Research On The Fatigue Risk Assessment Of Air Traffic Controller

Air traffic controllers are key players in ensuring the safe operation of aircraft.They are responsible for preventing aircraft from colliding,maintaining air traffic order,and improving operational efficiency.The research shows that about 20% of controllers have been fatigued when implementing control tasks.In the event of air traffic insecurity,work fatigue related to controller fatigue accounts for about 18%.Controller fatigue has become an artificial factor in air traffic control at home and abroad hot issues in the field.This article designed a controller fatigue cause and fatigue degree questionnaire,using the Cronbach’s coefficient to test the reliability and validity of the questionnaire to meet the research requirements,through AHP,the weight of each cause was obtained.taking 183 controllers from a domestic three air traffic control units as the research object,the quality of sleep will be studied.Index,workload,and alertness are the identification factors of the controller’s fatigue status.A controller fatigue status recognition model based on BP neural network is constructed.The network structure and training process are designed.The model is validated by MATLAB program.The results show that the fatigue value of the controller’s fatigue obtained by the network training is somewhere between ?0.2,and the sample’s overall regression coefficient is R=0.99996.The model’s recognition accuracy is higher.Based on the controller’s fatigue values and the statistical results of fatigue insecurity in the past 10 years,the extremum theory,threshold(POT)model and Q-Q plot were used to analyze the tendency of the sample to have a thick tail distribution.The integrated Hill map and the mean overreach function(MEF)In the figure,the POT model threshold and the controller fatigue over-threshold sample random variable are obtained.The generalized Pareto(GP)distribution approximates the distribution of random variables,and the distribution function of the controller’s fatigue threshold over the sample is obtained.The maximum likelihood is obtained.The estimation method calculates the distribution parameters and obtains the functional relationship between the controller’s fatigue status and the risk probability.A controller fatigue risk probability evaluation model is established.The MATLAB program is used to calculate the cumulative probability of the different fatigue values and the K-S is used.Methods The goodness-of-fit test was performed on the distribution.The results show that the controller’s fatigue threshold value sample obeys the shape parameters,scale factor,location parameter’s GP distribution,and the sample’s overall loss distribution function can well estimate the controller’s fatigue risk probability.The controller’s Fatigue Risk Index(FRI)was obtained by defining the severity of therisk consequences,and the reference for the fatigue risk rating of the controller was established.Through the statistics of the annual flight guarantee volume and the person-to-person ratio of different ATC units,different ATCs were compared and analyzed.The average fatigue risk index for units,different regulatory authorities,and different duty hours,and finally put forward measures for prevention and control of fatigue risks.The results show that the average FRI of the controllers of units with higher per-unit ratios is higher(FRI = 0.81);the average FRI of controllers in the regional control room is the highest(FRI = 0.76)in the different control departments;The average FRI of the controllers on duty at at 22:00--07:00(the next day)was the highest(FRI=0.82).