Real and Artificial Forces in the Control of Manipulators: Theory and Experiments Export

@phdthesis{vol90, abstract = {This research has explored the problems of moving through an environment while avoiding obstacles, approaching and stably contacting a surface in that environment, and applying force to the surface. To this end, a theoretical and experimental analysis of the following topics has been performed: explicit force and impedance control, impact control, and obstacle avoidance and approach with artificial potentials. The issue of force control has been addressed in two ways. First, a dynamic model of the manipulator / sensor / environment system has been theoretically developed and experimentally verified. With this model it has been possible to perform the second step of a detailed analysis of previously proposed force control schemes. Moreover, a theoretical framework has been developed which encompasses most control schemes proposed to date, including impedance control and explicit force control. This framework has provided a means by which to compare the stability properties of the various schemes. This theoretical analysis has been supported by experimental implementation and analysis on the CMU DDarm II. The theoretical framework developed for force control, also yielded new insight into the problem of transition from motion through the environment to contact with it. During this transition there is impact. The developed force control framework revealed a new method of effectively controlling the impact phase and providing stable transition. This scheme has also been successfully implemented on the CMU DDarm II. To move through an obstructed environment (possibly with moving obstacles) a local obstacle avoidance scheme based on superquadric artificial potentials has been developed. This potential formulation blends the best features of previous ones by eliminating local minima for simple environments, while not unnecessarily removing some parts of workspace. Two forms of the potential energy function provide either object avoidance or approach capability. This scheme has been implemented on the CMU DDarm II and has shown successful avoidance of multiple obstacles in real-time.}, author = {Volpe, R.}, citeulike-article-id = {1244595}, keywords = {commands, haptics, marsokhod, microcomputer}, posted-at = {2007-09-20 17:32:10}, priority = {2}, school = {Carnegie Mellon University}, title = {Real and Artificial Forces in the Control of Manipulators: Theory and Experiments}, year = {1990} }

TY - THES ID - vol90 L3 - citeulike-article-id:1244595 N2 - This research has explored the problems of moving through an environment while avoiding obstacles, approaching and stably contacting a surface in that environment, and applying force to the surface. To this end, a theoretical and experimental analysis of the following topics has been performed: explicit force and impedance control, impact control, and obstacle avoidance and approach with artificial potentials. The issue of force control has been addressed in two ways. First, a dynamic model of the manipulator / sensor / environment system has been theoretically developed and experimentally verified. With this model it has been possible to perform the second step of a detailed analysis of previously proposed force control schemes. Moreover, a theoretical framework has been developed which encompasses most control schemes proposed to date, including impedance control and explicit force control. This framework has provided a means by which to compare the stability properties of the various schemes. This theoretical analysis has been supported by experimental implementation and analysis on the CMU DDarm II. The theoretical framework developed for force control, also yielded new insight into the problem of transition from motion through the environment to contact with it. During this transition there is impact. The developed force control framework revealed a new method of effectively controlling the impact phase and providing stable transition. This scheme has also been successfully implemented on the CMU DDarm II. To move through an obstructed environment (possibly with moving obstacles) a local obstacle avoidance scheme based on superquadric artificial potentials has been developed. This potential formulation blends the best features of previous ones by eliminating local minima for simple environments, while not unnecessarily removing some parts of workspace. Two forms of the potential energy function provide either object avoidance or approach capability. This scheme has been implemented on the CMU DDarm II and has shown successful avoidance of multiple obstacles in real-time. TI - Real and Artificial Forces in the Control of Manipulators: Theory and Experiments PB - Carnegie Mellon University KW - commands KW - haptics KW - marsokhod KW - microcomputer AU - Volpe, R PY - 1990/// ER -