Air Traffic Controllers Cognitive Load Comprehensive Evaluation Based On The Multiple Physiological Parameter

Many researchers have been only evaluate objective(observed) load when come to evaluate ATC controller workload, while ignoring the subjective(cognitive) load. Cognitive load inconvenience for record and time, hard to quantify, but are the most important factors. High cognitive load will seriously affect the person's work efficiency, operation reliability and physical and mental health, ultimately affect the efficiency and reliability of the man-machine system. Therefore, the research of controllers' cognitive load scientific quantitative play a very important role in evaluation the air traffic controllers' workload comprehensively, and have some help in setting controller's training and working system.This paper deeply analyzes the controllers' cognitive load evaluation theoretical and practical significance, summarizes the evaluation method of cognitive load and its advantages and disadvantages. Advanced a new cognitive load evaluation method from the perspective of cognitive physiology, overcoming the subjectiveness and indeterminacy in former evaluation methods.This paper first combined with cognitive psychology and controllers operating characteristics, given the cognitive structure in controller performing the task. Then detailed demonstrated the process of experiment design and experiment, screening the collected data, uses variance analysis to get the indexes sensitive to changes in cognitive load: fixation duration, fixation frequency, the diameter of the pupil, heart rate, RMSSD, LF.In the study of the relations between controllers' cognitive load and aircraft condition, considering from aircraft's relative position relationship and operation stage. Based on the visual information sources, and the method of dynamic interest area classification is adopted, and the radar display interface is divided into eight areas. The eye movement data of participants are recorded and the method of Markov chain is used to analyze one-step transition probability of the visual gaze data in various visual areas. Experiment result shows that the cognitive load caused by approaching aircraft is significantly higher than departing aircraft, the cognitive load caused by radar signs is higher than the aircraft itself.Finally, establishes BP neural network model which is conducted feasibility test with NASA-TLX subjective assessment model, which shows a high consistency. Making a correlation analysis between cognitive load model and dynamic density model, showing logarithmic relationship, which is also consist with the theory of limitations in human cognitive resources.