Workload Prediction of Aircraft Borescope Inspection
Borescope inspection is the most important detection item in aircraft maintenance.In the process of borescope inspec-tion,the workload is directly related to the quality of detection.Under the condition of high load,it often induces safety risks such as wrong detection,forgetting detection and missing detection,resulting in maintenance errors.In order to solve the non real-time measurement of load such as NASA-TLX,the eye tracking data of borescope inspection under different workload mea-sured by Tobii Glasses 2 without contact and then the key eye tracking indexes of constructing workload prediction model are found by ANOVA.In addition,because there is a nonlinear relationship between eye tracking index and workload,SVR is se-lected to build the workload prediction model,and GASA algorithm is used to optimize the SVR parameters to obtain the work-load prediction model of borescope inspection with sufficient accuracy and generalization ability.So as to help the maintenance and repair organization supervise borescope inspection personnel in real-time,reduce the risk caused,and provide the basis for the CAAC to formulate corresponding regulations on the management of borescope inspection.
NETL
NSTL
万方数据
维普
