Research On Hard Landing Risk Of Navigation Training Based On Flight Data

Approach and landing stages are the most important and dangerous stages in flight.In flight training,the over-limit events are often caused by the control problems of flight cadets,among which the frequency of hard landing is very high.Hard landing will not only bring damage to each part of the aircraft fuselage,but also threaten the personal safety of trainees and instructors and seriously affect the quality of flight training.Therefore,based on the actual situation of over-limit events in flight training and the flight parameter data provided by GARMIN-1000 avionics system,this thesis carries out the assessment of hard landing risks and the research on the trend prediction of hard landing risks,so as to realize the scientific control of hard landing events.The main research overviews of this thesis are as follows:(1)Description of flight training and over-limit events.First,the relevant content of flight training is introduced to clarify the practical significance of this research model Cessna-172 R.Then the definition of hard landing event is determined by statistical analysis of over-limit event.Finally,from the perspective of flight cadet control,the causes of hard landing are analyzed from the aspects of attention distribution,point and line,direction and position,trim,speed and so on.(2)GARMIN-1000 avionics system and flight parameter data.By introducing avionics system,the characteristics of flight data recorded by the system are described,and the data are classified from the three angles of flight environment parameters,flight state parameters and engine parameters.The existing problems in the original flight data such as missing values and blank values are processed and explained to improve the availability of flight data.(3)Hard landing risk assessment.By analyzing the flight parameters data of a large number of historical landing samples,the evaluation index parameters of the model are determined.Then refer to the security manual to classify the risk level,and using the C-OWA operator to optimize the matrix element coding,and build a risk assessment model based on the interaction matrix in the end.Finally,the risk assessment is completed with the instance substituted into the model,and the feasibility of the evaluation model is verified by analyzing the record of the instance in the flight training system.(4)Forecast of trend in the occurrence of hard landing risk.Based on the analysis of stationary and non-stationary time series modeling process and the characteristics of hard landing historical statistical data,a risk trend prediction model based on ARIMA time series analysis was established.Then through model identification,grading,and testing,and then select the optimal model for prediction.The final analysis of the model results against the actual values shows that the model has a great predictive effect.