On Field calibration is increasingly considered as the best performing approach for air quality monitor devices. Field recorded sensor data together with co-located reference data build suitable dataset that are more representative of the complexity of real world conditions. However, many researchers pointed out the possible lack of generalization due to the strong dependence on the condition encountered during the field recordings. This work, for the first time, try to assess the robustness of this approach to relocation of the sensor nodes. This is particular relevant for mobile deployments and for guaranteeing the scalability properties of this calibration approach in pervasive deployments. Neural Networks have been used to provide for a nonlinear multivariate calibration algorithm. An extensive dataset, recorded in Oslo during 2015-16, provided the ground for a multi-node/multi-weeks assessment. The observed differences account for a greater influence of seasonal changes on the performances with respect to relocation effects. © 2017 IEEE.

Is on field calibration strategy robust to relocation?

Di Francia, G.;Fattoruso, G.;Salvato, M.;De Vito, S.;Esposito, E.
2017

Abstract

On Field calibration is increasingly considered as the best performing approach for air quality monitor devices. Field recorded sensor data together with co-located reference data build suitable dataset that are more representative of the complexity of real world conditions. However, many researchers pointed out the possible lack of generalization due to the strong dependence on the condition encountered during the field recordings. This work, for the first time, try to assess the robustness of this approach to relocation of the sensor nodes. This is particular relevant for mobile deployments and for guaranteeing the scalability properties of this calibration approach in pervasive deployments. Neural Networks have been used to provide for a nonlinear multivariate calibration algorithm. An extensive dataset, recorded in Oslo during 2015-16, provided the ground for a multi-node/multi-weeks assessment. The observed differences account for a greater influence of seasonal changes on the performances with respect to relocation effects. © 2017 IEEE.
9781509023912
Calibration algorithms;Machine learning;Mobile chemical sensing;Distributed air quality monitors
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/20.500.12079/6073
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus ND
social impact