You are on page 1of 13

Prediction system of hydroponic plant

growth and development using algorithm


Fuzzy Mamdani method
Cite as: AIP Conference Proceedings 1818, 020052 (2017); https://doi.org/10.1063/1.4976916
Published Online: 10 March 2017

I. Made Sudana, Okta Purnawirawan, and Ulfa Mediaty Arief

ARTICLES YOU MAY BE INTERESTED IN

High stenghth concrete with high cement substitution by adding fly ash, CaCO3, silica sand,
and superplasticizer
AIP Conference Proceedings 1818, 020068 (2017); https://doi.org/10.1063/1.4976932

Implementation of Industrial Work Practice management at vocational high school


AIP Conference Proceedings 1818, 020069 (2017); https://doi.org/10.1063/1.4976933

Forward chaining method on diagnosis of diseases and pests corn crop


AIP Conference Proceedings 1818, 020038 (2017); https://doi.org/10.1063/1.4976902

AIP Conference Proceedings 1818, 020052 (2017); https://doi.org/10.1063/1.4976916 1818, 020052

© 2017 Author(s).
Prediction System of Hydroponic Plant Growth and
Development Using Algorithm Fuzzy Mamdani Method

I Made Sudana1,a), Okta Purnawirawan2, Ulfa Mediaty Arief 3


1Informatics
and Computer Engineering Education.
2Department of Electrical Engineering. Engineering Faculty.
3Semarang State University. Semarang Indonesia.

a)
Corresponding author: sudana@mail.unnes.ac.id

Abstract. Hydroponics is a method of farming without soil. One of the Hydroponic plants is Watercress (Nasturtium
Officinale). The development and growth process of hydroponic Watercress was influenced by levels of nutrients,
acidity and temperature. The independent variables can be used as input variable system to predict the value level of
plants growth and development. The prediction system is using Fuzzy Algorithm Mamdani method. This system was
built to implement the function of Fuzzy Inference System (Fuzzy Inference System/FIS) as a part of the Fuzzy
Logic Toolbox (FLT) by using MATLAB R2007b. FIS is a computing system that works on the principle of fuzzy
reasoning which is similar to humans’ reasoning. Basically FIS consists of four units which are fuzzification unit,
fuzzy logic reasoning unit, base knowledge unit and defuzzification unit. In addition to know the effect of
independent variables on the plants growth and development that can be visualized with the function diagram of FIS
output surface that is shaped three-dimensional, and statistical tests based on the data from the prediction system
using multiple linear regression method, which includes multiple linear regression analysis, T test, F test, the
coefficient of determination and donations predictor that are calculated using SPSS (Statistical Product and Service
Solutions) software applications.

INTRODUCTION
Hydroponics is a method of cultivation without soil. Its lexical meaning of hydro is water. However, this
method requires other planting media such as gravel, sand, coconut fiber, a substance silicate, broken rock or
reef, pieces of wood , and foam [7].
One of the hydroponic plants is Watercress (Nasturtium officinale). In the growth and development process
of watercress is influenced by independent variables, namely levels of nutrients, acidity, and temperature [5].
Those independent variables can be used as input variable system to predict the watercress’ growth and
development. The prediction system used in this study was Fuzzy Mamdani algorithm. Fuzzy Mamdani method
works according to the rules of linguistics. The Mamdani method is also known by the name of Max- Min
method [2].
This prediction system was built to implement the function of Fuzzy Inference System (FIS) which is a part
of the Fuzzy Logic Toolbox (FLT) by using application software MATLAB R2007b [1]. FIS is a computing
system working on the principle of fuzzy reasoning which is to that of humans.
Basically FIS consists of four units which are fuzzification unit, fuzzy logic reasoning unit, base knowledge
unit and defuzzification unit [3] [12]. In addition, three dimensional output surface FIS function and the result of
prediction system were analyzed to visualize the influence of independent variables on the plants growth and
development of Watercress [1].

