Flight Landing And Departure Sequencing Optimization Strategy Based On Improved Particle Swarm

Flight departure sequencing is a key link to improving the efficiency of civil aviation transportation.The single-objective sequencing with the least time or economic loss will lead to uneven distribution of delay time and the sequencing result is contrary to the principle of fairness,while the multi-objective sequencing has the defects of slow convergence of the algorithm and easy to fall into local optimum when solving,which leads to low computational efficiency and poor optimization effect.To address the above problems,this paper investigates the construction of a multi-objective model for off-field flights and the improved method of the particle swarm algorithm.Firstly,the sequencing fairness is abstracted into the standard deviation of delay time,the multi-objective function is defined with the total delay time and the minimum standard deviation,the flight sequencing priority is proposed,the delay time window constraint and wake interval constraint are constructed,and an exit sequencing model considering the delay fairness is formed.Then,the related theories of particle swarm optimization and multi-objective optimization are elaborated,with a focus on Pareto optimal theory.In order to enhance the global search intensity,improve the change of the inertia weight coefficient in each iteration,introduce the inertia weight exponential reduction strategy,and in order to accelerate the convergence efficiency of the algorithm,the fixed static learning factor is changed to a dynamic learning factor affected by the individual learning factor,social learning factor and the number of iterations,and form a particle swarm algorithm with linear decreasing inertia weight index.In terms of iterative velocity,the velocity control factor is introduced to control the velocity update,the particle velocity formula is constructed,and the particle swarm algorithm based on velocity control is proposed.Finally,combined with the inbound and outbound flight data of Xiamen Gaoqi Airport,the results show that compared with the first-come-first-served method and inertial weight linearly decreasing particle swarm algorithm,the total delay time and 27% and 28% of the delay standard deviation in the uncrowded scenario are reduced by 72% and 26% compared with the first-come,first-served method and inertial weight linear decreasing particle swarm algorithm.In the crowded scenario,the proposed method reduces the total delay time compared with FCFS by 69%.In the case of flight approach conflict rate as high as 90%,the total delay cost of traditional multi-objective particle swarm algorithm,particle swarm algorithm with linear decreasing inertial weight exponential and particle swarm algorithm based on velocity control is 5040 yuan,3300 yuan and 2200 yuan,respectively,which is reduced by 82%,86% and 92%compared with FCFS,and the total delay cost of particle swarm algorithm based on velocity control is reduced by 56% and 33%,respectively.The SMPSO algorithm is 94% less than the FCFS method and 26% less than the traditional MOPSO algorithm.