基于PSO ANN方法的飞行学员前庭错觉研究  

A Study on Vestibular Illusion of Flight Cadets Based on PSO ANN Method

作　　者：司海青[1] 潘亭 梁思远 汪海波[1] 汤子宁 赵家辉 SI Hai-qing;PAN Ting;LIANG Si-yuan;WANG Hai-bo;TANG Zi-ning;ZHAO Jia-hui(Nanjing University of Aeronautics and Astronautics,Nanjing 211000,China)

机构地区：[1]南京航空航天大学,江苏南京211000

出　　处：《航空计算技术》2022年第1期6-10,共5页Aeronautical Computing Technique

基　　金：国家自然科学基金与民航基金联合重点项目资助(U2033202);教育部2020年第一批产学研合作协同育人项目资助(202002035025);教育部2021年第一批产学研合作协同育人项目资助(202101042005);中央高校基本科研业务费专项资金项目资助(NS2020051);国家级大学生创新创业训练计划项目资助(202010287049Z)。

摘　　要：探讨飞行学员在不同转速、不同头动状态下反旋转错觉的心率变异性的特征。通过飞行学员前庭功能训练实验模拟飞行学员在60°/s、90°/s和120°/s三种转速下的反旋转错觉,其中,转向分为正向和反向两种,实时获取被试者的心率变异性等生理数据;通过双因素方差分析法对转椅运动过程中的转速和转向对生理数据的显著性影响进行分析;在显著性分析的基础上运用粒子群对人工神经网络模型进行优化,构建飞行学员反旋转错觉辨识模型。研究结果表明,飞行学员在前庭功能模拟训练过程中,旋转方向比旋转速度对飞行学员的影响更大,心率变异性差异更明显,从而为飞行学员训练微观行为考核与评估提供理论依据与技术支持。The characteristics of the heart rate variability of anti rotation illusion in flight cadets caused by different rotational speeds and head movements were explored.Firstly,the anti rotation illusions of flight cadets at 60°/s,90°/s and 120°/s were simulated through vestibular function training experiment,in which the steering was divided into forward and reverse directions.And the physiological data of heart rate variability were obtained in real time.Secondly,the significant effect of rotating speed and steering on physiological data was analyzed by the two way analysis of variance.Finally,the artificial neural network model was optimized by using particle swarm based on the results of significance analysis to construct a model for the recognition of counter rotation illusion of flight trainees.Results show that the direction of rotation has more influence on the flight cadets than the speed of rotation during the vestibular function simulation training,and the difference of heart rate variability is more obvious.Thus,it provides a theoretical basis and technical support for the assessment and evaluation of flight cadet training microscopic behavior.

关 键 词：飞行训练 前庭觉 心率变异性 人工神经网络 粒子群算法 

分 类 号：R852[医药卫生—航空、航天与航海医学] V328[医药卫生—临床医学]