Research On Collaborative Obstacle Avoidance Algorithm Based On Multi-sensor Network

In recent years,China has accelerated the opening of low-altitude areas,the application of unmanned aerial vehicle have been more common.But the low-altitude environment is complex,for the safety of the flight,we have to protect the unmanned aerial vehicle to avoid obstacles and collision avoidance with other unmanned aerial vehicle in real-time when it is accomplishing the task,so collision avoidance research of unmanned aerial vehicle is particularly important.In this thesis,a collaborative obstacle avoidance algorithm based on multi-sensor network model is proposed to solve the problem of low-altitude collision avoidance,which combines with the characteristics of the overall motion between the agents in the multi-sensor network through local cooperation.This article focuses on the algorithm of multi-sensor network collaborative obstacle avoidance,mainly to solve the following four basic problems:1)How to form a sensor network for the aerial vehicles involved in the possible collision;2)How to avoid collisions and obstacles in the multi-sensor network;3)How to join in the existing net for a new comer;4)How to leave from the network to its own direction for each aerial vehicle.The simulation results show that the algorithm can solve the above four problems.When multiple aerial vehicles are involved in a possible collision,a multi-sensor network is cooperatively built to avoid collisions,each agent can complete the collision avoidance in the network,the algorithm successfully solves the problem when a new comer joins in the network.After the agents finishing the collision avoidance task,they will leave the network one by one to complete the initial task until arriving to the destination.Finally,we use multiple unmanned aerial vehicle work in the low-altitude environment to verify the correctness of the algorithm,the flight paths show on the ground station,the experiment shows the multi-sensor network-based collision-free trajectory planning algorithm can successfully keep the safety when unmanned aerial vehicle fly in the sky.Compared with the previous multi-sensor network research,the main contributions of this article are:1)Establish an MSN-based framework of collision avoidance for multi-aerial vehicle.2)A hard distance keeping is introduced by considering the maximum velocity and acceleration of the aerial vehicles.