| Visual is an important way for human access to information,the eye movement candirect response to people’s psychological process,the visual characteristics of the pilot hasbeen the focus of researchs.Completion of the flight and combat missions has closelyrelated to efficient visual search strategies.Over the past decade,more eye movementmeasurement methods used in the pilot’s visual behavior,the eye tracker is based oninfrared light recording the corneal reflex,to enable real-time recording of eye movementsin the current context.After recording is completed the process of eye movement analysisavailable throughout the flight the pilot’s attention allocation and cognitive load.In order tofurther investigate the military pilots visual search features,which can enable us tounderstand the laws of distribution of attention and visual information processingconditions.In order to obtain better quality pilot visual search feature,Due to thespecialized nature of the pilot,this study used a paradigm: expert–novice.Throughtraining that the instrument flash and convinced instrument,use eye tracker to record theinformation,analysis and comparison of the different levels of pilots on the meter readings to determine the reaction,accuracy,and attention to the distribution of the three groups ofdifferent levels of test and eye movement differences between different regions of interestduring the flight subjects in "convinced instrument".In order to understand and explore thecognitive models of different levels of pilots in different cognitive load,the difference invisual search patterns and characteristics,the findings provide a reference for the novicepilot training and flight performance evaluation.The study included two experiments：First experiment,Select the instrument flash training in meter spatial orientationtraining simulator.The training were viewing flash of instrument,then flash reading of theinstrument to judge,and at the same time measuring and recording the reaction of thereading process.The training effects of different levels of the subjects in the differentinstruments present time on the meter readings to determine the reaction time andaccuracy differences and to explore its impact on cognitive processing capacity and visualstrategy.3×3×2mixed design, the independent variables for the test level (good, average,novice),excellent pilots,flight time in more than1500h(containing1500h);novicepilots,flight time in the800h (including800h),and ground soldiers the novice pilots selectthe practical experience of the flight,instrument type (speedometer,altimeter,compass) andinstrument presentation time (500ms,1000ms),which belongs to the trial level inter-groupfactors,and the rest are group factor.The dependent variables were asked to judge thereaction of the reading and the correct rate.The second experiment select convinced instrument in the meter space orientationtraining simulator,The subjects analog location only through the cockpit instrument togather information, to keep the aircraft flight and complete the task. At the same time,thesubjects need to wear eye tracker throughout the flight,we can observe and record the eyemovements was comparing different levels is in the eye movement behavior underdifferent cognitive load and complete the task the difference.3×2mixed design,theindependent variable test levels,divided into good,generally,these three groups of novice excellent pilot flight time of1500h and the novice pilot flight time of800h and morenovice pilots selectedno ground soldiers flight practical experience;Task kinds of task1and task2,task-based task, that is to keep the aircraft original state flight, the task is a taskbased on the need to complete an additional task;trial level belongs to the groupfactors,and the rest are group factors.Achievements was to complete the task as the dependentvariable.Results:Experiment one:The statistical analysis of the results of the instrument to determinethe correct rate:1.different types of instrument accuracy rate difference significant (F=4.108, P=0.023) after further LSD test showed that: the the speedometer correct rate(0.810±0.020) and compass accuracy rate (0.820±0.020) was significantly greater thanthe the altimeter correct rate (0.770±0.010),while the speedometer and compass accuracyrate difference was not significant;2.different presentation time correct rate significantdifference (F=47.713,P=0.000),after further LSD test showed that: the correct rate ofunder500ms (0.750±0.010) was significantly lower than the1000ms the correct rate(0.860±0.020);3.different correct level was significantly(F=18.870,P=0.000), furtherLSD test:the correct rate of the outstanding group (0.880±0.200) was significantly higherthan the normal group (0.810±0.020) and novice group (0.730±0.020),and then