a function of the altitude Air_height _i of the ith control unit. When the hourly aircraft support flight volume of the ith control unit is ≥ the preset flow threshold Air_FT _i , the ith control unit is determined to be in the peak flow period. When the ith control unit is in the non-peak flow period, Air_F _i4 = 0. Otherwise, Air_F _i4 = Air_num_high _i × Air_high_time _i / Air_work_time _i . Air_F _i5 = Air_num_night _i × Air_night_time _i / Air_work_time _i , is a function of the night duty time Air_night_time _i of the ith control unit, Air_F _i6 is a function of the control method Air_manage _i of the ith control unit, Air_F _i7 is a function of the control area Air_area _i of the ith control unit, and Air_F _i8 is the weighted coefficient of the non-duty time of the ith control unit. 7. A controller fatigue warning device, characterized in that the device comprises: a first data acquisition module, for acquiring fatigue warning parameters Air_P=(Air_P ₁ , Air_P ₂ , Air_P ₃ , ..., _{Air_PN} ) of all control units within a first preset time period when a current demand meets a first fatigue determination demand, wherein N is the number of all control units, wherein the fatigue warning parameters Air_P _i of the i-th control unit at least include unit type Air_type _i , altitude Air_height _i , control mode Air_manage _i , control area Air_area _i , total number of on-duty controllers Air_num _i , total number of on-duty controllers at night Air_num_night _i , total number of on-duty controllers during peak traffic period Air_num_high _i , duty time of controllers Air_time _i , duty time of controllers at post Air_work_time _i , duty time of controllers at post at night Air_night_time _i , duty time of controllers at post during peak traffic period Air_high_time _i , and non-duty time of controllers Air_rwork_time _i , the duty time of the controller for other operations Air_other_time _i , 1≤i≤N; The first data processing module is used to obtain the overall fatigue coefficient IoF of all control units and/or the unit fatigue coefficient Air_IoF of all control units based on the fatigue warning parameter Air_P = [Air_IoF ₁ , Air_IoF ₂ , Air_IoF ₃ , ..., Air_IoF _N ], wherein: Unit fatigue coefficient of the i-th control unit _{Air_Fi1} = 1, which is the weighted coefficient of the air traffic controller's duty time Air_work_time _i of the ith control unit. When the ith control unit is a civil airport, _{Air_Fi2} = 0. Otherwise, _{Air_Fi2} = Air_fm _i × Air_other_time _i / Air_work_time _i . Air_fm _i is the weighted coefficient of other working time of the ith control unit. _{Air_Fi3} is a function of the altitude Air_height _i of the ith control unit. When the hourly aircraft support flight volume of the ith control unit is ≥ the preset flow threshold Air_FT _i , the ith control unit is determined to be in the peak flow period. When the ith control unit is in the non-peak flow period, _{Air_Fi4} = 0. Otherwise, _{Air_Fi4} = Air_num_high _i × Air_high_time _i / Air_work_time _i , _{Air_Fi5} = Air_num_night _i × Air_night_time _i / Air_work_time _i , is a function of the night duty time Air_night_time _i of the ith control unit; Air_F _i6 is a function of the control method Air_manage _i of the ith control unit; Air_F _i7 is a function of the control area Air_area _i of the ith control unit; Air_F _i8 is a weighted coefficient of the non-duty time of the ith control unit; A first fatigue warning module is used to perform fatigue warning on the entire regulated industry according to the overall fatigue coefficient IoF, and/or to perform fatigue warning on regulated units of a group with a higher fatigue coefficient among different regulated unit types based on the unit fatigue coefficient Air_IoF and the fatigue warning parameter Air_P; A second data acquisition module is used for acquiring a test fatigue value Set_PreIoF=[Set_PreIoF ₁ , Set_PreIoF ₂ , ..., Set_PreIoF _M ] of a team controller of an i-th control unit in a second preset time period before taking up the post when the current demand meets the second fatigue determination demand, where M is the total number of the team controllers, wherein the test fatigue value Set_PreIoF _k of the k-th controller is obtained according to at least one of the video test data, alertness test data and subjective scale data of the k-th controller, and 1≤k≤M; The second data processing module is used to calculate the warning value Set_AL=[Set_AL ₁ , Set_AL ₂ , ..., Set_AL _M ] of the team controller, wherein Set_AL _k =Set_PreIoF _k -Set_ALP _k , Set_ALP _k is the fatigue warning threshold of the k-th controller regarding the post to be on duty, and the team fatigue warning threshold Set_ALP=[Set_ALP ₁ , Set_ALP ₂ , ..., Set_ALP _M ] is obtained according to the warning threshold parameters in the specified time interval of the team to be on duty, and