论文部分内容阅读
提出了基于自组织映射(self organizing map,SOM)神经网络的自组织算法,把任务分配融入网络训练过程中.通过竞争获胜函数值在网络训练(任务分配)中的决定作用,并结合算法自身任务分配的实时性,由各情感机器人根据情感等因素提供竞争获胜函数值,并对值进行强化学习调整.这样使情感直接有效地参与任务分配决策,优化了算法性能.最后,通过仿真实验验证了本文所提出算法的有效性,特别是随着情感机器人团队规模扩大,追捕时间会比现有算法缩短一半以上.
This paper proposes a self-organizing algorithm based on self organizing map (SOM) neural network, which integrates task assignment into the network training process.By competing for the decisive role of the winning function value in network training (task assignment) and combining with the algorithm itself According to emotion and other factors, the emotional robots provide the competitive winning function value and adjust the value intensively, which makes the emotion directly and effectively participate in the task allocation decision, and optimizes the performance of the algorithm.Finally, through the simulation experiment, The effectiveness of the proposed algorithm in this paper, especially with the expansion of the emotional robot team size, hunt time will be shortened than the existing algorithm more than half.