Control And Simulation Of Pollution Based On Cyber-Physical System(CPS)

With the rapid development of computer and Internet technologies, the research ofCyber-Physical System has been widely used. The CPS integrates computing, communication andcontrol capabilities. It can monitor one or more entities dependably, safely, securely and efficientlyin the physical world. The main research of CPS is how to enhance the performance of perceptionand coordination in the complex time varying environment.This thesis conducts a thorough study on the problem of decision-making, task-assigning,area-covering, and obstacle-avoiding in dealing with pollution in our environment. The main workis listed as follows:First, the “Win-Stay, Lose-Swith”(WSLS) algorithm is proposed in this thesis based on theperformance of the human team. By using this algorithm, robot team can make decision based onsome rewards securely, efficiently and real-timely like human being.Second, the CWVT algorithm is proposed in this thesis based on weighted distance andVoronoi technology. It can easily solve the problem of WSLS algorithm, which cannot perceive orcover the pollution area very well. Beginning with the definition of Voronoi diagrams, someproperties and algorithms related to Voronoi diagrams are comprehensively summarized. Then anew partition method, based on distance and discrete method, is proposed to simplify the generationof Voronoi diagrams.Third, the CWVT-Allocation algorithm is proposed to solve the problem of task allocating.The Moving Least Squares (MLS) is used to fit the curve of pollution function. Then determine thenumber of robots for each pollution resource. Each robot weights between the impact of distance toresource and spreading of pollution, and then chooses one resource to clear. Obtained algorithm canreduce the time of tracking and clearing resource effectively.Fourth, an improved CWVT-Obstacle-Avoidance algorithm is also proposed for obstacleavoidance in the environment where different sources and obstacles are allowed. This methodintegrates virtual world and reality world. In virtual world, we do not consider the impact ofobstacles, but in the reality world, obstacles should be considered. This method can simplifytrajectories for the robot to follow and expand applications fields of Voronoi diagrams.Fifth, the simulation platform Diff-MAS2D is used to illustrate the effectiveness of ouralgorithms. Simulation results have been shown that robot can track resource timely, avoid obstacle effectively and control the diffusion process very well.