Deep Deterministic Policy Gradient to Regulate Feedback Control Systems Using Reinforcement Learning

Al‐Bawri, Samir Salem; Islam, Mohammad Tariqul; Singh, Mandeep Jit; Jamlos, Mohd Faizal; Narbudowicz, Adam; Ammann, Max J.; Schreurs, Dominique

doi:10.32604/cmc.2022.021917

Cited by 4 publications

(1 citation statement)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The use of reinforcement learning method in the controller design can change the situation that traditional controller design relies too much on the precise mathematical model of the controlled object and can also solve the problem that traditional controller can be interfered and fails in the control process. Due to its unique characteristics, reinforcement learning has been applied in many fields, including reducing the difficulty of control in control [6], playing a role in task scheduling [7], and showing strong adaptability in image processing [8].…”

Section: Introductionmentioning

confidence: 99%

Generation and Simulation of Basic Maneuver Action Library for 6-DOF Aircraft by Reinforcement Learning

Wang¹,

Teng²,

He³

et al. 2022

Journal on Internet of Things

View full text Add to dashboard Cite

The development of modern air combat requires aircraft to have certain intelligent decision-making ability. In some of the existing solutions, the automatic control of aircraft is mostly composed of the upper mission decision and the lower control system. Although the underlying PID (Proportional Integral Derivative) based controller has a good performance in stable conditions, it lacks stability in complex environments. So, we need to design a new system for the problem of aircraft decision making. Studies have shown that the behavior of an aircraft can be viewed as a combination of several basic maneuvers. The establishment of aircraft basic motion library will effectively reduce the difficulty of upper aircraft control. Given the good performance of reinforcement learning to solve the problem with continuous action space, in this paper, reinforcement learning is used to control the aircraft's rod and rudder to generate a basic maneuver action library, and the flight of the aircraft under the 6 degrees of freedom (6-DOF) simulation engine is effectively controlled. The simulation results verify the feasibility of the method on a visual simulation platform.

show abstract

Section: Introductionmentioning

confidence: 99%

Generation and Simulation of Basic Maneuver Action Library for 6-DOF Aircraft by Reinforcement Learning

Wang¹,

Teng²,

He³

et al. 2022

Journal on Internet of Things

View full text Add to dashboard Cite

show abstract

Joint Power Allocation and User Fairness Optimization for Reinforcement Learning Over mmWave-NOMA Heterogeneous Networks

Sobhi-Givi,

Nouri,

Shayesteh

et al. 2024

IEEE Trans. Veh. Technol.

View full text Add to dashboard Cite

Improving the Survival Time of Multiagents in Social Dilemmas through Neurotransmitter-Based Deep Q-Learning Model of Emotions

Hassan

Shahid

Hayat

et al. 2022

Journal of Healthcare Engineering

View full text Add to dashboard Cite

In multiagent systems, social dilemmas often arise whenever there is a competition over the limited resources. The major challenge is to establish cooperation among intelligent virtual agents for solving the situations of social dilemmas. In humans, personality and emotions are the primary factors that lead them toward a cooperative environment. To make agents cooperate, they have to become more like humans, that is, believable. Therefore, we hypothesize that emotions according to the personality give birth to believability, and if believability is introduced into agents through emotions, it improves their survival rate in social dilemma situations. The existing researches have introduced different computational models to introduce emotions in virtual agents, but they lack emotions through neurotransmitters. We have proposed a neurotransmitters-based deep Q-learning computational model in multiagents that is a suitable choice for emotion modeling and, hence, believability. The proposed model regulates the agents’ emotions by controlling the virtual neurotransmitters (dopamine and oxytocin) according to the agent’s personality. The personality of the agent is introduced using OCEAN model. To evaluate the proposed system, we simulated a survival scenario with limited food resources in different experiments. These experiments vary the number of selfish agents (higher neuroticism personality trait) and the selfless agents (higher agreeableness personality trait). Experimental results show that by adding the selfless agents in the scenario, the agents develop cooperation, and their collective survival time increases. Thus, to resolve the social dilemma problems in virtual agents, we can make agents believable through the proposed neurotransmitter-based emotional model. This proposed work may help in developing nonplayer characters (NPCs) in games.

show abstract

Deep Deterministic Policy Gradient to Regulate Feedback Control Systems Using Reinforcement Learning

Cited by 4 publications

References 17 publications

Generation and Simulation of Basic Maneuver Action Library for 6-DOF Aircraft by Reinforcement Learning

Generation and Simulation of Basic Maneuver Action Library for 6-DOF Aircraft by Reinforcement Learning

Joint Power Allocation and User Fairness Optimization for Reinforcement Learning Over mmWave-NOMA Heterogeneous Networks

Improving the Survival Time of Multiagents in Social Dilemmas through Neurotransmitter-Based Deep Q-Learning Model of Emotions

Contact Info

Product

Resources

About