Neural Joint Space Implicit Signed Distance Functions for Reactive Robot Manipulator Control
نویسندگان
چکیده
In this letter, we present an approach for learning a neural implicit signed distance function expressed in joint space coordinates, that efficiently computes distance-to-collisions arbitrary robotic manipulator configurations. Computing such distances is long standing problem robotics as approximate representations of the robot and environment geometry can lead to overly conservative constraints, numerical instabilities expensive computations – limiting real-time reactive control task success. Leveraging GPU parallelization differentiable nature proposed allows fast calculation gradients with respect network inputs, providing continuous repulsive vector field directly space. We show learned high-resolution collision representation be used achieve by i) formulating it collision-avoidance constraint quadratic programming (QP) inverse kinematics (IK), ii) introducing cost sampling-based model predictive controller (MPC). For reaching benchmark 7DoF dynamic obstacles intentionally obstructing robot's path average 250 Hz frequency QP-IK 92 MPC, showing accelerated performance 15% 40% MPC over baseline computation techniques.
منابع مشابه
Tracking control based on neural network strategy for robot manipulator
This study presents a sliding-mode neural-network (SMNN) control system for the tracking control of an n rigid-link robot manipulator to achieve high-precision position control. The aim of this study is to overcome some of the shortcomings of conventional robust controllers such as a model-based adaptive controller requires the system dynamics in detail; the fuzzy rule learning scheme has a lat...
متن کاملModel-Independent Control of a Flexible-Joint Robot Manipulator
Flexibility at the joint of a manipulator is an intrinsic property. Even “rigid-joint” robots, in fact, possess a certain amount of flexibility. Previous experiments confirmed that joint flexibility should be explicitly included in the model when designing a high-performance controller for a manipulator because the flexibility, if not dealt with, can excite system natural frequencies and cause ...
متن کاملNeural network impedance force control of robot manipulator
Performance of impedance controller for robot force tracking is aaected by the uncertainties in both the robot dynamic model and environment stiiness. The purpose of this paper is to improve the controller ro-bustness by applying the neural network(NN) technique to compensate for the uncertainties in the robot model. NN control techniques are applied to two impedance control methods : torque-ba...
متن کاملadaptive control of two-link robot manipulator based on the feedback linearization method and the proposed neural network
This paper proposes an adaptive control method based on the feedback linearization technique and a proposed neural network, for tracking and position control of an industrial manipulator. At first, it is assumed that the dynamics of the system are known and the control signal is constructed by the feedback linearization method. Then to eliminate the effects of the uncertainties and external d...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: IEEE robotics and automation letters
سال: 2023
ISSN: ['2377-3766']
DOI: https://doi.org/10.1109/lra.2022.3227860