scholarly journals Indoor Point-to-Point Navigation with Deep Reinforcement Learning and Ultra-Wideband

2021 ◽  
Author(s):  
Enrico Sutera ◽  
Vittorio Mazzia ◽  
Francesco Salvetti ◽  
Giovanni Fantin ◽  
Marcello Chiaberge
Keyword(s):  
Reinforcement Learning ◽  
Ultra Wideband ◽  
Point To Point

scholarly journals Zermelo’s problem: Optimal point-to-point navigation in 2D turbulent flows using reinforcement learning

2019 ◽  
Vol 29 (10) ◽  
pp. 103138 ◽  
Author(s):  
L. Biferale ◽  
F. Bonaccorso ◽  
M. Buzzicotti ◽  
P. Clark Di Leoni ◽  
K. Gustavsson
Keyword(s):  
Reinforcement Learning ◽  
Turbulent Flows ◽  
Optimal Point ◽  
Point To Point

scholarly journals Pick and Place Operations in Logistics Using a Mobile Manipulator Controlled with Deep Reinforcement Learning

2019 ◽  
Vol 9 (2) ◽  
pp. 348 ◽  
Author(s):  
Ander Iriondo ◽  
Elena Lazkano ◽  
Loreto Susperregi ◽  
Julen Urain ◽  
Ane Fernandez ◽  
...  
Keyword(s):  
Reinforcement Learning ◽  
Path Planning ◽  
Mobile Manipulator ◽  
Complex Tasks ◽  
Complex Control ◽  
Planning And Control ◽  
Pick And Place ◽  
Policy Gradient ◽  
Point To Point ◽  
And Control

Programming robots to perform complex tasks is a very expensive job. Traditional path planning and control are able to generate point to point collision free trajectories, but when the tasks to be performed are complex, traditional planning and control become complex tasks. This study focused on robotic operations in logistics, specifically, on picking objects in unstructured areas using a mobile manipulator configuration. The mobile manipulator has to be able to place its base in a correct place so the arm is able to plan a trajectory up to an object in a table. A deep reinforcement learning (DRL) approach was selected to solve this type of complex control tasks. Using the arm planner’s feedback, a controller for the robot base is learned, which guides the platform to such a place where the arm is able to plan a trajectory up to the object. In addition the performance of two DRL algorithms ((Deep Deterministic Policy Gradient (DDPG)) and (Proximal Policy Optimisation (PPO)) is compared within the context of a concrete robotic task.

scholarly journals The Control Method of Twin Delayed Deep Deterministic Policy Gradient with Rebirth Mechanism to Multi-DOF Manipulator

Electronics ◽  
2021 ◽  
Vol 10 (7) ◽  
pp. 870
Author(s):  
Yangyang Hou ◽  
Huajie Hong ◽  
Zhaomei Sun ◽  
Dasheng Xu ◽  
Zhe Zeng
Keyword(s):  
Reinforcement Learning ◽  
Control Method ◽  
Kinematic Model ◽  
Degree Of Freedom ◽  
Learning Ability ◽  
Reward Function ◽  
Policy Gradient ◽  
Evaluation Algorithm ◽  
Improved Learning ◽  
Point To Point

As a research hotspot in the field of artificial intelligence, the application of deep reinforcement learning to the learning of the motion ability of a manipulator can help to improve the learning of the motion ability of a manipulator without a kinematic model. To suppress the overestimation bias of values in Deep Deterministic Policy Gradient (DDPG) networks, the Twin Delayed Deep Deterministic Policy Gradient (TD3) was proposed. This paper further suppresses the overestimation bias of values for multi-degree of freedom (DOF) manipulator learning based on deep reinforcement learning. Twin Delayed Deep Deterministic Policy Gradient with Rebirth Mechanism (RTD3) was proposed. The experimental results show that RTD3 applied to multi degree freedom manipulators is in place, with an improved learning ability by 29.15% on the basis of TD3. In this paper, a step-by-step reward function is proposed specifically for the learning and innovation of the multi degree of freedom manipulator’s motion ability. The view of continuous decision-making and process problem is used to guide the learning of the manipulator, and the learning efficiency is improved by optimizing the playback of experience. In order to measure the point-to-point position motion ability of a manipulator, a new evaluation index based on the characteristics of the continuous decision process problem, energy efficiency distance, is presented in this paper, which can evaluate the learning quality of the manipulator motion ability by a more comprehensive and fair evaluation algorithm.

