Air Spring Controlled by Reinforcement Learning Algorithm

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F46747885%3A24210%2F20%3A00008299" target="_blank" >RIV/46747885:24210/20:00008299 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.engmech.cz/im/im/page/proc" target="_blank" >https://www.engmech.cz/im/im/page/proc</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.21495/5896-3-428" target="_blank" >10.21495/5896-3-428</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Air Spring Controlled by Reinforcement Learning Algorithm

Popis výsledku v původním jazyce

The paper deals with the replacement of the analog PID stroke controller of a bellows pneumatic spring, by machine learning algorithms, specifically deep reinforcement learning. The Deep Deterministic Policy Gradient (DDPG) algorithm used consists of an environment, in this case a pneumatic spring, and an agent which, based on observations of environment, performs actions that lead to the cumulative reward it seeks to maximize. DDPG falls into the category of actor-critic algorithms. It combines the benefits of Q-learning and optimization of a deterministic strategy. Q-learning is represented here in the form of critic, while optimization of strategy is represented in the form of an actor that directly maps the state of the environment to actions. Both the critic and the actor are represented in deep reinforcement learning by deep neural networks. Both of these networks have a target variant of themselves. These target networks are designed to increase the stability and speed of the learning process. The DDPG algorithm also uses a replay buffer, from which the data from which the agent learns is taken in batches.

Název v anglickém jazyce

Air Spring Controlled by Reinforcement Learning Algorithm

Popis výsledku anglicky

The paper deals with the replacement of the analog PID stroke controller of a bellows pneumatic spring, by machine learning algorithms, specifically deep reinforcement learning. The Deep Deterministic Policy Gradient (DDPG) algorithm used consists of an environment, in this case a pneumatic spring, and an agent which, based on observations of environment, performs actions that lead to the cumulative reward it seeks to maximize. DDPG falls into the category of actor-critic algorithms. It combines the benefits of Q-learning and optimization of a deterministic strategy. Q-learning is represented here in the form of critic, while optimization of strategy is represented in the form of an actor that directly maps the state of the environment to actions. Both the critic and the actor are represented in deep reinforcement learning by deep neural networks. Both of these networks have a target variant of themselves. These target networks are designed to increase the stability and speed of the learning process. The DDPG algorithm also uses a replay buffer, from which the data from which the agent learns is taken in batches.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

21100 - Other engineering and technologies

Projekt

—

Návaznosti

S - Specificky vyzkum na vysokych skolach

Rok uplatnění

2020

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Engineering Mechanics 2020

ISBN

978-80-214-5896-3

ISSN

1805-8248

e-ISSN

—

Počet stran výsledku

4

Strana od-do

428-431

Název nakladatele

Brno University of Technology Institute of Solid Mechanics, Mechatronics and Biomechanics

Místo vydání

Brno

Místo konání akce

Brno

Datum konání akce

1. 1. 2020

Typ akce podle státní příslušnosti

EUR - Evropská akce

Kód UT WoS článku

000667956100099