DECISION SUPPORT SYSTEM FOR SELECTING CROP AND

IDENTIFYING PLANTING METHODOLOGY BASED ON

ENVIRONMENT AND SPATIAL FACTOR USING FUZZY LOGIC
APPROACH

Wilkin F. Simo1, Randy S. Gamboa2

1
Davao Oriental State College of Science and Technology, [email protected]
2
University of Southeastern Philippines, [email protected]

ABSTRACT:
The study aims to help the farmers identify the suitable crop to be planted on an area base on
environment and spatial factor using fuzzy logic algorithm. Iterative and incremental development
methodology were used for the development of the system which undergone planning, data gathering,
data analysis, system design and implementation. The result of the study found out that the
environment and spatial factor being considered in the study affects the decision of identifying the
crop to be planted in an area. By predicting the weather data, we were able to determine the crops to
be planted on a specific planting date together with the soil properties and slope percentage. The
predicted rainfall data based on weather historical data shows a higher mean average percentage error.
Using fuzzy logic as basis for selecting crops helps the system decide whether a certain plant is
suitable in a certain area. The system were able to provide the most suitable crops to be planted and
gives information on crop water requirements, lime application, land preparations, plant calendar, and
initial planting preparation. In conclusion, the developed system demonstrated the ability to solve the
problem in selecting the suitable crops to be planted on a certain area and deliver aid in crop planning
and decision making for a desired planting date.

Keywords: fuzzy logic; fuzzy inference system; crop selection; decision support system

1. INTRODUCTION Pablo (Bopha) on 4th of December, 2013 that

In the part of Southeast Asia, Philippines is one devastated over 300,000 hectares of land wiping
of the country that are being hardly hit by out entire coconut plantations, and it teared over
typhoons, floods and droughts. This disaster 100 years of farming livelihood.(“Social Impact
leaves a negative effects in terms of economic of Thunderstorm Bopha in Davao Oriental and
and environmental impacts on affected areas Compostella Valley”, 2013). These
which leads to damage crop production, foods circumstances greatly causes significant changes
supply falls and even affect farming livelihood. on the farmers’ decisions in their conduct of
Moreover the unwelcome consequences and new crop planning and it has become a problem
continuous expose to natural disaster makes since it is hard to determine what is the most
agriculture and natural resources become appropriate crop to be planted and which
vulnerable. Because of this scenario, the planting methodology best fits the situation. To
stability of crop production becomes a threat. It address such problems, a decision support
is expected that agricultural production level system was created in order to help identify
will be lowered and makes it production become what would be the suitable seeds or crops to be
unpredictable (Stern Review 2006; Fischer et. planted in an area and the most effective crops
al. 1993; and IPCC 2007). One of the recent on a certain area based on the environment and
tragedy that gave a negative impact in Davao spatial data of the field.
Oriental, Philippines was the Super Typhoon
2. PURPOSE OF THE STUDY
The purpose of the study is to create a decision
support system that will be used for selecting
appropriate crop to be planted in an area.
Predictive analysis was used in order to make
such decisions by collecting historical data on
weather, temperature and field precipitation as
well as other environmental and spatial factor. Figure 2. Trapeziodal Membership Function
The study also aims to develop a system that
will provide a land preparation methods that will This function is described using the following
improve and support the crops being suggested equation:
by the system. 0 ; x < a
3. METHODOLOGY x–a ; a < x <b
In this study, the methodology used for the b–a
development of the system was Iterative and
Incremental Development Methodology. The Trapezoid 1 ; b < x <c (1)
Iterative and Incremental model was developed (x;a,b,c,d) =
to overcome the weaknesses of the waterfall d–x ; c < x <d
model (Larman and Basili, 2003). d–c

0 ; d < x

Where:
x-is real value (Crisp Value) within the universe
of discourse.
a, b, c, d - is the x- coordinates of the four
heads of the trapezoidal and values should
validated the following condition:

