Automatically assigned DDC number: 00631

Manually assigned DDC number: 00631

Title: Reinforcement Learning for. . .

Subject: Reinforcement Learning for. . .

Description: This report proposes a method that handles these tasks without using a dynamical model. A simulation of a robot manipulator with two links is used to demonstrate the 1 ################ ### ######## ######### #### ## ##### ####### ## ###### ######### ########## ########### ########### ########### ####### ######### ### #### ### ###### # ################## ################# #### ### #### # ############ ############# ######## ### ############ ########### ####### ability of the learning controller in moving the end-eector of the manipulator from a start point to some desired destination. A signal which indicates when a collision with an obstacle or joint limit has occurred or when the destination has been reached is required. Joint position and velocity information from sensors are used as inputs to the system in our experiments, but the method can be modied to accept inputs from a range sensor or a vision system. The appropriate torques to each joint are obtained which not only enables the manipulator to reach its destination, but also avoid collisions with obstacles in the workspace. This report expands on the ndings described in [11] and is organized as follows. Related work are described briey in Section 2. Section 3 describes the simulation of the robot and the task in greater detail. Section 4 contains details about the implementation of CMACs in the experiments. Section 5 provides an overview of reinforcement learning and Section 6 describes the theory of associative stochastic learning automata { the learning algorithm we employed is based on the theory explained in these two sections. Section 7 presents the reinforcement schedule used to facilitate learning of useful behaviours. Section 8 contains exp...

Contributor: The Pennsylvania State University CiteSeer Archives

Publisher: unknown

Date: 1992-07-24

Pubyear: 1992

Format: ps

Identifier: http://citeseer.ist.psu.edu/140107.html

Source: ftp://svr-ftp.eng.cam.ac.uk/pub/reports/tham_tr104.ps.Z

Language: en

Rights: unrestricted

Graph

<?xml   version="1.0"   encoding="UTF-8"?>

<references_metadata>

      <rec   ID="SELF"   Type="SELF"   CiteSeer_Book="SELF"   CiteSeer_Volume="SELF"   Title="Reinforcement   Learning   for.   .   .">

            <identifier   Org="ISBN:0262112450"   Paper_ID="SELF"   Extracted="0262112450"   />

            <identifier   Org="ISBN:0387950664"   Paper_ID="SELF"   Extracted="0387950664"   DDC="621.36/7"   Normalized_DDC="621367"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:0444899316"   Paper_ID="SELF"   Extracted="0444899316"   DDC="152.3"   Normalized_DDC="1523"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:047166054X"   Paper_ID="SELF"   Extracted="047166054X"   DDC="519.7/03"   Normalized_DDC="519703"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:0792394070"   Paper_ID="SELF"   Extracted="0792394070"   DDC="006.3/1"   Normalized_DDC="00631"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:1586030787"   Paper_ID="SELF"   Extracted="1586030787"   DDC="629.8/9263"   Normalized_DDC="62989263"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:1586032399"   Paper_ID="SELF"   Extracted="1586032399"   DDC="629.8/9263"   Normalized_DDC="62989263"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:3211836519"   Paper_ID="SELF"   Extracted="3211836519"   />

            <identifier   Org="ISBN:3540206469"   Paper_ID="SELF"   Extracted="3540206469"   DDC="006.3"   Normalized_DDC="0063"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:3540222820"   Paper_ID="SELF"   Extracted="3540222820"   DDC="006.31"   Normalized_DDC="00631"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:3540252606"   Paper_ID="SELF"   Extracted="3540252606"   DDC="006.3"   Normalized_DDC="0063"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:3540410562"   Paper_ID="SELF"   Extracted="3540410562"   DDC="004/.35"   Normalized_DDC="00435"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:3540413855"   Paper_ID="SELF"   Extracted="3540413855"   DDC="510   s"   Normalized_DDC="51"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:3540421858"   Paper_ID="SELF"   Extracted="3540421858"   DDC="629.8/92"   Normalized_DDC="629892"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:3540427325"   Paper_ID="SELF"   Extracted="3540427325"   DDC="006.3"   Normalized_DDC="0063"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:3540440747"   Paper_ID="SELF"   Extracted="3540440747"   DDC="006.3/2"   Normalized_DDC="00632"   Normalized_Weight="0.06666666666666667"   />

            <identifier   Org="ISBN:3540680020"   Paper_ID="SELF"   Extracted="3540680020"   />

            <identifier   Org="ISBN:3540691545"   Paper_ID="SELF"   Extracted="3540691545"   />

            <identifier   Org="ISBN:3540740236"   Paper_ID="SELF"   Extracted="3540740236"   />

            <identifier   Org="ISBN:3540878041"   Paper_ID="SELF"   Extracted="3540878041"   DDC="006.3"   Normalized_DDC="0063"   Normalized_Weight="0.06666666666666667"   />

      </rec>

</references_metadata>

www.000webhost.com