Adaptive architecture towards portability of greenhouse models

L. Miranda, G. Schillaci
This work deals with the portability of greenhouse models, as we believe that this is a challenge to their practical usage in control strategies under production conditions. We address this task by means of adaptive neural networks, which re-adjust their weights when transferred to new conditions. Such an adaptive account for computational models is typical of the field of developmental robotics, which investigates learning of motor control in artificial systems inspired on infants' development. Similarly, to robots, greenhouses are complex systems comprising technical and biological elements, whose state can be measured and modified through control actions. We present an adaptive model architecture to perform online learning. This learning process makes use of an episodic memory and of online re-training, which modify the plasticity of the model according to the prediction error. This allows for adaptation without the need for a complete new training, which might be prohibitive if the data under the new conditions is scarce (in comparison to a research facility). Current experiments focus on how a model of tomato photosynthesis, developed in a research facility, can adapt itself to a new environment in a production greenhouse. The models presented as a proof-of-concept estimate the transpiration and photosynthesis of a hydroponic tomato crop by using measurements of the climate and the state of the actuators as inputs. In the thought experiment, the models are trained and tested using data from a greenhouse in Berlin, Germany. Thereafter, the adaptive architecture is fed with data from a production greenhouse in southern Germany, where other tomato cultivars were grown under different irrigation and climate strategies. The proposed adaptive architecture represents a promising tool for spreading the use of models produced by high-tech research centres to the greenhouse production sector.
Miranda, L. and Schillaci, G. (2020). Adaptive architecture towards portability of greenhouse models. Acta Hortic. 1296, 25-32
DOI: 10.17660/ActaHortic.2020.1296.4
tomato, photosynthesis, transpiration, deep learning, transfer learning, online model learning from greenhouse data

Acta Horticulturae