Engineering International Conference (EIC) 2016


AIP Conf. Proc. 1818, 020052-1–020052-12; doi: 10.1063/1.4976916
Published by AIP Publishing. 978-0-7354-1486-0/$30.00

020052-1
This present study aims to determine:
1) How a rule base system was created by looking at the difference of a set of fuzzy values on the levels of
nutrients, acidity (pH), and temperature to Watercress (Nasturtium officinale) ?
2) How a system which can simulate to predict the growth and development of Watercress (Nasturtium
officinale) was realized by using MATLAB software ?
3) How to figure out the effect of different value levels of the content of nutrients, acidity (pH), and
temperature on the watercress plant growth and development (Nasturtium officinale) based on fuzzy
mamdani algorithm method ?

METHOD
The steps used in this study were adopted from [9], [4]. The steps determining the attributes of the system
such as fuzzy variables as input and output of the prediction system, fuzzy set, the universe discourse of fuzzy
variables, the domain value of each linguistic variable.

Determining Fuzzy Variable


Variables used in this study were the variables of input system:
a. Variable of the nutrition levels.
b. Variable of acidity levels.
c. Variable of temperature.

The output system variables are the growth and development of Watercress. The process is illustrated be
illustrated in Fig. 1.

FIGURE 1. Input and Output System Predictions

1) Determining Fuzzy Set:


The linguistic variables included:
x Variable nutrition levels [less, enough, more].
x Variable levels of acidity [acid, neutral, alkaline].
x Variable temperature [cold, medium, hot].
x Variable growth and development [low, normal, high].

2) Determining Universal Discourse of Fuzzy Variable.


Universal discourse of fuzzy variables consisted of:
x Nutrition Contents Variable [0 – 2500].
x Acidity level variable [3 – 10].
x Temperature variable [15 – 50].
x The growth and development of watercress variable [1 – 40].

020052-2
3) Determine the Domain Value In Each Linguistic Variable.
The domain values of each linguistic variables are as follows:
a. Nutrition level variable:
x Less : [0 – 250 – 550 – 650].
x Enough : [651 – 1000 – 1700 – 1800].
x More : [1801 – 2000 – 2400 – 2500].
b. Acidity level variable:
x Acid : [3 – 5 – 6 – 6,9].
x Neutral : [7 – 7 – 7].
x Alkaline : [7,1 – 8 – 9 – 11].
c. Variable of temperature:
x Cold : [15 – 18 – 21 – 24].
x Medium : [24,1 – 25 – 29 – 30].
x Hot : [30,1 – 33 – 37 – 40].
d. Variable of Watercress Growth and Development:
x Low : [1 – 11 – 15 – 20,9].
x Normal : [21 – 25 – 28 – 30,9].
x High : [31 – 33 – 36 – 40].
The domain values can be illustrated in Figure 2 to 5.

FIGURE 2. Variable of Content Nutrition

FIGURE 3. Variable of Acidity Level

020052-3
FIGURE 4. Variable of Temperature

FIGURE 5. Variable of Plant Growth and Development

After determining the attributes of the system used in the construction of this prediction system, then the
prediction system architecture of hydroponic plants Watercress growth and development using mamdani fuzzy
algorithm method was constructed, as shown in Fig. 6.

FIGURE 6. The Architecture of Prediction System

020052-4
More explanation about the design of prediction system of hydroponic Watercress growth and development
using algorithm fuzzy mamdani method can be shown in Fig. 7.

FIGURE 7. Design Flow of The Prediction System

According to the Figure 6 and 7, the process of the system to predict the growth and development of
hydroponic Watercress has the processing units as follows:

Fuzzification Unit
It is a unit that serves to change input values that is crisp fuzzy set into fuzzy set which has a degree of
membership with a range of 0 to 1 [4]. To obtain the degree of membership in each variable, this prediction
system used the membership function representation of triangular and trapezoid curve according to Figure 8 to 9
[8], [6].

