Improvement Of Ant Colony Algorithm And Its Application In Route Planning

Ant Colony Algorithm(ACA) is a metaheuristic approach for solving combinatorial optimization problems. It was first proposed by Dorigo in 1991. From then on, it has been successfully used to solve a series of combinatorial problems, such as: Traveling Salesman Problem, Quadratic Assignment Problem, Vehicle Routing Problem, Graph Coloring Problem and so on. For the great performance of ACA, researchers show great interest in the research and applications on ACA nowadays.Based on the basic principle of the first ant algorithm, ant system(AS), this article lays a strong emphasis on advantage and disadvantage of many kinds of basic ant colony algorithm (Ant System with elitist strategy, Rank-Based Version of System, Ant Colony System, Max-Min Ant System). In the paper, the original Ant Colony System(ACS) is improved, and a new Q/a0 adaptive ant colony algorithm is proposed, through adaptive changing of the pheromone constant and evaporation rate in the global updating rule. In allusion to characteristic of route planning, using the suitable route performance index, the Q/a0 adaptive ant colony algorithm is applied in route planning. The paper takes the orientative point of aerial vehicle entering the enemy defensive area as the initial station, the campaign attack goal location as the target point. It connects the initial station and the target point through carrying on the grid division to the planning space to forming network chart. The node of the network chart is defined as aerial vehicle feasible way node, in the algorithm. Through the probability choice function the best feasible spots are determined, forming a most superior route. The paper carries on the algorithm simulation using the Matlab language and compares it with AS and ACS. The result shows that the algorithm can find better path at higher convergence speed, and solve route planning well.To further prove the relevance and effectiveness of the proved ant colony algorithm. This paper uses it to solve the low-altitude 3-D route planning issues. Low-altitude route planning can use radar blind spot caused by topography to find flying optimal routes, or take advantage of the terrain and conditions to evade artillery threats and enhance the survival probability of the aerial vehicle. Therefore, low-altitude route planning must minimize the flying altitude, and at the same time, it also must meet other constraints, such as the mobile performance of aerial vehicle. First, we combine ant colony algorithm with the idea of dynamic programming to search out the optimal route in the horizontal direction, which meet the restrictions on mobility of the aircraft. Then in the vertical plane of the terrain determined by the horizontal route, slope and curvature of the terrain are smoothed to meet the restrictions on vertical mobility of the aircraft. Finally, a route, which can improve performance, but also meet the restrictions on mobility of the aircraft, is planned through simulation.