Research On Optimization Method Based On Complex Network For Crossing Waypoints Location

With the rapid development of civil aviation, the existing air route network (ARN) can't meet the growing needs of air traffic flow, so it is necessary to design the ARN scientifically and efficiently. The Crossing Waypoints Location Problem (CWLP) is a crucial problem in the optimal design of ARN, which is aimed at providing reasonable layout for communication, navigation and monitoring facilities, distributing and using airspace resources scientifically, improving the air transport efficiency. At present, people mainly solve CWLP from the economic viewpoint and can get satisfactory results when the air traffic density is considerably low. However, with the development of the civil air industry in China, airspace congestion and flight delay has become more and more serious for increasing air traffic flow. Lacking of effective treatments for these new problems becomes the main bottleneck of existing methods, airspace congestion has become an urgent problem in CWLP.According to the inherent characteristic of CWLP, we present traffic flow model of ARN, propose a single-objective particle swarm optimization with betweenness guiding and a multi-objective particle swarm optimization introducing airspace capacity. The main research works on CWLP of this thesis are as follows:(1) The existing method based on flying efficiency may cause ARN to fall into congestion when getting maximum flying efficiency. Aiming at this problem, this paper gets heuristic knowledge for guiding ARN planning through traffic flow modeling analysis. Then make the airspace congestion as one of constraints. Synthesizing the congestion rules from modeling analysis, we propose a heuristic particle swarm algorithm with betweenness guiding to solve CWLP, to maximize the flight efficiency under the condition of ARN's congestion does not appear.(2) The existing method which only considers flying efficiency ignores an important traffic performance index of ARN: airspace capacity, the optimized ARN will quickly fall into congestion with increasing air traffic flow. Aiming at this problem, this paper propose a multi-objective particle swarm algorithm introducing airspace capacity for CWLP, which consider both flight efficiency and airspace capacity. First, give one kind of the quantitative description method for airspace capacity through complex network modeling; Then, construct multi-objective optimization model based on the total airline cost (TAC) and the standard deviation of betweenness (SDB). At last, adapt multi-objective particle swarm algorithm to get non-dominated solutions. This method takes flying efficiency and airspace capacity into account, the optimized ARN can delay the emergence of airspace congestion effectively when facing the growth of air traffic.