the warning threshold parameters at least include the team's duty time, predicted flight volume, predicted duty duration, predicted weather conditions, predicted special conditions, and predicted attributes of the controller itself; A second fatigue warning module, used for issuing a warning based on the warning value Set_AL; a third data acquisition module, for acquiring, when the current demand meets the third fatigue determination demand, a controller comprehensive fatigue value Person_IoF=Person_C ₁ ×F_pose+Person_C ₂ ×F_face+Person_C ₃ ×F_voice of a controller Person, wherein F_pose is a current controller fatigue value based on behavior posture, F_face is a current controller fatigue value based on facial features, F_voice is a current controller fatigue value based on ground-air communication, Person_C ₁ is a weight coefficient of F_pose, Person_C ₂ is a weight coefficient of F_face, and Person_C ₃ is a weight coefficient of F_voice; The third data processing module is used to calculate the warning value Person_AL of the controller Person=Person_IoF-Person_ALP, wherein the individual warning threshold Person_ALP of the controller is obtained according to the warning threshold determination parameters set in the future time area in the post duty time period of the controller Person on duty, and the warning threshold determination parameters at least include the predicted flight volume related to the post on duty, the predicted weather conditions, the predicted special situation, and the predicted attributes of the controller Person itself; A third fatigue warning module, used for issuing a warning based on the warning value Person_AL; Based on the unit fatigue coefficient Air_IoF and the fatigue warning parameter Air_P, fatigue warning is performed on control units with higher fatigue coefficients in different control unit types as follows: first, all control units are classified based on the fatigue warning parameter Air_P; second, each type of control unit is sorted according to the size of the fatigue coefficient; then, each type of control unit after sorting is graded using a clustering method to be divided into groups with lower, medium, and higher fatigue coefficients; finally, in each type of control unit, a warning prompt is given to the control unit that enters a group with a higher value after clustering; Performing fatigue warning on the entire regulated industry according to the overall fatigue coefficient IoF includes: issuing a warning when the overall fatigue coefficient IoF ≥ the overall fatigue coefficient warning threshold Total_ALP, wherein the overall fatigue coefficient warning threshold Total_ALP is obtained by data-driven iteration based on the overall fatigue coefficients of all regulated units and the number of unsafe events of all regulated units obtained in multiple different historical time periods; The acquisition of the team fatigue warning threshold Set_ALP includes the following steps: S331, obtaining the warning threshold parameters within the specified time interval of the team's upcoming duty time, wherein the warning threshold parameters at least include the team's duty time, predicted flight volume, predicted duty duration, predicted special conditions, predicted weather conditions, and predicted attributes of the controller itself, and the duration of the specified time interval has a value range of [20 minutes, 300 minutes]; S332, obtaining a team basic fatigue value Set_P=[Set_P ₁ , Set_P ₂ , ..., Set_P _M ] of the team according to the warning threshold parameter, wherein Set_P _k represents the basic fatigue value of the kth controller of the team; S333, based on Set_P, obtaining the team fatigue warning threshold Set_ALP = [Set_ALP ₁ , Set_ALP ₂ , ..., Set_ALP _M ] of the team, wherein Set_ALP _k = Set_C × Set_P _k , and Set_C is the team position adjustment coefficient; The acquisition of the controller's individual fatigue warning threshold Person_ALP includes the following steps: S431, obtaining the warning threshold determination parameters set in the future time zone in the duty time period of the controller Person on duty, wherein the warning threshold determination parameters at least include the predicted flight volume, the predicted weather conditions, the predicted special conditions, and the predicted attributes of the controller Person himself, and the duration of the set future time zone has a value range of [3 minutes, 20 minutes]; S432, obtaining the basic required value of fatigue level of the controller's duty post Person_F _b according to the warning threshold determination parameter; S433, according to the basic requirement value of fatigue degree of the duty post Person_F _b, obtain the individual fatigue warning threshold value of the controller Person_ALP = (Person_A1 + Person_A2 × Person_A3) × Person_F _b , where Person_A1 is the human circadian rhythm coefficient of the controller during the duty time, Person_A2 is the post coefficient, which is used to indicate the requirement of fatigue degree of the controller's post; Person_A3 is the post task requirement coefficient, which is used to indicate the requirement of fatigue degree of the controller due to the expected traffic flow situation of the flight plan. 8. The device according to claim 7 is characterized in that the value range of Air_F _i3 is [0.00, 8.00]; the value range of Air_F _i6 is [0.00, 6.00]; the value range of Air_F _i7 is [0.00, 7.00], and the value range of Air_F _i8 is [0.20, 1.00]. 