Skip to main navigation Skip to main content
  • E-Submission

JKSPE : Journal of the Korean Society for Precision Engineering

OPEN ACCESS
ABOUT
BROWSE ARTICLES
EDITORIAL POLICIES
FOR CONTRIBUTORS
REGULAR

인공지능 머신러닝을 이용한 카트-펜듈럼 이동제어 시스템 성형입력 설계

Input Shaping Design for Cart-Pendulum Motion Control System by Using Machine Learning of Artificial Intelligence

Journal of the Korean Society for Precision Engineering 2022;39(6):395-402.
Published online: June 1, 2022

1 가천대학교 대학원 기계공학과

2 가천대학교 기계공학과

1 Department of Mechanical Engineering, Graduate School, Gachon University

2 Department of Mechanical Engineering, Gachon University

#E-mail: mskang@gachon.ac.kr, TEL: +82-31-750-5524
• Received: January 27, 2022   • Revised: April 22, 2022   • Accepted: April 29, 2022

Copyright © The Korean Society for Precision Engineering

This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

  • 24 Views
  • 2 Download
  • 1 Crossref
  • 1 Scopus
prev next

Citations

Citations to this article as recorded by  Crossref logo
  • Dynamic Force‐Shaped Input Control With Adjustable Maneuvering Time for Payload Transportation Systems
    Abdullah Mohammed, Abdulaziz Al-Fadhli, Khalid Alghanim, Emad Khorshid, Petko Petkov
    Journal of Control Science and Engineering.2025;[Epub]     CrossRef

Download Citation

Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.

Format:

Include:

Input Shaping Design for Cart-Pendulum Motion Control System by Using Machine Learning of Artificial Intelligence
J. Korean Soc. Precis. Eng.. 2022;39(6):395-402.   Published online June 1, 2022
Download Citation

Download a citation file in RIS format that can be imported by all major citation management software, including EndNote, ProCite, RefWorks, and Reference Manager.

Format:
Include:
Input Shaping Design for Cart-Pendulum Motion Control System by Using Machine Learning of Artificial Intelligence
J. Korean Soc. Precis. Eng.. 2022;39(6):395-402.   Published online June 1, 2022
Close

Figure

  • 0
  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
Input Shaping Design for Cart-Pendulum Motion Control System by Using Machine Learning of Artificial Intelligence
Image Image Image Image Image Image Image Image Image
Fig. 1 Schematic of cart-pendulum system
Fig. 2 Block diagram of machine learning with FIR filter
Fig. 3 Desired cart motion profile
Fig. 4 Response to cart reference position input profile
Fig. 5 Designed shaped input when λx = 1, λθ = 1, λw = 0
Fig. 6 Shaped input, cart and pendulum responses during updating processes
Fig. 7 Shaped input, cart and pendulum responses according to constant weights
Fig. 8 Desired and designed cart position and velocity profiles
Fig. 9 Shaped input, cart and pendulum responses according to constant and exponential weights
Input Shaping Design for Cart-Pendulum Motion Control System by Using Machine Learning of Artificial Intelligence
Step 1. Assume initial input vector w(0) and let k = 0
Step 2. Calculate cart & pendulum responses
x(k) & θ(k) for w(k) by Eq. (12)
Step 3. Calculate cost function J(k) by Eq. (13)
Step 4. If J(k) ≤ Jmin finish update process
Step 5. Calculate gradient Jkwk by Eq. (14)
Step 6. Update input vector w(k + 1) by Eqs. (15) & (16)
Step 7. k = k +1 and go to Step 2
Parameters Value
System parameters m = 10 [kg], M = 20 [kg], L = 1 [m]
c = 50 [Ns ⁄m], C = 5 [Ns ⁄m]
Control parameters ςd = 0.7 ωn = 2π [rad⁄s], α = 5ςdωn
Kp = 6987.7, Kd = 873.6, Ki = 26045.0
Sampling interval Δt = 0.02 [s]
Parameters Value
Maximum acceleration/
Velocity
amax = 0.8 [m/s2]
vmax = 0.4 [m⁄s]
Settling time tf = 3 [s]
Case Weights
1 λx = 10, λθ = 1, λw = 10
2 λx = 1, λθ = 1, λw = 10
3 λx = 1, λθ = 10, λw = 10
4 λx = 1, λθ = 100, λw = 10
5 λx = 1, λθ = 10e0.2t, λw = 10
Case Cart Pendulum
Rise time
[s]
Maximum
overshoot
[%]
Peak value
[m]
Settling time
[s]
Desired 3.0 0 Minimum
1 3.3 1.2 3.3 10.9
2 3.4 1.6 1.8 6.0
3 3.6 2.3 1.5 5.7
4 4.4 2.8 0.9 5.7
5 3.6 2.1 1.4 4.6
§Settling time: Based on 0.1 deg. band criterion
Table 1 Input design process by machine learning
Table 2 Parameters for simulation
Table 3 Constraints of cart motion
Table 4 Weights in simulations
Table 5 Control performances in each case