Finding the Right Problems: Some HREM Applications to Interfaces

1984 ◽  
Vol 42 ◽  
pp. 376-379
Author(s):  
D. Cherns
Keyword(s):  
Grain Boundaries ◽  
High Resolution Electron Microscopy ◽  
Boundary Structure ◽  
Atomic Structures ◽  
Grain Boundary Structure ◽  
Resolution Electron ◽  
Point To Point ◽  
The Right ◽  
New Generation ◽  
Better Than

The use of high resolution electron microscopy (HREM) to determine the atomic structure of grain boundaries and interfaces is a topic of great current interest. Grain boundary structure has been considered for many years as central to an understanding of the mechanical and transport properties of materials. Some more recent attention has focussed on the atomic structures of metalsemiconductor interfaces which are believed to control electrical properties of contacts. The atomic structures of interfaces in semiconductor or metal multilayers is an area of growing interest for understanding the unusual electrical or mechanical properties which these new materials possess. However, although the point-to-point resolutions of currently available HREMs, ∼2-3Å, appear sufficient to solve many of these problems, few atomic models of grain boundaries and interfaces have been derived. Moreover, with a new generation of 300-400kV instruments promising resolutions in the 1.6-2.0 Å range, and resolutions better than 1.5Å expected from specialist instruments, it is an appropriate time to consider the usefulness of HREM for interface studies.

An x-ray microprobe based upon a close-coupled focal-spot / glass capillary arrangement

1992 ◽  
Vol 50 (2) ◽  
pp. 1758-1759
Author(s):  
D. A. Carpenter ◽  
M. A. Taylor
Keyword(s):  
Imaging Techniques ◽  
Fluorescence Analysis ◽  
High Sensitivity ◽  
Specimen Preparation ◽  
X Rays ◽  
Glass Capillary ◽  
Close Coupling ◽  
X Ray ◽  
Point To Point ◽  
Beam Damage

The development of intense sources of x rays has led to renewed interest in the use of microbeams of x rays in x-ray fluorescence analysis. Sparks pointed out that the use of x rays as a probe offered the advantages of high sensitivity, low detection limits, low beam damage, and large penetration depths with minimal specimen preparation or perturbation. In addition, the option of air operation provided special advantages for examination of hydrated systems or for nondestructive microanalysis of large specimens.The disadvantages of synchrotron sources prompted the development of laboratory-based instrumentation with various schemes to maximize the beam flux while maintaining small point-to-point resolution. Nichols and Ryon developed a microprobe using a rotating anode source and a modified microdiffractometer. Cross and Wherry showed that by close-coupling the x-ray source, specimen, and detector, good intensities could be obtained for beam sizes between 30 and 100μm. More importantly, both groups combined specimen scanning with modern imaging techniques for rapid element mapping.

Atomic structure imaging of the Si/SiO2 interface with high-resolution electron microscopy

1986 ◽  
Vol 44 ◽  
pp. 390-391
Author(s):  
J.L. Batstone ◽  
J.M. Gibson ◽  
Alice.E. White ◽  
K.T. Short
Keyword(s):  
Electron Microscopy ◽  
High Resolution ◽  
Atomic Structure ◽  
High Resolution Electron Microscopy ◽  
Medium Voltage ◽  
Voltage Electron ◽  
Resolution Electron ◽  
Structural Image ◽  
Electron Microscopes ◽  
Point To Point

High resolution electron microscopy (HREM) is a powerful tool for the determination of interface atomic structure. With the previous generation of HREM's of point-to-point resolution (rpp) >2.5Å, imaging of semiconductors in only <110> directions was possible. Useful imaging of other important zone axes became available with the advent of high voltage, high resolution microscopes with rpp <1.8Å, leading to a study of the NiSi2 interface. More recently, it was shown that images in <100>, <111> and <112> directions are easily obtainable from Si in the new medium voltage electron microscopes. We report here the examination of the important Si/Si02 interface with the use of a JEOL 4000EX HREM with rpp <1.8Å, in a <100> orientation. This represents a true structural image of this interface.

Sign in / Sign up

Export Citation Format

Share Document