Reconstructing Null-space Policies Subject to Dynamic Task Constraints in Redundant Manipulators

نویسندگان

  • Matthew Howard
  • Sethu Vijayakumar
چکیده

We consider the problem of direct policy learning in situations where the policies are only observable through their projections into the null-space of a set of dynamic, non-linear task constraints. We tackle the issue of deriving consistent data for the learning of such policies and make two contributions towards its solution. Firstly, we derive the conditions required to exactly reconstruct null-space policies and suggest a learning strategy based on this derivation. Secondly, we consider the case that the null-space policy is conservative and show that such a policy can be learnt more easily and robustly by learning the underlying potential function and using this as our representation of the policy.

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

منابع مشابه

Dynamics of Internal Motion of Redundant Manipulators

In the paper the dynamic properties of the internal motion of redundant manipulators are considered. The motion is decoupled into the end-effector motion and the internal motion. Next, the dynamic model of a redundant manipulator is derived. A special attention is given to the inertial properties of the system in the space where internal motion is taking place; we define a null space effective ...

متن کامل

Maximum Allowable Load On Wheeled Mobile Manipulators (RESEARCH NOTE)

This paper develops a computational technique for finding the maximum allowable load of mobile manipulators for a given trajectory. The maximum allowable loads which can be achieved by a mobile manipulator during a given trajectory are limited by the number of factors; probably the dynamic properties of mobile base and mounted manipulator, their actuator limitations and additional constraints a...

متن کامل

Trajectory Optimization of Cable Parallel Manipulators in Point-to-Point Motion

Planning robot trajectory is a complex task that plays a significant role in design and application of robots in task space. The problem is formulated as a trajectory optimization problem which is fundamentally a constrained nonlinear optimization problem. Open-loop optimal control method is proposed as an approach for trajectory optimization of cable parallel manipulator for a given two-end-po...

متن کامل

Planning and Control of Two-Link Rigid Flexible Manipulators in Dynamic Object Manipulation Missions

This research focuses on proposing an optimal trajectory planning and control method of two link rigid-flexible manipulators (TLRFM) for Dynamic Object Manipulation (DOM) missions. For the first time, achievement of DOM task using a rotating one flexible link robot was taken into account in [20]. The authors do not aim to contribute on either trajectory tracking or vibration control of the End-...

متن کامل

A Comparison of Control Algorithms for Redundant Manipulators from Viewpoint of Collision Forces

In the paper some control algorithms for redundant manipulators are compared from the viewpoint of the behaviour of the system during the collision with an obstacle. The emphasis is given to the contact forces and their minimization, and how they can utilize the self-motion caused by the contact forces to avoid the obstacles and to minimize the tracking errors in the task space. A controller wh...

متن کامل

ذخیره در منابع من


  با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید

برای دانلود متن کامل این مقاله و بیش از 32 میلیون مقاله دیگر ابتدا ثبت نام کنید

ثبت نام

اگر عضو سایت هستید لطفا وارد حساب کاربری خود شوید

عنوان ژورنال:

دوره   شماره 

صفحات  -

تاریخ انتشار 2007