FIGURE 8. The Membership Function Representation of Curves Triangle

020052-5
FIGURE 9. Membership Function Representation Trapezoid Curve

1. Fuzzy Logic Reasoning Unit.


It is a unit that serves to input value that has been transformed into a fuzzy set to the rule base system [1].
2. Knowledge Base Unit.
This unit was divided into two parts as follows, a) data Base, which contains the functions in each
membership of the fuzzy sets related to the value of linguistic variables which were used, and b) Rule Base,
which contains the rules in the form of fuzzy implications.
The unit of this knowledge database used the MIN method, illustrated in Figure 10, to process the rule base
system that fulfills the membership degree of all input variables.

FIGURE 10. Example Using the Method MIN


3. Defuzzification Unit.
It is a unit that serves to change the input value that has been processed by the base unit of knowledge in the
form of the value of fuzzy sets, and then reshaped into crisp sets (crisp) which has a degree of membership 1 or
0 [6]. The method used is a method that can be seen on MAX Fig. 11.

020052-6
FIGURE 11. Example Using the Method MAX

Method MAX is used if a condition in which the rule base system that fills the membership degree of all
input variables is more than one.
Then to confirm the value that will be output by the prediction system uses method LOM. Method LOM
(Largest of Maximum) is a method to restore the crisp values (crisp) of fuzzy values obtained by taking the
largest value of a domain that has a maximum membership degree value, which can be seen in Fig. 12.

FIGURE 12. Implementation of Method LOM

RESULTS AND DISCUSSION


Rule base system formed 27 rules used in the system. Rule base system can be seen of the Table 1.

020052-7
TABLE 1. Rule Base System
Rule Base Nutrition Acidity Level Temperature Growth and
The - Level Development
1 Less Acid Cold Low
2 Less Neutral Cold Low
Rule Base Nutrition Acidity Level Temperature Growth and
The - Level Development
3 Less Alkaline Cold Low
4 Less Acid Medium Low
5 Less Neutral Medium Low
6 Less Alkaline Medium Low
7 Less Acid Hot Normal
8 Less Neutral Hot Low
9 Less Alkaline Hot Low
10 Enough Acid Cold Normal
11 Enough Neutral Cold Low
12 Enough Alkaline Cold Low
13 Enough Acid Medium Normal
14 Enough Neutral Medium Low
15 Enough Alkaline Medium Low
16 Enough Acid Hot Normal
17 Enough Neutral Hot Low
18 Enough Alkaline Hot Low
19 More Acid Cold High
20 More Neutral Cold Normal
21 More Alkaline Cold Low
22 More Acid Medium High
23 More Neutral Medium Normal
24 More Alkaline Medium Low
25 More Acid Hot High
26 More Neutral Hot Normal
27 More Alkaline Hot Low

of the Table 1, can be example rule base system, as follows: (a) If the nutrition level is less, the acidity level is
acid, and the environmental temperature is cold, then the plant growth is low; (b) If the nutrition level is less, the
acidity level is neutral and the environmental temperature is cold, then the plant growth is low; (c) If the
nutrition level is less, the acidity level is alkaline and the environmental temperature is cold, then the plant
growth is low.

Reference hows that mamdani fuzzy method is able to be used to figure out the results of growth and
development of watercress [6]. The results of the prediction system of hydroponic Watercress growth and
development using fuzzy mamdani algorithm method can be seen in Table 3.
TABLE 2. The Result of Prediction System of Watercress Growth and Development
NL AL T PGD
1900 5,5 29,5 37,5
200 6 27 15,8
1500 5 17 28,7

The explanation:
NL : Nutrition Levels (ppm)
AL : The acidity level (H + / liter)
T : Temperature (° C)
PGD : Plants Growth and Development (cm)