9. The device according to claim 7 or 8 is characterized in that, based on the unit fatigue coefficient Air_IoF and the fatigue warning parameter Air_P, fatigue warning is given to control units in the group with higher fatigue coefficients among different types of control units as follows: first, all the control units are classified based on the fatigue warning parameter Air_P; secondly, each type of control units is sorted according to the size of the fatigue coefficient; then, each type of control units after sorting is graded using a clustering method and divided into groups with lower, medium and higher fatigue coefficients; finally, in each type of control units, a warning prompt is given to the control units that enter the group with higher values after clustering. 10. The device according to claim 7 is characterized in that fatigue warning is performed on the regulated industry as a whole based on the overall fatigue coefficient IoF, including: issuing a warning when the overall fatigue coefficient IoF ≥ the overall fatigue coefficient warning threshold Total_ALP, wherein the overall fatigue coefficient warning threshold Total_ALP is obtained by data-driven iteration based on the overall fatigue coefficients of all regulated units obtained in multiple different historical time periods and the number of unsafe events of all regulated units. 11. The device according to claim 7, characterized in that the acquisition of the team fatigue warning threshold Set_ALP comprises the following steps: S331, obtaining the warning threshold parameters within the specified time interval of the team's upcoming duty period, wherein the warning threshold parameters at least include the team's duty time, predicted flight volume, predicted duty duration, predicted weather conditions, predicted special conditions, and predicted attributes of the controller itself, and the duration of the specified time interval has a value range of [20 minutes, 300 minutes]; S332, obtaining a team basic fatigue value Set_P=[Set_P ₁ , Set_P ₂ , ..., Set_P _M ] of the team according to the warning threshold parameter, wherein Set_P _k represents the basic fatigue value of the kth controller of the team; S333, based on Set_P, obtain the team fatigue warning threshold Set_ALP = [Set_ALP ₁ , Set_ALP ₂ , ..., Set_ALP _M ] of the team, wherein Set_ALP _k = Set_C × Set_P _k , and Set_C is the team position adjustment coefficient. 12. The device according to claim 7, characterized in that the acquisition of the controller's individual fatigue warning threshold Person_ALP comprises the following steps: S431, obtaining the warning threshold determination parameters set in the future time zone in the duty time period of the controller Person on duty, wherein the warning threshold determination parameters at least include the predicted flight volume, the predicted weather conditions, the predicted special conditions, and the predicted attributes of the controller Person himself, and the duration of the set future time zone has a value range of [3 minutes, 20 minutes]; S432, obtaining the basic required value of fatigue level of the controller's duty post Person_F _b according to the warning threshold determination parameter; S433, according to the basic requirement value of fatigue degree of the duty post Person_F _b, obtain the individual fatigue warning threshold value of the controller Person_ALP = (Person_A1 + Person_A2 × Person_A3) × Person_F _b , where Person_A1 is the human circadian rhythm coefficient of the controller during the duty time, Person_A2 is the post coefficient, which is used to indicate the requirement of fatigue degree of the controller's post; Person_A3 is the post task requirement coefficient, which is used to indicate the requirement of fatigue degree of the controller due to the expected traffic flow situation of the flight plan. 13. A controller fatigue warning system, the system comprising a processor and a non-transitory storage medium for storing at least one instruction or at least one program, characterized in that the at least one instruction or the at least one program is loaded and executed by the processor to implement the method described in any one of claims 1-6.