Research On Multi-robot Cooperation Task Assignment Based On Group Affective Tone

Artificial emotion is an important symbol of robot intelligence level,and multi-robot cooperation task assignment is the core of multi-robot cooperation system(MRCS),researching the emotional robot cooperation task assignment in the emotional multi-robot cooperation system is a more complex job.At present,in the emotional multi-robot cooperation system,researchers mainly introduce the emotional state(such as emotion,personality,willingness,etc.)from the individual level to improve the robot's autonomous behavior decision-making ability,optimize the emotional robot cooperation task allocation and improve the group effectiveness.However,researchers lack the perspective of group level emotional state to investigate the influence of positive group affective tone(PGAT)on the group cooperation ability,and emotional robot cooperation task assignment and group effectiveness.This thesis mainly includes the following aspects:(1)Positive group affective tone modeling at group emotional level.According to the characteristics of emotional multi-robot cooperation system,this thesis proposes a relatively perfect positive group affective tone(PGAT)model based on group cooperation.The model has good interpretability and computability,besides,the model comprehensively analyzes the influence of individual level factors(personality,emotion,emotional decay)and group level factors(relationship coefficient between group members,group members' emotional expression ability and emotional reception ability,emotional contagion)on the positive group affective tone.The emotional multi-robot cooperative pursuit simulation experiment proves that by giving full play of positive group affective tone in task allocation,it can effectively reduce the risk of group disbanding due to group member's emotional decay,and help to improve pursuit group's cooperative ability and the system effectiveness.(2)Emotional multi-robot cooperative task assignment problem.In the emotional multi-robot cooperative system,in order to improve the group cooperation ability and the effectiveness of the system,this thesis proposes an emotional robot cooperation task allocation algorithm based on the improved bacterial foraging optimization algorithm.The algorithm improves the standard bacterial foraging optimization algorithm,for example,in the bacterial chemotaxis operation,simulated annealing algorithm is used to improve the ability of the algorithm to jump out of the local optimum;in the bacterial reproduction operation,PGAT model is used to improve the group cooperation ability;in the bacterial dispersion operation,self-organizing feature map algorithm clustering function is used to improve the algorithm convergence speed and global optimization ability.Finally,in the emotional multi-robot cooperative pursuit simulation experiment,the algorithm effectiveness of this thesis is verified by comparing with other algorithms.