Communication Dans Un Congrès Année : 2024

Parameter estimation of a dynamic growth model for lettuce in an adaptive vertical farm

Résumé

We present an innovative and sustainable vertical farming technology, named adaptive vertical farm (AVF), and investigate parameter estimation for a dynamic growth model of the crops cultivated therein. Unlike existing vertical farming solutions with fixed shelves, the main novelty of the AVF concept consists of cultivating crops in stacked and automatically mobile layers that adapt continuously to the growth of the crops instead of fixing the distance between the shelves to the maximum height at harvest. This allows the installation of a larger number of shelves in the same vertical space. In order to properly control the movements of the shelves, it is crucial to have at disposal an accurate crop growth model, so as to predict the height of the plants at each time step. Toward this end, we focus on a crop growth dynamic model for lettuce, which depends on a set of parameters to be tuned. Based on dry mass production as a growth indicator, an accurate estimate of the parameters is essential for a precise height prediction. Thus, we propose an approach to estimate the parameters affecting structural and non-structural dry mass production. The goal is first to numerically study the identifiability of the parameters of the growth model starting from information on the dry mass that can be collected on the field, and then use the identified model to forecast the growth of plants using real- world measurements. Parameter estimation is performed by solving an optimization problem that aims at minimizing the difference between the measured and predicted dry mass over a given temporal window. Preliminary numerical results are presented to assess the effectiveness of the proposed estimation approach and of considered crop growth model, using datasets made up of both synthetic and real-world measurements.

Domaines

Fichier non déposé

Dates et versions

hal-04600269 , version 1 (04-06-2024)

Identifiants

  • HAL Id : hal-04600269 , version 1

Citer

Echrak Chnib, Patrizia Bagnerini, Mauro Gaggero, Ali Zemouche. Parameter estimation of a dynamic growth model for lettuce in an adaptive vertical farm. 20th International Conference on Automation Science and Engineering, CASE 2024, Aug 2024, Puglia, Italy. ⟨hal-04600269⟩
157 Consultations
0 Téléchargements

Partager

  • More