Robot Cognitive Model And Algorithm With Emotional Mechanism

Different disciplines such as brain science,cognitive neuroscience,and cognitive psychology have demonstrated that emotions play a key role in human learning and understanding of the environment and can effectively improve learning efficiency.The interaction and relationship between emotion and cognition has been studied from many disciplines.In traditional cognitive robot research,there are problems such as low learning efficiency and large demand for training samples,therefore,in the process of robot cognition,the introduction of emotional factors to improve the learning efficiency of robots is the main goal of this paper.In this regard,this paper combines artificial emotion modeling theory,fuzzy control,and reinforcement learning methods to study the cognitive model and algorithm of robots with affective mechanism,and the main work content is as follows:(1)Combining artificial emotion and memory mechanism,a cognitive model of maze robot with emotion and memory mechanism is proposed.Firstly,based on the one-dimensional emotion model theory,artificial emotion modeling is carried out to construct the emotion generation system of the model;then,based on the theoretical framework of reinforcement learning,the internal reward generated by the emotional system is fused with the memory to form a new reward mechanism,and the corresponding cognitive learning algorithm is designed;Finally,the experimental with maze robot search tasks that require "energy replenishment" results show that the proposed cognitive model can effectively improve the search efficiency and learning stability of maze robots.(2)Based on the maze robot cognitive model with emotion and memory mechanism,a robot cognitive model integrating perceptual and rational behavior decision-making mechanism is proposed.Firstly,based on OCC(Ortony,Clore,Collins)cognitive evaluation emotion theory and PAD(Pleasure-Arousal-Dominance)emotion model theory,personality factors are introduced to allow robots to generate simple emotions,and complex emotions are generated by combining fuzzy control theory;after that,the robot behavior decision-making system is designed by integrating emotional and rational factors,so as to realize the autonomous switching between robot perceptual decision-making and rational decision-making;finally,experiments are carried out in a more complex "multi-trap" maze environment,and the experimental results show that the cognitive model under the fusion decision-making mechanism has better risk resistance and faster learning efficiency.Through the principle of simulated human learning and decision-making,this paper introduces emotional factors into the cognitive process of robots,which provides a new direction for the research of robot cognition;the study of artificial emotion modeling provides a new idea for machine emotion generation;the designed cognitive model and learning algorithm can promote the development of cognitive robots and emotional robots.