Monitoring and prediction of environmental parameters is a challenging task for researchers and technical operators. Data loggers for environmental monitoring are compact devices equipped with microprocessor input channels and data storage. In this paper we propose a new sampling strategy for hydrometric level sensor that can self-adapt based on the error between predicted and observed water level timetrend. Using this procedure it will be possible to dynamically improve the measurement accuracy of the peak stage during a flood event. Support Vector Machines will be used to predict the hydrometric level given a limited number of previous samples. The effectiveness of the method has been tested on a real-world stage-discharge dataset.
Adaptive Sampling for Embedded Software Systrems Using SVMs: Application to Water Level Sensors
DE LEONE, Renato;MAPONI, Pierluigi
2012-01-01
Abstract
Monitoring and prediction of environmental parameters is a challenging task for researchers and technical operators. Data loggers for environmental monitoring are compact devices equipped with microprocessor input channels and data storage. In this paper we propose a new sampling strategy for hydrometric level sensor that can self-adapt based on the error between predicted and observed water level timetrend. Using this procedure it will be possible to dynamically improve the measurement accuracy of the peak stage during a flood event. Support Vector Machines will be used to predict the hydrometric level given a limited number of previous samples. The effectiveness of the method has been tested on a real-world stage-discharge dataset.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.