there arealso significant differences between the two,P<0.054.level of rendering time andinstrument type interaction was not significant.Determine the response of the instrument:1.the reaction of the different types of instrument differences significant (F=22.881,P=0.000),by further LSD test showed that the: speedometer reaction when (2724.740±80.330) and compass reaction(2741.610±110.100)was significantly shorter than thealtimeter response(3343.700±133.220),speedometer and compass accuracy ratedifference was not significant;2.different rendering time to react significantly different (F=24.750,P=0.000), further LSD test showed that: under500ms response (3136.910±118.640)was significantly longer than1000ms reaction (2736.460±79.120);3.the reaction of the different levels of the subjects were significantly different (F=20.090,P=0.000),further LSD test showed that: the reaction of the excellent group(2530.990±171.510) and group (2577.360±162.300) was significantly higher than the novice group(3701.690±134.540),while the formerboth there was no significant difference.4.level ofrendering time and instrument type interaction was not significant.Experiment two: the results display:1.task results: main effect of instrument type(F=358.990,P=0.000);significant main effect of the level of task (F=614.14,P=0.000);was thetrial level main effect of significantly forward (F=8.697,P=0.004)2.pupil size: the maineffect of the type of task significantly with (F=18.776,P=0.005); was the main effect of thetest level significantly forward (F=8.804,P=0.000);was the trial level and tasklevelinteractions was significant (F=62.410,P=0.000):3.watching frequency:significantmain effect of task type with (F=104.610,P=0.000);was the main effect of the test levelsignificantly with (F=33.530,P=0.000);4.of the average viewing time: a significant maineffect of task type(F=28.070,P=0.000);the level significant main effect (F=6.120,P=0.007);5.saccade frequency: significant main effect of task type (F=109.010,P=0.000);level was a significant main effect (F=4.210,P=0.000);interaction between the trial leveland the level of task significant (F=4.450,P=0.020).6.the average eye jump time: the maineffect of task type (F=49.160,P=0.000);is the trial level main effect is not significant with(F=1.820,P=0.180)is the trial level and task level interaction a significant effect (P>0.05)7.average eye jump amplitude: the main effect of the type of task significantlysignificantly(F=34.300,P=0.000);significant is test the level of the main effect (F=7.825,P=0.002).8.average eye jump speed: main effect of task type (F=25.880,P=0.000);singnific-ant is test the level of the main effect(F=8.131,P=0.002);9.watching time percentage:significant main effect of region of interest(F=111.130,P=0.000);area of interest,and thetrial level interactions significant(F=22.100,P=0.000);significant interaction between theregion of interest and the level of task (F=37.470,P=0.000).10.the percentage of thefixation point: significant main effect of region of interest (F=72.776,P=0.000);is the trial level main effect is not significant with (F=2.552,P=0.090);interaction between the regionof interest and the level is significantly(F=39.130, P=0.000);between the region of interestand the level of task interaction significant (F=13.760,P=0.000);region of interest and thelevel is a significant interaction (F=22.100,P=0.000);interaction between the region ofinterest and the level of task significant (F=37.470,P=0.000).Conclusion:1.During the task,expert pilots make good use of their own experience, modularinformation processing, decision-making capacity is also strong.2Expert and novice take a different visual search strategies, and expert strongercognitive skills.3.Experts with more efficient visual search patterns and strategies, the studypreliminary to and summarized as follows: the pupil of the rate of change of theoutstanding group of pilots in the experiment is smaller, shorter fixation time, saccadefrequency, saccade amplitude and fixation frequency, saccade amplitude larger.4.Experts Meter importance and experience to optimize the allocation of attentionmode, the novice take widespread attention, and therefore result in lower efficiency of itsvisual search.5.Excellent group and general group not only has a more efficient visual search mode,and also it has a good flight performance.In summary, the experts have been a professional flight training and flight experiencea certain flight duration, and novice there are some differences in the mission and eyemovement and behavior indicators. The expert visual search more efficient, better flightperformance, reminds us that the visual search models and strategies can be used as areference pilot training and flight performance evaluation. |
PDF Full Text Request