This is an open access article under the  CC–BY-SA  license.  I




Download 0,75 Mb.
Pdf ko'rish
bet2/9
Sana09.12.2023
Hajmi0,75 Mb.
#114189
1   2   3   4   5   6   7   8   9
Bog'liq
01#Повторная калибровка модельного емкостного датчика для измерения влажности почвы IoT

 
This is an open access article under the 
CC–BY-SA
 license. 
I.
 
I
NTRODUCTION
 
Many researchers are interested in calibrating sensors to 
measure soil moisture content. Soil moisture sensor 
calibration is required to be able to develop a low-cost 
automatic irrigation system. Considering that the performance 
of soil moisture sensors is highly dependent on the physical 
and chemical properties of the soil, the development of a 
durable low-cost soil moisture monitoring automation system 
depend on the quality of the calibrated sensor [1]. 
A thermo-gravimetric, which technique by drying the soil 
sample in an oven, have been developed to measure soil 
moisture content. So, dry soil is given a certain volume of 
water and then the moisture sensor value is measured. 
Because the moisture sensor is a qualitative measurement, the 
soil moisture sensor value is expected to change as the water 
volume increases. This technique commonly as the regular 
reference 
because 
it 
can 
provide 
more 
accurate 
measurements. Extended technique of thermos gravimetric be 
carried out based on data-driven [2]. 
Polynomial and linear regression models are frequently 
used to calibrate soil moisture sensors data. Ordinary 
polynomial regression analysis is often used for curve fitting 
while the linear regression model is used for linear fitting. 
Both models express R Squared as a relationship 
measurement. The order of the polynomial necessary to 
explain the dependent variable.
But, computational effort 
required [3]. 
This research aims to provide an application, which 
interactively allows retrieving a calibration equation, based 
on user defined requirements. This application makes 
gravimetric tests easier and accommodates the use of non-
linear models to determine the best model. Interactive web-
based calibration application as has been done previously by 


JAIC 
e-ISSN: 2548-6861 
Re-Calibration of Model-Based Capacitive Sensor for IoT Soil Moisture Measurements 
(Iman Setiawan, Mohammad Dahlan Th. Musa, Saskia Amalia Putri) 
151 
[4] but not using capacitive soil moisture sensor. Our previous 
research [5], [6], we calibrated a percentage-based soil 
moisture sensor. The percentage is calculated using 
observation data when the sensor is in the air and immersed 
in water. There are many missing values found in the stored 
data. This indicates that the sensor calibration process is not 
adaptive, especially if used for a long period of time. So that, 
this research aims to visualise data points and the estimated 
linear and non-linear model in a web-based application. 
Additional features are an interactive web application handles 
small sample sizes problem for model selection by using 
adjusted R squared. The coefficients model obtained is then 
programmed into an Arduino Uno which is connected to a 
mobile phone so that an IoT model-based soil moisture sensor 
is formed. 

Download 0,75 Mb.
1   2   3   4   5   6   7   8   9




Download 0,75 Mb.
Pdf ko'rish

Bosh sahifa
Aloqalar

    Bosh sahifa



This is an open access article under the  CC–BY-SA  license.  I

Download 0,75 Mb.
Pdf ko'rish