Research On Task Allocation Of Multi Emotional Robot

Emotional robot is a kind of robot which owns the similar human emotion.Human endow robot with emotion by artificial methods and technologies. Thenrobots can own autonomy and can adjust their behavior selection by the effect ofemotion mechanism. Multi emotional robot task allocation problem is to introduceemotion mechanism into the traditional task allocation problem. It mainly studieshow robots select the suitable task through the effect of emotion mechanism andhow robots perform the task effectively in order to get the maximum team gain. Theproblem becomes one hotspot of multi robot system. Solving the problemeffectively can improve the stability of the multi-robot system and can exert thesuperiority of multi-robot system.This thesis is funded by the national science foundation of china "research oncomplex system causal relationship discovery and robustness infer under thedynamic environment " and the natural science foundation of Anhui province"research on collaboration and competition technology of emotional robot based onquantum game theory ".The mainly study is how to construct an emotion model andhow to apply the emotion mechanism to the multi robot task allocation problem.The main content is as follows:Firstly, a reasonable emotion model is constructed in this thesis. Based on thefoundation of emotion psychology knowledge, a dynamic emotion transfer model ofthe emotional robot with different personality under the different externalstimulation is presented. The maximum similarity matching emotion transformationmodel based on mapping between state space and probability space is proposed inthis thesis. Firstly, the current emotion probability is calculated by HMM process.Then the current concrete emotion state is got by maximum similarity matching.The experiment results show the emotion transformation process which is simulatedin the model corresponds with the general rule of human emotion transformation.Secondly, the emotion mechanism is introduced into the traditional taskallocation problem in this thesis. Achieve efficient allocation of tasks througheffect of emotion mechanism. An auction algorithm based on collaborationintention is proposed and the concept of collaboration intention is defined in thisthesis. In this algorithm, robots decide to perform the task or not based on theircollaboration intention. Task manager assigns tasks to suitable robots based on the principle of maximizing the gain of the system. Then, the robots which haveassigned a task form groups and recruit leisure robots into their groups to performthe task together. The experiment results validate this algorithm is real-time,efficient and can get better gain of the system.