Multiparametric monitoring in equatorian tomato greenhouses (III): Environmental measurement dynamics

Mayra Erazo-Rodas, Mary Sandoval-Moreno, Sergio Muñoz-Romero, Mónica Huerta, David Rivas-Lalaleo, José Luis Rojo-álvarez

Research output: Contribution to journalArticlepeer-review

7 Scopus citations

Abstract

World population growth currently brings unequal access to food, whereas crop yields are not increasing at a similar rate, so that future food demand could be unmet. Many recent research works address the use of optimization techniques and technological resources on precision agriculture, especially in large demand crops, including climatic variables monitoring using wireless sensor networks (WSNs). However, few studies have focused on analyzing the dynamics of the environmental measurement properties in greenhouses. In the two companion papers, we describe the design and implementation of three WSNs with different technologies and topologies further scrutinizing their comparative performance, and a detailed analysis of their energy consumption dynamics is also presented, both considering tomato greenhouses in the Andean region of Ecuador. The three WSNs use ZigBee with star topology, ZigBee with mesh topology (referred to here as DigiMesh), and WiFi with access point topology. The present study provides a systematic and detailed analysis of the environmental measurement dynamics from multiparametric monitoring in Ecuadorian tomato greenhouses. A set of monitored variables (including CO2, air temperature, and wind direction, among others) are first analyzed in terms of their intrinsic variability and their short-term (circadian) rhythmometric behavior. Then, their cross-information is scrutinized in terms of scatter representations and mutual information analysis. Based on Bland–Altman diagrams, good quality rhythmometric models were obtained at high-rate sampling signals during four days when using moderate regularization and preprocessing filtering with 100-coefficient order. Accordingly, and especially for the adjustment of fast transition variables, it is appropriate to use high sampling rates and then to filter the signal to discriminate against false peaks and noise. In addition, for variables with similar behavior, a longer period of data acquisition is required for the adequate processing, which makes more precise the long-term modeling of the environmental signals.

Original languageEnglish
Article number2557
JournalSensors (Switzerland)
Volume18
Issue number8
DOIs
StatePublished - 4 Aug 2018

Bibliographical note

Funding Information:
This work was supported in part by Research Grants PRINCIPIAS and FINALE (TEC2013-48439-C4-1-R and TEC2016-75161-C2-1-R) from Spanish Government and by Research Grant PRICAM (S2013/ICE-2933) from Comunidad de Madrid; in part by the Universidad de las Fuerzas Armadas ESPE under Research Grant Energy Efficiency in Wireless Sensor Networks (2016-PIC-043 and 2015-PIT-004); and in part by the Thematic Network RiegoNets (CYTED project 514RT0486), and PLATANO project from Universidad Politécnica Salesiana, Cuenca, Ecuador.

Funding Information:
Acknowledgments: This work was supported in part by Research Grants PRINCIPIAS and FINALE (TEC2013-48439-C4-1-R and TEC2016-75161-C2-1-R) from Spanish Government and by Research Grant PRICAM (S2013/ICE-2933) from Comunidad de Madrid; in part by the Universidad de las Fuerzas Armadas ESPE under Research Grant Energy Efficiency in Wireless Sensor Networks (2016-PIC-043 and 2015-PIT-004); and in part by the Thematic Network RiegoNets (CYTED project 514RT0486), and PLATANO project from Universidad Politécnica Salesiana, Cuenca, Ecuador.

Publisher Copyright:
© 2018 by the authors. Licensee MDPI, Basel, Switzerland.

Keywords

  • Greenhouses
  • Mutual information
  • Parametric modeling
  • Residuals
  • Rhythmometric

Fingerprint

Dive into the research topics of 'Multiparametric monitoring in equatorian tomato greenhouses (III): Environmental measurement dynamics'. Together they form a unique fingerprint.

Cite this