Research On UAV Hedding Method Based On Improved Ant Colony Algorithm

In recent years,unmanned aerial vehicles are widely used in various industries due to their low cost and flexible characteristics.With the increase of the number of commercial unmanned aerial vehicles,the accident rate of UAV collisions is also increasing year by year,and the flight safety of UAV is increasingly valued by people.Whether UAV have the ability to avoid risks independently is the key to their survival when they are on missions,and it is also an essential research content in the development of UAV.The core of UAV risk avoidance system is realized by risk avoidance algorithm.At present,the research on the flight avoidance technology of UAV on a two-dimensional plane at the same height is relatively mature,but the complexity of UAV movement determines that the flight avoidance technology on a two-dimensional plane alone cannot meet the requirements of autonomous flight of UAV.In the complex three-dimensional environment,the UAV risk avoidance technology needs to be further studied.First of all,this paper makes a comprehensive analysis of UAV flight risk avoidance,summarizes the constraints of UAV flight,and expounds their characteristics and the problems to be paid attention to when studying UAV flight risk avoidance.According to the flight characteristics of UAV,a complex three-dimensional environment simulation model was established.In order to prevent the environment model from affecting the accuracy of subsequent hedging algorithms,the environment model was processed by using the two-dimensional cubic convolution interpolation method to make the environment model more consistent with the requirements of UAV for hedging.Secondly,the research on UAV hedging algorithm is started.Ant colony algorithm is finally selected as the UAV hedging algorithm by comparing the advantages and disadvantages of the intelligent algorithm.On the basis of ant colony system,an improved ant colony algorithm is proposed,which takes the distance between obstacles and UAV into account of the pheromone update of ant colony algorithm,so that the transfer probability can select the direction of the ant in the optimal allocation way,improve the convergence speed,and prevent the ant from falling into the local optimal.Through MATLAB simulation verification,the proposed algorithm has higher accuracy,in the complex static three-dimensional simulation environment,successfully completed the autonomous flight of the UAV to avoid risk.Finally,aiming at the problem that the convergence rate of ant colony algorithm cannot meet the needs in the dynamic three-dimensional environment,a particle swarm optimization algorithm based on the improved ant colony system is proposed.Particle swarm optimization is used to form the sub-optimal solution,which is the initial condition of the distribution of pheromone.In order to improve the convergence rate of the fusion algorithm,an adaptive condition is added to the hybrid particle swarm optimization algorithm.Two improved intelligent algorithms are fused to form the improved particle swarm ant colony fusion algorithm.The performance of the fusion algorithm and the other four algorithms are tested by using four test functions.The results show that the convergence and convergence accuracy of the fusion algorithm are better than those of the other four algorithms.Finally,the simulation is carried out in a dynamic three-dimensional environment.On the premise that the UAV successfully avoids static obstacles,it makes evasive actions to the dynamic aircraft in time.This algorithm has strong universality and can be used in the remote path planning and target recognition of UAV,which has important research significance and wide application value.