Research On Safety Evaluation And Prediction Of Civil Aviation Air Traffic Control System

With the development of social economy and the change of science and technology,civil aviation transportation industry has been constantly development,it has brought great convenience to people life,thus occupying more and more important position in human life.Air traffic management system as a key component of the civil aviation transportation system.Its safe operation or not directly affects the civil aviation safety.Therefore,to evaluate the safety of ATC system and ATC unsafe events forecast can make us understand security status and future trends of.air traffic control systems.Allowing departments to conduct macro-management,and developing appropriate policies and standards,and taking preventive measures,thus to guarantee the operation of ATC system safe and reliable and reduce the aviation accident caused by air traffic control,and enhance the security level of civil aviation.This paper embarks from the research background and significance of research,reviewed the civil aviation safety assessment and prediction research status of domestic and foreign,clarifies the research content and main innovation points,and introduces the related theory in detail.This article mainly conducts the research from the following two aspects.Firstly,study on the safety evaluation of ATC system.Based on the analysis of the ATC system and clear the principle of index selection basis,summarizing and analyzing.through the four aspects of human factors,equipment factors,environmental factors,management factors,extracting index,establishing the evaluation index system of ATC safety.In the aspect of evaluation method,choose the the extenics theory that can quantify value evaluation results and comprehensively reflect the comprehensive level of things;In terms of weight,taking into account the fuzziness of human judgment,so in combination with fuzzy and order relation analysis method to determine the weight.Through the study of the example of an air traffic control unit safety research,verified the feasibility and validity of the method.Secondly,to predict the number of ATC incidents.Due to air traffic control system is a complex nonlinear system and less sample data,and least squares support vector machine in tackling small sample,nonlinear problems have unique advantages,and given the parameters of least squares support vector machine,using improved Particle Swarm Optimization algorithm to optimize the parameters,establishing an improved Particle Swarm Optimization algorithm optimizing least squares support vector machines prediction model,selecting six impact factors in the index system as input,the ATC unsafe events as output,simulating and achieving ATC unsafe events predicted.Meanwhile,the prediction results are compared with that of the least squares support vector machines model that not optimized,which shows that the model has higher prediction accuracy in prediction.