Research On Key Techniques Of Aircraft Surface Trajectory Prediction And Path Planning In Airport

With the rapid development of the global commercial aviation transport industry,airport surface traffic is becoming busier and busier,and the layout of airport surface has become more complex.The trend of multi-runways,multi-taxiways and multiterminals has been developing.Airport operations have gradually become a bottleneck restricting the development of commercial aviation traffic.The frequent occurrence of slight collision and flight delays have increased year by year,making the domestic and international commercial airports more and more longing for the need for collision-free,taxiing distance and taxiing time of the surface aircraft taxiing path.For this reason,this thesis has conducted an intensive research on the key technologies of aircraft surface trajectory prediction and path planning.First of all,this thesis analyzed the operation process of surface traffic and the related constraints that influence the decision-making of surface traffic operations,as well as making an statement on the taxiway collision.It also dealt with the different intersection types in the taxiway system,and the surface structure diagrams were restructured.The connection relationship between the path nodes in the network structure was represented by adjacency matrix.Secondly,aiming at the conflict of surface taxiing path,this thesis analyzed the characteristics of taxiway trajectory data,made a preprocessing such as probability sampling,first-order difference and standardization on the data,and constructed a trajectory prediction model based on Long Short Term Memory networks(LSTM).Combined with the change of the motion state of the surface aircraft,the attenuation memory window was introduced to improve the hidden layer structure to further enhance the prediction accuracy in the model,and it was compared and verified under different parameters.Then,the method of curve fitting was used for dealing with the prediction results by polynomial fitting,aiming at reducing the prediction error caused by the training of data sets under different sampling factors,achieving the medium and long-term position prediction of surface aircraft,and laying the foundation for the implementation of taxiing conflict detecting and resolution.Finally,through the analysis of the surface taxiing path planning problem,a mathematical model of surface path planning was established,and corresponding conflict detecting and resolution methods were provided for three different types of taxiway collision.On the basis of genetic algorithm,the real number form of path node coding was designed,an initial feasible solution population was generated by heuristic path search algorithm,and the crossover and mutation operator of adaptive change was formulated.Combined with trajectory prediction technology,a dynamic path planning algorithm with automatic conflict detecting and resolution characteristics was designed.Then,through the algorithm validation and case analysis,the effectiveness and security of the algorithm were verified.This thesis enriches the application research of surface path planning technology and benefits the safe and efficient operation of surface traffic.