Study On Redundancy Measurement And Optimization Of Aviation Network Under Severe Weather

As an important intercity transportation mode,air transportation has the advantages of high speed,strong maneuverability,and high comfort.With the rapid development of China’s aviation system,it is particularly important to ensure the stable operation of the aviation network.However,the normal operation of the aviation system is easily affected by weather factors.Faced with increasingly uncertain risks and disturbances,the redundancy in the aviation network can effectively resist the disturbances brought about by the occurrence of disasters,and maintain the relative stability of the network’s performance when facing disaster attacks.Based on this,this paper abstracts the real aviation system as a topological network,and studies the redundancy of aviation network under severe weather events.The work content of this paper is as follows:(1)Construction of two-layer aviation network model under bad weather.By analyzing the attack characteristics of severe weather on aviation network,this paper simulates the spatial influence area of severe weather events.Then,based on the traditional network construction model,a modeling method of dual layer aviation network is proposed.Finally,according to the disaster distribution characteristics of the network,the disaster probability of different airports is calculated,and the random attack model is extended to a weighted random attack model to simulate the aviation network when it is attacked by bad weather,so that the improved attack strategy is more in line with the actual disaster attack situation.(2)Research on redundancy measurement and optimization of aviation network.In terms of measurement,this paper evaluates the changes of network performance after disaster,and then calculates the redundancy of aviation network.In terms of optimization,this paper uses the link prediction model to obtain the alternative set of routes,and considers the closeness between alternative nodes and the public neighbor nodes,as well as the influence of the closeness between the public neighbor nodes on the contribution allocation in the network,and improves and expands the functional functions in the local naive Bayes link prediction model.According to the alternative set of routes obtained by the prediction model,from the perspective of network topology and disaster characteristics,an appropriate edging strategy is designed to select the edging nodes,so as to determine the appropriate routes to be added to the network,so as to improve network redundancy.(3)Empirical study on China’s aviation network.In this paper,airport and flight data from China’s aviation timetable were used to construct the aviation network model.Daily meteorological data from the ground climate dataset in China between 2011-2020 were collected as disaster attack data,with strong winds,heavy rain,and blizzards selected as representative severe weather events for case studies.First,statistical indicators were used to calculate the network’s static geometric characteristics,and the topological structure of China’s aviation network was analyzed.Next,the performance change curve of the network after being attacked by disasters was visualized,and the trend of structural and functional changes in the network after the disaster was discussed,and the network redundancy was calculated.Finally,the strategy of adding redundant edges based on the link prediction model was used to increase the network redundancy,and the effectiveness and superiority of the strategy were verified.The research results showed that the optimized network had improved both in structure and redundancy.The paper is based on real data and its proven results are reliable.The above research is of practical significance in understanding the structural characteristics of China’s aviation network,analysing the functional loss of the network after a disaster,and improving the disaster resilience of the network.At the same time,the redundancy measurement and optimisation model of aviation networks proposed in this paper also enriches the theoretical results in the field of complex transportation.