Abstract:A self-organizing algorithm based on self-organizing map (SOM) neural network is proposed and task allocation is integrated into the training process of the network. Taking advantages of the decisive effect of the competitive winning function value in network training (task allocation) and combining with the real-time performance of task allocation in the proposed algorithm, the emotional robot provides the competitive victory function value, which is adjusted by reinforcement-learning, according to the emotional factors. In this way, the emotion directly and effectively participates in task allocation decisions, and the algorithm performance is optimized. Finally, simulation experiments are carried out to verify the effectiveness of the proposed algorithm. Especially, with the scale of emotional robot team growing, the pursuit time can be shortened by more than 50% compared with the existing algorithm.
[1] Oh K G, Jang M S, Kim S J. Automatic emotional expression of a face robot by using a reactive behavior decision model[J]. Journal of Mechanical Science and Technology, 2010, 24(3):769-774.
[2] Murphy R R, Lisetti C L, Tardif R, et al. Emotion-based control of cooperating heterogeneous mobile robots[J]. IEEE Transactions on Robotics and Automation, 2002, 18(5):744-757.
[3] Fang B F, Chen L, Wang H, et al. Research on multirobot pursuit task allocation algorithm based on emotional cooperation factor[J]. The Scientific World Journal, 2014, 2014:No.864180.
[4] Watada S, Obayashi M, Kurenoto T, et al. A new decision-making system of an agent based on emotional models in multi-agent system[C]//18th International Symposium on Artificial Life and Robotics. Oita, Japan:Alife Robotics Co, Ltd, 2013:452-455.
[5] 谷学静,王志良,魏哲华,等.基于人工心理理论的情感模型构建方法研究[J].微计算机信息,2006,22(5):264-266.Gu X J, Wang Z L, Wei Z H, et al. The research of constructing affective model based on artificial psychology theoretics[J]. Microcomputer Information, 2006, 22(5):264-266.
[6] Yuan W, Zeng B. Multi-robot task allocation using compound emotion algorithm[C]//7th International Symposium on Advanced Parallel Processing Technologies. Berlin, Germany:Springer-Verlag, 2007:545-550.
[7] Banik S C, Watanabe K, Izumi K. Task allocation with a cooperative plan for an emotionally intelligent system of multi-robots[C]//SICE Annual Conference. Piscataway, USA:IEEE, 2007:1004-1010.
[8] Kohonen T. Self-organizing maps:Optimization approaches[C]//1991 International Conference on Artificial Neural Networks. Amsterdam, Netherlands:Elsevier, 1991:981-990.
[9] Digman J M. Personality structure:Emergence of the 5-factor model[J]. Annual Review of Psychology, 1990, 41(1):417-440.
[10] Wang G J, Teng S D, Fu K C. Simulating emotion and personality for intelligent agent[C]//International Conference on Intelligent Computation Technology and Automation. Piscataway, USA:IEEE, 2010:304-308.
[11] 王浩,张权益,方宝富,等.基于状态空间与概率空间映射的极大相似度匹配情感模型[J].模式识别与人工智能,2013,26(6):552-560.Wang H, Zhang Q Y, Fang B F, et al. Maximum similarity matching emotion model based on mapping between state space and probability space[J]. Pattern Recognition and Artificial Intelligence, 2013, 26(6):552-560.
[12] Wang S C, Song Z X, Ding H, et al. An improved reinforcement Q-learning method with BP neural networks in robot soccer[C]//4th International Symposium on Computational Intelligence and Design. Piscataway, USA:IEEE, 2011:177-180.
[13] Jiang Y L. Task allocation of multiple robotic fishes based on self-organizing map neural network[J]. Applied Mechanics and Materials, 2014, 563:308-311.
[14] 朱大奇,曹翔.多个水下机器人动态任务分配和路径规划的信度自组织算法[J].控制理论与应用,2015,32(6):762-769.Zhu D Q, Cao X. An improved self-organizing map method for multiple autonomous underwater vehicle teams in dynamic task assignment and path planning[J]. Control Theory and Applications, 2015, 32(6):762-769.
[15] Ma Z J, Cui Y Y. Optimal hierarchical allocation in deregulated electricity market under PSP auction mechanism[C]//26th Chinese Control and Decision Conference. Piscataway, USA:IEEE, 2014:2256-2261.