MIMO Control of Climate in Livestock Production Houses
Authors
Reynskor, Jóannes Højgaard ; Riisager, Mads Jacob
Term
4. term
Education
Publication year
2022
Submitted on
2022-06-01
Pages
112
Abstract
Dette speciale undersøger, hvordan man styrer indeklimaet i et produktionshus til fjerkræ, så det både opfylder dyrenes behov og de setpunkter, som landmanden vælger. Målet er at designe en MIMO-regulator (multiple input, multiple output), der kan holde de vigtigste klimavariabler på deres referencer. Den vigtigste variabel er temperaturen, som skal ligge på reference med en tolerance på ±2°C. Arbejdet er udført i samarbejde med SKOV A/S, som har leveret data fra et produktionshus på Mosegaarden og stillet et Air Physical Laboratory til rådighed som testfacilitet. På baggrund af disse data er der opstillet en ikke-lineær model af huset, og modellens parametre er estimeret med felt- og laboratoriedata. Med modellen er der designet et Extended Kalman Filter (EKF), som fungerer som tilstandsobservatør og forbedrer estimeringen af de indre tilstande ud fra målte signaler. Der er derudover udviklet regulatorer baseret på Linear Quadratic Regulator (LQR) og Linear Quadratic Integrator (LQI). Test i laboratoriet viste, at LQI-regulatoren kunne holde temperaturen på reference inden for kravet på ±2°C.
This thesis examines how to control the indoor climate of a poultry production house so it meets the birds' needs and the farmer's chosen setpoints. The goal is to design a MIMO (multiple-input multiple-output) controller that keeps key climate variables at their references. The most important variable is temperature, which must stay at the reference within a ±2°C tolerance. The work was carried out with SKOV A/S, which provided data from a production house at Mosegaarden and access to an Air Physical Laboratory used as a test facility. Using these data, a nonlinear model of the house was built and its parameters were estimated from field and laboratory measurements. Based on the model, an Extended Kalman Filter (EKF) was designed to act as a state observer, improving estimates of internal states from measured signals. Controllers based on a Linear Quadratic Regulator (LQR) and a Linear Quadratic Integrator (LQI) were developed. Laboratory testing showed that the LQI controller kept the temperature at its reference within the ±2°C requirement.
[This abstract was generated with the help of AI]
Documents