论文部分内容阅读
该文提出了一种模拟人类学习与进化过程的机器人智能体系结构,微小机器人利用设计人员事先设计的机器人基本行为,根据实际环境和具体任务要求,采用增强学习方式,通过群体行为进化,自主创建满足任务要求和适应环境的具体动作。克服了设计人员在采用基于符号的传统人工智能方法时,由于对外部环境和任务的认识不足而造成的局限性,使机器人的行为动作更适合环境和任务要求。
This paper presents a robot intelligent architecture that simulates the process of human learning and evolution. The micro-robots utilize the basic behaviors of robots designed by designers in advance. Based on the actual environment and specific tasks, the micro-robots adopt the method of enhanced learning, evolve through group behavior, Meet the mission requirements and adapt to the environment of the specific actions. Which overcomes the limitations caused by the lack of understanding of the external environment and tasks when designers use the traditional symbol-based artificial intelligence method to make the robot’s behavior more suitable for the environment and the task requirements.