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)
153
air dry conditions for 1000 ml of soil. The dry soil
measurement is then used as
𝑚
𝑑𝑟𝑦
.
The next step is that measurements are carried out at five
levels
of increase in water volume, namely air dry, 100 ml,
200 ml, 300 ml and 400 ml. For each increase in water
volume, measurements are also
carried out using an IoT-
based soil moisture sensor. The results of the gravimetric test
experiment are as follows.
T
ABLE
3.
G
RAVIMETRIC
T
EST
Table 3 shows that the greater the soil + container weights,
the smaller the sensor measurement. The same pattern was
also obtained for volumetric soil moisture content (
𝜃
𝑣
). The
𝜃
𝑣
value is obtained using
𝜌
𝑠𝑜𝑖𝑙
=
0.57. To obtain predictions
of soil moisture, the
𝜃
𝑣
and sensor
measurement values are
then modelled into the linear and polynomial regression
models.
C. Model
Volumetric soil moisture content and raw sensor data were
analyzed using linear regression and polynomial regression
models. The volumetric soil moisture
content becomes the
response variable (y) and raw soil moisture sensor data
becomes the predictor variable (x). Adjusted R Squared is
then carried out for each model as follows:
T
ABLE
4
M
ODEL
S
ELECTION