020052-8
Based on Table 1, when the nutrition level was 1900 ppm, the acidity level was 5.5 H+/liter and the
temperature was 29.5 °C, then the system predicted the value of growth and development of hydroponic
Watercress was 37.7 cm. However, if the nutrition level is 200 ppm, the acidity level was 6 H+/liter and the
temperature was 27 °C, the system predicted the value of growth and development of hydroponic Watercress
was 15.8 cm, and if the nutrition level was 1500 ppm, the acidity level was 5 H+/liter and the temperature was
17 °C, the system predicted the value of growth and development of hydroponic Watercress was 28.7 cm.
The results of the prediction system is also based on the observation of the process of growth and
development of watercress plant, then used as system parameters in calculating the input system. Values above
the result of the prediction system MATLAB R2007B, using by algorithms fuzzy mamdani method as flow
processes in the system predictive value of growth and development of watercress plant.
The results of output surface FIS (Fuzzy Inference System) visualization diagram formed by software
application of MATLAB R2007B were observed and analysed to know the effect of different levels value of
the independent variables which are the levels of nutrition, acidity (PH), and temperature, and the dependent
variables which are the watercress growth and development. This result is consistent with reference [1].
Shows that the high levels of nutrition and acidity are in the acid value. Therefore, watercress plants can
grow and develop highly, and vice versa, can be seen of the Figure 13. Than shows that the high levels of
nutrition and temperature are in the medium condition. Therefore watercress can grow and develop highly, and
vice versa, can be seen of the Figure 14. Than shows that in the acid conditions and medium temperature
watercress can grow and develop well and vice versa, can be seen of the Fig. 15.

FIGURE 13. The Effect of Nutrition Levels and Acidity Levels for The Growth and Development of Hydroponic
Watercress

FIGURE 14. The Effect of Nutrition Levels and Temperature for the Hydroponic Watercress Growth and Development

020052-9
FIGURE 15. The Effect of Acid and Temperature for the Hydroponic Watercress Growth and Development

Multiple linear regression analysis, T- test , F - test , the coefficient of determination and donations predictor
were conducted using software application SPSS (Statistical Product and Service Solutions) to know the
influence of the independent variables on the control variables [10].
1) Test Multiple Linear Regression Analysis

FIGURE 16. Table of Coefficients Using SPSS

Figure 16 shows the results of multiple linear regression analysis using SPSS. Constant of 27.683, levels of
nutrition of 0,004, the acidity level of -1.615, and temperature environment of 0,029 were necessary to figure
out the multiple linear regression equation
Constant is a constant regression equation or also known as the intercept. While 0.004, -1.615, and 0.29
respectively are constant for X1 X2 and X3.Therefore, the multiple linear regression equation can be expressed
as follows : Y = 27.683 + 0.004 X1 + (-1.615) X2 + 0,29X3.
2) T Test ( Partial Test)
T test was carried out to determine whether the independent variables (levels of nutrients, acidity, and
temperature) had partially significant effect or not on the dependent variable (the growth and development of
watercress) [10].

020052-10
TABLE 3. The Result of T Test
H P N K
H1 Vn influenced Vpt 4,132 H1 was accepted, H0 was rejected
H2 Vk influenced Vpt -4,903 H1 was accepted, H0 was rejected
H3 Vs influenced Vpt 0,034 H1 was rejected H0 was accepted

The explanation:
H : hypothesis
Vn : nutritional variables
Vk : acidity variable
Vs : temperature Variable
VPT : growth Variable

3) F Test (Simultaneous Test)


F test was used to determine whether the independent variables (levels of nutrients, acidity, and temperature)
had simultaneously significant effects on the dependent variables (plant growth and development watercress)
[10]

FIGURE 17. Table of Anova Using SPSS

Fig. 17 shows that F value = 13.039 with the probability value ( sig ) = 0.000. The F value (13.039) > F
table (2.81 ), and sig value was smaller than the probability value 0.05 or 0.000 < 0.05. Therefore, H1 was
accepted and the H0 was rejected meaning that simultaneously the levels of nutrition, acidity, and temperature
significantly influenced the growth and development watercress. H0 was rejected, with a significant level of
confidence or 0.00 or 100% .
4) The Coefficient of Determination (R2)
The coefficient of determination was used to know how contributive the independent variables on the
independent variable in unit of percentage.

