【摘要】 针对一种新型仿生机器鼠平台的运动规划问题主要包括跟踪、避障等问题,依据仿生机器鼠的双目视觉系统以及多自由度的机械结构,根据驱动电机的特性,建立仿生机器鼠的扭腰、抬身、攀爬等动力学模型,采用"记忆"目标鼠的特征、"学习"最优避障路径、"推断"目标鼠的运动方向等方法,实现了仿生机器鼠的跟踪与避障行为,提出了一种基于仿生学原理的仿生机器鼠的跟踪与避障策略。在一种新型仿生机器鼠平台上的实验结果表明,相比与基于虚拟阻抗模型的控制方法,该跟踪与避障策略能够更好地实现仿生机器鼠的跟踪与避障行为,从而验证了策略的可行性与有效性。更多还原

【Abstract】 This strategy mainly focuses on the motion planning problems such as tracking and obstacle avoidance method for bionic intelligent rat platform.According to the binocular vision system and multiple degrees of freedom of the mechanical structure,the dynamic model of the bionic intelligent rat(twisting waist,rising body,and climbing etc.) was based on the characteristics of driving motors.By using the method of remembering the feature of the target intelligent rat,learning the optimal obstacle avoidance path,inferring the movement direction of the target rat,the bionic intelligent rat can track the target rat and avoid the obstacle.Accordingly,this strategy proposed a tracking and obstacle avoidance method that is based on the principle of bionics.A new kind of bionic intelligent rat is used for the motion experiments.The results demonstrate that compared with the virtual impedance model,the proposed motion algorithms can realize tracking and obstacle avoidance of the intelligent rat,and thus show the feasibility and validity of the proposed algorithms.更多还原