The Research Of Crew Error Prediction Model Based On Task

In recent decades, the development of civil aviation has seen human factors as the main reason for aviation accidents. Human errors occurred in the execution of flight tasks have significant impacts on safety issues. A number of serious accidents were caused by human errors in the past 20 years. Therefore, figuring out how to prevent or reduce flight crew errors is crucial to the reduction of aviation accident rate. Feasible methods to reduce aviation accident rate are: finding out the reasons for flight crew errors using human error prediction technology; improving cockpit design and carrying out more training of crews.Based on the demand for a safer civil aviation industry, as well as requirements for more handy cockpit designs from Europe and the US human factors seaworthiness terms, this paper investigates human error prediction technology development and its application. It also sets up a human error prediction model to predict human errors in the execution of flight tasks. Main components are as follows:(1) Studying current human error research methods and grouping them into five classes according to different focus of each method;(2) Contrasting current classification of human errors and grouping them into three classes based on their time of appearance(provides reference for the setting up flight crew error classification of this paper);(3) Contrasting current methods of flight task analysis. Using HTA method to analyze flight tasks. Improvements to HTA method are suggested in this section. As a part of the prediction model, the new analysis process and stopping rules suggested in this section provides operation sequence for the model;(4) Groping crew errors into nine classes based on different features of flight tasks. Developing prediction model on the basis of this classification;After establishing the model, the paper applies it to a real-life take off task. The feasibility of the model is tested through the prediction of possible human errors in the execution of this task. Improvements to cockpit design are also suggested.