Figure 1.Iterative and Incremental Methodology a<b<c<d

Model
The rules generated helps to determine the crop
3.1 Requirements and Planning Phase based on soil type, soil pH, temperature and
This is where the idea of the system was taken rainfall requirements. Below are the example of
place. Interviews from agricultural-related the if-then rules applied in the system:
offices and agencies took place and planned the
initial flow of the system. R1: IF Soil type is “not favorable” AND soil ph
level is “low ph” AND temperature is “low
3.2 Analysis and Design Phase temp” and rainfall requirement is “low”
This is where the analysis of the data took place. THEN Not recommended
Data that were collected went data analysis R2: IF Soil type is “not favorable” AND soil ph
which include Time Series Analysis to predict level is “low ph” AND temperature is “low
future weather data. In this phase the fuzzy temp” and rainfall requirement is
inference system created to help select the crops “favorable” THEN Not recommended
which includes membership functions of each R3: IF Soil type is “not favorable” AND soil ph
crops and rules to be considered for selection. level is “low ph” AND temperature is “low
temp” and rainfall requirement is “high”
THEN Not recommended
R4: IF Soil type is “not favorable” AND soil ph
level is “low ph” AND temperature is
“favorable” and rainfall requirement is
“low” THEN Not recommended
A total of 50 rules were being created in this analysis. As the analysis of data finish it
study. The fuzzy inference system use Weighted provides the model of data as shown on Table 1.
Average Method for deffuzification with an This model will be used later by the system.
equation of:
Environmental Factors Model
(2)
Minimum Temperature Tt = 25.644 - 0.0132t
Maximum Temperature Tt = 30.079 - 0.0280t
Air Temperature Tt = 26.049 + 0.0127t
where ∑ denotes the algebraic summation and x Rainfall Amount Tt = 319.680 - 2.186t
is the element with maximum membership Rainfall Intensity Tt = 0.490-0.003t
function. Rainfall Duration Tt = 67.734-0.180t
Humidity Tt = 83.487-0.0340t
Graphical User Interface design was also created Table 1. Mean Absolute Percentage Error of predicted
weather data
in this phase.
Environmental Factors MAPE
3.3 Implementation and Testing Phase
The system was created in this phase and the Minimum Temperature .120
Maximum Temperature .052
system design generated during analysis and
Air Temperature 0.016
design phase was implemented and tested. Rainfall Amount 1.365
Rainfall Intensity 10.61
Rainfall Duration 0.753
Humidity 0.020
Table 2. Mean Absolute Percentage Error of predicted
weather data

Table 2 shows the MAPE value of the predicted

weather data. The predicted rainfall data has
high percentage error. Having a high error on
the predicted rainfall affects the selection of the
the crops.

Figure 3.Conceptual Framework

The proposed system was implemented based on

the conceptual framework generated above
(Figure 3). It maintains an online database
which stores all of the data being inputted in the
system. Other data like spatial data and
knowledge domain were collected from various Figure 4. Crop Selector Input Page
government and private agency and was
consolidated for data analysis. The collected
data from external sources undergo data analysis
and reasoning which was used to give
recommendations to the farmers. Result coming
from data analysis were stored on the system.
Knowledge base was also created where all
valuable resource related on crop management
and planting methodology was stored.
Figure 5. Result Page
4. FINDINGS/RESULT
The weather data being collected went to data Farmer or Agricuturist must input the required
analysis for prediction using time series farm data like the area location (latitude,
longitude), soil pH, soil type, slope type and
target planting date in the crop selector input Larman, C., & Basili, V. R. (2003). Iterative and
page (Figure 4). These data were processed incremental developments. a brief history.
together with the predicted weather data. After Computer, 36(6), 47-56.
the of analysis of the input and predicted data
are finish through the fuzzy inference system, all
of the results are stored in the database then
redirected to the result page (Figure 5). The
result page shows the farm data inputted in the
system together with the predicted data and
display the crop selected by the system. The
result page also shows the planting guide of the
crop selected together with the expected growth
date of the crop along with the water irrigation
of crops daily needs on a monthly basis, farm
lime requirements and the land preparation
method which was based on Soil and Water
Conservation Methods.

5. CONCLUSION
Creating a decision support system in
selecting crops leads to an effective method in
order to select appropriate crops based on spatial
and environment attributes of an area. There is a
great potential of developing a web-based
decision support system by using the spatial and
environmental factor as to be considered for
selection of crops. This provide a supporting
tool that gives information and knowledge to
farmers on what is the most suitable crop in an
area. The developed decision support system
needs more accurate prediction for rainfall data
to predict accurately the type of crops to
planted.

6. REFERENCES
IPCC. (2007). Contribution of Working Group
II to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change”,
M.L. Parry, O.F. Canziani, J.P. Palutikof, P.J.
van der Linden and C.E. Hanson (eds), in
Climate Change 2007, Fourth Assessment
Report. Cambridge University Press,
Cambridge, United Kingdom and New York,
NY, USA, online at:
http://www.ipcc.ch/publications_and_
data/ar4/wg2/en/contents.html

Fischer, M. and P. Nijkamp (1993). Geographic

Information System, Spatial Modeling, and
Policy Evaluation. Berlin & New York:
Springer-Verlag, pg. 42.