Control of the Cart-Pole System: Model-based vs. Model-free Learning

Fu, Minyue

10.1016/j.ifacol.2023.10.588

2023

22nd World Congress of the International Federation of Automatic Control (IFAC)

2405-8963

IFAC PAPERSONLINE

56

2

JUL 09-14, 2023

null,Yokohama,JAPAN

RADARWEG 29, 1043 NX AMSTERDAM, NETHERLANDS

ELSEVIER

In this paper, we study the control design problem for a cart-pole system without the prior knowledge of its physical parameters. The control task involves both swing-up and balancing. Two control methods are compared: 1) model-free reinforcement learning; 2) systemidentification based control design. The former uses a popular deep deterministic policy gradient (DDPG) algorithm, whereas the latter uses a model-based design together with a system identification method for parameter estimation. The results show that system-identification based control design is far superior than reinforcement learning in terms of training, computation complexity and control performance, but requiring skills for system modelling, parameter estimation and control design. Copyright (c) 2023 The Authors.

Reinforcement learning optimal control system identification.

第一 ; 通讯

英语

CPCI-S

Automation & Control Systems

Automation & Control Systems

WOS:001196708400685

Web of Science

Southern Univ Sci & Technol, Sch Syst Design & Intelligent Mfg, Shenzhen, Peoples R China

Fu, Minyue. Control of the Cart-Pole System: Model-based vs. Model-free Learning[C]. RADARWEG 29, 1043 NX AMSTERDAM, NETHERLANDS:ELSEVIER,2023.