The application region extended by the development of the robot,the unmanned aerial vehicles are used to explore the terrain or executing investigation mission,disaster relief robots are utilized to rescue.In this context,how to use the equipment available and how the rescue robot can reach the target as quickly as possible,improving the efficiency of search and rescue become a research hotspot.However,under some circumstance,the GPS signal is bad or the environment is too cluttered,the search and rescue robots have to carpet search due to their limitations on its own perception.If there are aerial robot to provide perspective for ground robots and make route planning to guide the ground robots to search and rescue,so the ground robots may be perform search and rescue missions smoothly,which not only improves the efficiency of searching and rescuing,but also reduce the search and rescue time,therefore,and the cooperation between ground robot and aerial robot is becoming very important.In this work,we study the cooperative path planning between ground robot and aerial robot under this context and we studied and finished the following work:1.Analyzing the most widely used system ROS among robot systems,and build the entire system architecture based on the ROS.2.In machine vision,putting forward an adaptive threshold method in dynamic environment to solve the problem of color changes which caused by the light changes or the robot the perspective changes or the white balance when robot moves.3.In path planning,for the basic ant colony algorithm in robot path planning environment model is easy to fall into local optimum and the speed of convergence in maximum and minimum ant colony algorithm is slow,an improved ant colony algorithm combined with the grid environment model for path planning is put forward.4.In robots cooperative model,against the current center collaboration model is harmful to the independence of a single robot and the parallel collaboration model is low in efficiency,we put forward the center-independence collaboration model.Finally,the experiment results show the validity of the model and algorithm. |