Broiler FCR Optimization using Norm Optimal Terminal Iterative Learning Control

Simon Vestergaard Johansen, Martin Riisgaard Jensen, Bing Chu, Jan Dimon Bendtsen, Jesper Mogensen, Eric Rogers

Research output: Contribution to journalJournal articleResearchpeer-review

Abstract

Broiler feed conversion rate optimization reduces the amount of feed, water, and electricity required to produce a mature broiler, where temperature control is one of the most influential factors. Iterative learning control provides a potential solution given the repeated nature of the production process, as it has been especially developed for systems that make repeated executions of the same finite duration task. Dynamic neural network models provide a basis for control synthesis, as no first-principle mathematical models of the broiler growth process exist. The final feed conversion rate at slaughter is one of the primary performance parameters for broiler production, and it is minimized using a modified terminal iterative learning control law in this work. Simulation evaluation of the new designs is undertaken using a heuristic broiler growth model based on the knowledge of a broiler application expert and experimentally on a state-of-the-art broiler house that produces approximately 40,000 broilers per batch.
Original languageEnglish
JournalI E E E Transactions on Control Systems Technology
ISSN1063-6536
DOIs
Publication statusE-pub ahead of print - 2020

Fingerprint

Iterative Learning Control
Dynamic Neural Networks
Norm
Temperature Control
Optimization
Growth Process
First-principles
Growth Model
Electricity
Neural Network Model
Batch
Heuristics
Mathematical Model
Synthesis
Model-based
Water
Temperature control
Evaluation
Mathematical models
Neural networks

Keywords

  • Iterative learning control
  • Biosystems
  • Neural networks

Cite this

Johansen, Simon Vestergaard ; Riisgaard Jensen, Martin ; Chu, Bing ; Bendtsen, Jan Dimon ; Mogensen, Jesper ; Rogers, Eric. / Broiler FCR Optimization using Norm Optimal Terminal Iterative Learning Control. In: I E E E Transactions on Control Systems Technology. 2020.
@article{723547d0d16c4ac6932d6b4dfed0e66b,
title = "Broiler FCR Optimization using Norm Optimal Terminal Iterative Learning Control",
abstract = "Broiler feed conversion rate optimization reduces the amount of feed, water, and electricity required to produce a mature broiler, where temperature control is one of the most influential factors. Iterative learning control provides a potential solution given the repeated nature of the production process, as it has been especially developed for systems that make repeated executions of the same finite duration task. Dynamic neural network models provide a basis for control synthesis, as no first-principle mathematical models of the broiler growth process exist. The final feed conversion rate at slaughter is one of the primary performance parameters for broiler production, and it is minimized using a modified terminal iterative learning control law in this work. Simulation evaluation of the new designs is undertaken using a heuristic broiler growth model based on the knowledge of a broiler application expert and experimentally on a state-of-the-art broiler house that produces approximately 40,000 broilers per batch.",
keywords = "Iterative learning control, Biosystems, Neural networks",
author = "Johansen, {Simon Vestergaard} and {Riisgaard Jensen}, Martin and Bing Chu and Bendtsen, {Jan Dimon} and Jesper Mogensen and Eric Rogers",
year = "2020",
doi = "10.1109/TCST.2019.2954300",
language = "English",
journal = "I E E E Transactions on Control Systems Technology",
issn = "1063-6536",
publisher = "IEEE",

}

Broiler FCR Optimization using Norm Optimal Terminal Iterative Learning Control. / Johansen, Simon Vestergaard; Riisgaard Jensen, Martin; Chu, Bing; Bendtsen, Jan Dimon; Mogensen, Jesper; Rogers, Eric.

In: I E E E Transactions on Control Systems Technology, 2020.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - Broiler FCR Optimization using Norm Optimal Terminal Iterative Learning Control

AU - Johansen, Simon Vestergaard

AU - Riisgaard Jensen, Martin

AU - Chu, Bing

AU - Bendtsen, Jan Dimon

AU - Mogensen, Jesper

AU - Rogers, Eric

PY - 2020

Y1 - 2020

N2 - Broiler feed conversion rate optimization reduces the amount of feed, water, and electricity required to produce a mature broiler, where temperature control is one of the most influential factors. Iterative learning control provides a potential solution given the repeated nature of the production process, as it has been especially developed for systems that make repeated executions of the same finite duration task. Dynamic neural network models provide a basis for control synthesis, as no first-principle mathematical models of the broiler growth process exist. The final feed conversion rate at slaughter is one of the primary performance parameters for broiler production, and it is minimized using a modified terminal iterative learning control law in this work. Simulation evaluation of the new designs is undertaken using a heuristic broiler growth model based on the knowledge of a broiler application expert and experimentally on a state-of-the-art broiler house that produces approximately 40,000 broilers per batch.

AB - Broiler feed conversion rate optimization reduces the amount of feed, water, and electricity required to produce a mature broiler, where temperature control is one of the most influential factors. Iterative learning control provides a potential solution given the repeated nature of the production process, as it has been especially developed for systems that make repeated executions of the same finite duration task. Dynamic neural network models provide a basis for control synthesis, as no first-principle mathematical models of the broiler growth process exist. The final feed conversion rate at slaughter is one of the primary performance parameters for broiler production, and it is minimized using a modified terminal iterative learning control law in this work. Simulation evaluation of the new designs is undertaken using a heuristic broiler growth model based on the knowledge of a broiler application expert and experimentally on a state-of-the-art broiler house that produces approximately 40,000 broilers per batch.

KW - Iterative learning control

KW - Biosystems

KW - Neural networks

U2 - 10.1109/TCST.2019.2954300

DO - 10.1109/TCST.2019.2954300

M3 - Journal article

JO - I E E E Transactions on Control Systems Technology

JF - I E E E Transactions on Control Systems Technology

SN - 1063-6536

ER -