Control of link-elastic serial kinematic chains with multiple bending planes for safe, dependable and efficient physical human-machine-interaction

拟议项目的目的是在振动阻尼，定位和力调节的运动链与链接弹性出现在多个平面的机器人机构的控制概念的发展。在未来的小批量生产链中，制造专家被设想成为一个机器人管理员，负责整个机器人群。通过他对制造过程的了解，专家指导机器人完成特定的制造任务。当牛群中的每一个机器人都在处理自己的任务时，机器人管理员可以自由地指示其他机器人完成新的任务。人口结构的变化激发了灵活和物理交互式生产助理在家庭环境中的通用服务机器人的概念扩展。这些人工管家使我们从日常琐事中解脱出来。它们使我们能够保持独立和自决，直到老年。实现的关键是确保人与机器人之间物理交互的本质安全性和可靠性。这意味着从传统的高精度，但僵硬和大规模的机器人设计转向所谓的软机器人方法的应用程序的范式转变。首先，这种方法努力减少机器人的整体质量。在下一步中，被动弹性元件被有意地引入到与致动器串联的结构中。不期望的结构振动和挠度必须由新的控制器补偿。现有的原型仅在机器人关节中实现被动弹性。这里，弹性效应围绕关节致动轴线配置。新型复合材料或大量的机械构造努力被认为是为了确保轻质但刚性的连杆机构。如果连杆体是弹性的，则弹性特性沿着连杆结构分布，使得不期望的振荡和偏转的衰减成为更困难的控制挑战。鲁棒控制概念可以放松对机器人和类似机械的连杆刚性的严格要求。因此，可以用更简单的连杆几何形状、传统的材料和更薄的壁厚来实现整个机构质量的减少。这就产生了更轻、更高效、更实惠的机器人。可以采用更小和更小功率的致动器，这进一步提高了关于物理人机交互的安全性和可靠性。这就是为什么提议者看到了一个增长的需求，开发新的控制概念的链接弹性多体系统的软机器人设计范式。文献中报道的少数实验研究仅限于弹性仅存在于单个平面中。随着调查，特别是集中在连杆弹性在多个平面的机制，因此，拟议的项目扩展了科学的状态。

The proposed project aims at the development of control concepts for the oscillation damping, positioning and force regulation of kinematic chains with link elasticity emerging in multiple planes of a robot mechanism. In future small batch production chains, the manufacturing expert is envisioned to become a herdsman, who is responsible for an entire herd of robots. With his knowledge about the manufacturing process, the expert instructs the robots for a particular manufacturing task. While each and every robot of the herd is processing its individual task, the herdsman is free to instruct other robots for new tasks. The demographic change motivates the conceptual extension of a flexible and physically interactive production assistant to universal service robots in household environments. Such artificial butlers disburden us from tedious tasks of the daily grind. They enable us to retain independence and self-determination up to an old age. The key to the realization is to ensure intrinsic safety and dependability of the physical interaction between humans and robots. This implicates a paradigm shift from conventional highly accurate but stiff and massive robot designs towards the application of so called Soft-Robotics methods. In first place, such methods strive to reduce the overall robot mass. In the next step, passive elastic components are intentionally introduced into the structure in series with the actuators. Undesired structural oscillations and deflections must be compensated by novel controllers. Existing prototypes realize the passive elasticity solely in the robot joints. Here the elastic effects are collocated about the joint axis of actuation. Novel composite materials or substantial mechanical construction efforts are considered to ensure lightweight but stiff link bodies. If the link bodies are elastic, the elastic properties are distributed along the link structure, so that the attenuation of undesired oscillations and deflections becomes a more difficult control challenge. Robust control concepts can relax the strict requirement for rigidity in the links of robots and comparable machinery. The reduction of the overall mechanism mass can hence be accomplished with simpler link geometries, conventional materials and thinner wall thicknesses. This yields a lighter and more efficient and affordable class of robots. Smaller and less powerful actuators can be employed, which further improves safety and dependability with respect to physical human robot interaction. This is the reason why the proposer sees an increased demand for the development of novel control concepts for link elastic multi body systems following the Soft-Robotics design paradigm. The few experimental studies reported in literature are limited to the elasticity being present in just a single plane. With the investigations particularly focusing on link elasticity in multiple planes of the mechanism, the proposed project consequently extends the state of science.