Dynamic Programming for Optimal Planning and Control of Redundant Robot Manipulators

Kinematically redundant robots possess more degrees of freedom than those strictly required to execute a given task. Such a charac- teristic gives the system a higher degree of dexterity and mobility that can be exploited to optimize performance indices of interest. Kinematic redundancy can be found in anthropomorphic serial chains, humanoids, mobile manipulators (terrestrial, aerial), par- allel robots and systems of cooperating robots. The optimization problem that looks into the exploitation of redundancy to optimize a given performance index is commonly referred to as redundancy resolution. Many robotic tasks require to follow a speci c path, such as in welding, cutting, gluing and in some assembly and disassembly tasks. This is opposed to point-to-point motion, that is rather typical in pick-and-place operations. The de nition of the speed with which the path is tracked is also a decision variable that, in most applications, is necessary to optimize. The criterion to follow could be related to time minimization, but also tracking accuracy and, more generally, quality measures de ned for the speci c process. The optimization problem that looks into the de nition of velocity along a given path is commonly referred to as trajectory planning or time-parametrization. Nowadays, many applications require the robots to live and oper- ate in unstructured, unknown and highly-dynamical environments, which undoubtedly shifts the attention of research towards real- time planning and control. However, there still are situations, that are quite frequent in reality, where o -line planning is required, while the underlying process is not fully optimized. Indeed, one of the most frequent employments of anthropomorphic arms in manufacturing industries still concerns the execution of repetitive tasks in structured environments, where the task is planned once and executed cyclically. In the aerospace sector, o -line planning instances exist in the mission design phase for feasibility and bud- get assessments. .. [edited by Author]