Research On Flocking Control Of Amigo Bots

With the arrival of German Industry 4.0,the robotic industry enters a development period of spurt.On the one hand,industrial robots become the main force in the field of manufacturing.Numerous robots are used in auto industry and electronics manufacturing industry recently.On the other hand,many internet companies,such as Google and Baidu,begin to get involved in the robot industry with advanced Internet technology.Google commits to driverless vehicle formation,DJI also devotes to the flocking of UAV lately.Due to the application of Big Data,robot has been networked and intelligent.Intelligent robots have become a new synonym for robots.In this dissertation,AmigoBots produced by ActivMedia Robotics company are utilized as a research platform.The strategies of flocking control of AmigoBots in barrier-free and barrier access environment are respectively studied.The main work done in this dissertation are presented as follows.Firstly,the environment model of robots is builded with grid map,and the certainty of sonar sensor is discussed.The evaporation and enhancement mechanisms are introduced in this section which simulates the cognition of the environment by the flocking robots.By using evaporation and enhancement mechanism,the performance of the flocking control strategy for the robot in an obstacle environment is improved.Secondly,the strategy of flocking control of Amigo Bots in barrier-free environment is studied.An artificial potential function is designed to evaluate the relative position of any two adjacent robots.The total artificial potential function of each robot describes the properties of the currently flocking state.Then Newton method is introduced which is responsible for searching the optimal motion orientation of robots to enable robots to form the desired flocking pattern.In the experiment,the model of both leader agent flocking control and no-leader agent flocking control is designed.Experimental results verify the effectiveness of the designed methods for the flocking control problem of five AmigoBots.Finally,the strategy of flocking control of AmigoBots in barrier access environment is studied.The robots use the sonar sensor to detect the obstacle information.Environment model is constructed by using grid map with certainty.In this paper,a repelling force model between robot and obstacles is designed to control the motion orientation of robots to avoid obstacles.Then Newton method and repelling force are combined,which ensure AmigoBots achieve the desired formation in barrier access environment.