FIGURE 18. Table of Model Summary Using SPSS

020052-11
Based on Table 18, it can be seen that R Square was 0.46. It means that the influence of independent
variables, namely levels of nutrition, acidity, and temperature on the dependent variable which is the growth and
development of hydroponic Watercress was 46%. Meanwhile, the remaining of 54.0% was influenced by other
variables that were not studied. Because the value of R is 0.678 above 5%/0.05 or tend to approach a value of 1,
it could be concluded that the ability of independent variables (levels of nutrition, acidity, and temperature) in
explaining the variation of the variable was not limited to the level of standard error of the estimate of 5,145.
5) Donations Predictors
x Relative Contribution (RC)
The level of acidity variable had the highest relative contribution which was -363.3% and then the
temperature variable had the 271% relative contribution and level of nutrition variable was 192.3%. While the
total of the effective contribution of (-363.3) % + 271% + 192.3% = 100%.
x Effective Contribution (EC)
The level of acidity variable had the highest effective contribution was - 42.5% then the temperature variable
had 31.7% effective contribution and the effective contribution of nutrition level variable was 22.5%. While the
total of the effective contribution was (-42.5%) + 31.7% + 22.5% = 11.7%.

CONCLUSION
In the process of constructing the rule base system using fuzzy algorithm mamdani method, rule base system
that can be form was twenty-seven rules.
Prediction system of hydroponics plant growth and development by using fuzzy algorithm mamdani method
can be realized using a software application MATLAB R2007b and the function of Fuzzy Logic Toolbox (FLT).
It applied the process of the prediction system based on the basic concept of fuzzy inference systems (Fuzzy
Inference System / FIS).
Based on the output diagram surface FIS (Fuzzy Inference System) in the prediction system, a variable of
acidity level (PH) had a high impact on growth and development of watercress, then variable temperature and
variable levels of nutrition.
The results of multiple linear regression analysis showed that the influence nutrition level variable was
22,5%, acidity level variable (PH) was – 42,5% and temperature variable was 31,7% on the growth and
development of hydroponic Watercress.

REFERENCE
1. Dewi E, Rama and Dr. Nagaveni N., Int. J.Comput. Sci. , 7, 239-247 (2010).
2. Kusumadewi, Sri, “Analisa &amp; Desain Sistem Fuzzy (Menggunakan TOOLBOX MATLAB) (Graha
Ilmu: Jogjakarta. 2003)
3. M. Kalpana and Dr. A.V. Senthil Kumar. Int. J. Adv. Networking Appl.. 3, 1128 – 1134 (2011).
4. M. Kalpana and Dr. A.V. Senthil Kumar. S. Int. J. Adv. Res. Comput. Sci. Software Eng., 3, 1350-1358
(2013).
5. Rosliani, Rini and Nani Sumarni. “Budidaya Tanaman Sayuran dengan Sistem Hidroponik”. Jurnal
Monografi, No.27, Balai Penelitian Tanaman Sayuran. (2005)
6. Saleh, Alfa. “Implementasi Fuzzy Mamdani dalam Sistem Pendukung Keputusan Menetukan Kesesuaian
Bidang Peminatan Mahasiswa,” Seminar Nasional Teknologi Informasi dan Media, 2, 217-222 (2015).
7. Siswadi dan Teguh Yuwono. Jurnal Agronomika, 9, 257-264 (2015)
8. Suwadnyana, I Made Budi and Gede Bagus Ariana. Jurnal Nasional Pendidikan Teknik Informatika, 1, 153-
163 (2013).
9. Singh, Harsimranjit and Narinder Sharma, Int.J. Emerging Tech. Adv. Eng., 4, 278 – 283 (2014).
10. Sudjana. Metode Statistika. (PT. Tarsito, Bandung, 2005)
11. R., Arumugam, A., Balasubramanian, G., Balaji. V.S, Int.J. Eng. Tech. 5, 1863-1869 (2013).

020052-12

You might also like