Swarm Robot Multi-objective Search Task Allocation And Parameter Selection Research

Being one of typical artificial swarm systems, swarm robotics systemmainly researches how a group of robots with relatively simple structure andfunction to complete relatively complicated task by coordination andcooperation. The research field of swarm robotics system is very wide, includingtarget search,terrain coverage,task allocation and so on. This paper is focusedon studying task allocation and communication problems in multi-targetlocalization in swarm robotics, and the work is as follows:(1)This paper applies distance variable to response threshold model of waspswarm to solve task allocation in multi-target localization in swarm robotics, theresponse threshold model of wasp swarm is an effective model of division oflabor, applying distance variable to the model to solve task allocation inmulti-target localization in swarm robotics can effectively reduce the probabilityof searching a target for the robots, which are far away from the robot thatreceived the strongest signal, and to a certain extent can avoid too muchcommunication cost caused by data forwarding between robots that are far awayfrom each other belong the same subgroup. Simulation results have shown thatthis method can effectively enable the robots to allocate task based on differentsignals.(2)Considering the significant impact of parameters on the task allocationalgorithm based on the introduction of distance variable to the responsethreshold model of wasp swarm in multi-target localization in swarm robotics,this paper employs the uniform design method on optimization of parameterselection on the five important parameters in the task allocation algorithm, andfinally selects a set of parameters that can make the task allocation algorithm toachieve good results by direct analysis and linear regression analysis of the testresults.(3)For the communication problems in multi-target localization in swarmrobotics, this paper proposes a feasible asynchronous communication strategy based on parallel asynchronous communication mechanism, and simulationresults have shown that this method is feasible and effective. Secondly, theproblems of the task allocation algorithm in multi-target localization under thecondition of local communication have been shown through the simulation.Finally, this paper employs the uniform design method on optimization ofparameter selection on the important parameters in the task allocation algorithmin multi-target localization under the condition of local communication, andselects a set of parameters that can make the task allocation algorithm to achievegood results.