Precision agriculture seeks to optimize production processes by monitoring and analyzingenvironmental variables. For example, establishing farming actions on the crop requiresanalyzing variables such as temperature, a...Precision agriculture seeks to optimize production processes by monitoring and analyzingenvironmental variables. For example, establishing farming actions on the crop requiresanalyzing variables such as temperature, ambient humidity, soil moisture, solar irradiance,and Rainfall. Although these signals might contain valuable information, it is vital to mixup the monitored signals and analyze them as a whole to provide more accurate information than analyzing the signals separately. Unfortunately, monitoring all these variablesresults in high costs. Hence it is necessary to establish an appropriate method that allowsthe infer variables behavior without the direct measurement of all of them.This paper introduces a multi-sensor data fusion technique, based on a sparse representation, to find the most straightforward and complete linear equation to predict and understand a particular variable behavior based on other monitored environmental variablesmeasurements. Moreover, this approach aims to provide an interpretable model that allowsunderstanding how these variables are combined to achieve such results. The fusion strategy explained in this manuscript follows a four-step process that includes 1. data cleaning,2. redundant variable detection, 3. dictionary generation, and 4. sparse regression. Thealgorithm requires a target variable and two highly correlated signals. It is essential to pointout that the developed method has no restrictions to specific variables. Consequently, it ispossible to replicate this method for the semiautomatic prediction of multiple critical environmental variables.As a case study, this work used the SML2010 data set of the UCI machine learning repository to predicted the humidity’s derivative trend function with an error rate lower than 17%and a mean absolute error lower than 6%. The experiment results show that even thoughsparse model predictions might not be the most accurate compared to those of linearregression (LR), support vector machine (SVM), and extreme learning machine (ELM) sinceit is not a black-box model, it guarantees greater interpretability of the problem.展开更多
基金The authors acknowledge the Vice-rectory of research of the Universidad Militar Nueva Granada by founding the project INV-ING-2640.
文摘Precision agriculture seeks to optimize production processes by monitoring and analyzingenvironmental variables. For example, establishing farming actions on the crop requiresanalyzing variables such as temperature, ambient humidity, soil moisture, solar irradiance,and Rainfall. Although these signals might contain valuable information, it is vital to mixup the monitored signals and analyze them as a whole to provide more accurate information than analyzing the signals separately. Unfortunately, monitoring all these variablesresults in high costs. Hence it is necessary to establish an appropriate method that allowsthe infer variables behavior without the direct measurement of all of them.This paper introduces a multi-sensor data fusion technique, based on a sparse representation, to find the most straightforward and complete linear equation to predict and understand a particular variable behavior based on other monitored environmental variablesmeasurements. Moreover, this approach aims to provide an interpretable model that allowsunderstanding how these variables are combined to achieve such results. The fusion strategy explained in this manuscript follows a four-step process that includes 1. data cleaning,2. redundant variable detection, 3. dictionary generation, and 4. sparse regression. Thealgorithm requires a target variable and two highly correlated signals. It is essential to pointout that the developed method has no restrictions to specific variables. Consequently, it ispossible to replicate this method for the semiautomatic prediction of multiple critical environmental variables.As a case study, this work used the SML2010 data set of the UCI machine learning repository to predicted the humidity’s derivative trend function with an error rate lower than 17%and a mean absolute error lower than 6%. The experiment results show that even thoughsparse model predictions might not be the most accurate compared to those of linearregression (LR), support vector machine (SVM), and extreme learning machine (ELM) sinceit is not a black-box model, it guarantees greater interpretability of the problem.
