Water Turbidity Control System in Aquascape Plants

Study Case Water Turbidity Control System in Aquascape Plants Using an IoT-Based Fuzzy Control Method Michael Kurniawan (2018123456) University of Cincinnati, 2600 Clifton Ave, Cincinnati, OH 45221, United States kurniaml@mail.uc.edu Abstract Aquascaping is the art of arranging aquatic plants together with rocks, corals, driftwood, and other decorative elements in an aquarium in a natural and aesthetically pleasing manner, so that it looks like gardening underwater. The main purpose of aquascaping is to create an attractive underwater landscape while paying attention to plant maintenance aspects. Factors that affect aquascape conditions include water turbidity, temperature, lighting, and fish feeding to ensure the survival of organisms in the aquarium. This study was designed to optimize the aquascape system by using fuzzy logic on water turbidity, with parameters between 0.48 and 200 NTU, to keep plants healthy. This system regulates the speed of the water pump so that turbid water can be filtered through a filter box. A temperature sensor will turn on the heater to maintain stable water temperature. In addition, this study enables real-time control of lighting and fish feeding using ESP32. Water condition monitoring is carried out through IoT (Internet of Things) technology, which provides periodic monitoring and feeding notifications, thereby simplifying aquascape maintenance. The results of the study show a fuzzy method error of 0.1% on the water turbidity sensor. After more than two days of filtering, the turbidity value dropped from 200 NTU to 0.48 NTU, although some plants appeared slightly wilted. Keywords: Aquascape, Water Turbidity, Fuzzy, IoT MICHAEL KURNIAWAN 1 Study Case INTRODUCTION Aquascape is the art of arranging aquatic plants with elements such as rocks, corals, driftwood, and other decorations in an aquarium aesthetically, so that it looks like gardening underwater [4]. The main purpose of aquascaping is to create an attractive underwater landscape while paying attention to plant care. Important factors in aquascaping include water turbidity, temperature, lighting, and fish feeding. High turbidity can damage or kill plants. Water becomes turbid when it contains many suspended particles, decomposed organic matter, mud, sand, colloids, or floating objects. Water clarity affects plant growth, with an ideal turbidity standard of 5–25 NTU (Nephelometric Turbidity Unit) [4]. Water temperature is greatly influenced by climate; in the tropical climate of the United States, temperature fluctuations can cause the water to reach >28°C or <22°C, whereas the ideal temperature for aquascaping is 22–28°C. This study set a temperature set point of 25°C. Lighting is also important for plant photosynthesis, so LED lights are used for 7 hours per day. The aquarium has a capacity of 97 liters, and fish feeding must be regular to support biota life. Many aquascapers still pay little attention to feeding and lighting schedules. This study aims to optimize the aquascape system with fuzzy logic to regulate water turbidity between 0.48–200 NTU. The pump speed is controlled to filter turbid water, the temperature sensor turns on the heater to maintain stability, and the ESP32 controls lighting and fish feeding in real-time. Monitoring is carried out through IoT with automatic notifications to facilitate maintenance. RESEARCH METHOD Overall Block Diagram Block diagrams explain the working process of a device visually and are often used in engineering. MICHAEL KURNIAWAN 2 Study Case Figure 1: Block Diagram ESP32 acts as the “brain” of the system. Temperature sensors, water turbidity sensors, ultrasonic sensors, and RTC are calibrated to achieve minimum error. The temperature sensor activates the heater relay when it reaches the set point. The ultrasonic sensor sends a notification to the Blynk application when fish feed is running low. The RTC sets the schedule for the servo motor for fish feed and the LED lights for photosynthesis. The main water turbidity sensor is controlled by fuzzy logic on the ESP32 to adjust the pump speed based on input conditions. A 12V power supply is used, stepped down to 5V via a buck converter. Sensor outputs are monitored via Blynk or an LCD that displays turbidity, temperature, and RTC. MICHAEL KURNIAWAN 3 Study Case Figure 2: System Block Diagram This block diagram explains fuzzy control of the water pump speed. The turbidity sensor reads the set point value in NTU with error and delta error variables. The error variable has three linguistic values: Very Clean, Clean, Normal. The delta error also has three values: Slightly Turbid, Turbid, Very Turbid. The ESP32 processes this data with fuzzy logic, regulating the motor's PWM speed: slow, normal, fast. The sensor reads data continuously until the device is turned off. The water pump filters the water, so turbidity decreases over time. Mechanical Design Mechanical design was carried out to determine the physical specifications of the device in 3D. There are two mechanical components: an aquarium with a sensor/output layout, and an electronic box for electronic components. MICHAEL KURNIAWAN 4 Study Case Figure 3: Mechanical Design of Aquarium Glass aquarium: length 70 cm, width 40 cm, height 40 cm. Components: 1. SENO189 water turbidity sensor 2. DS18B20 temperature sensor 3. 12VDC pump motor 4. Heater 5. 3D printed fish feeder with ultrasonic sensor & servo motor 6. Filter box connected to the pump Figure 4: Electrical Box Design Electronic box: length 24 cm, width 22 cm, height 13 cm. Components: 1. I2C 20x4 LCD 2. Acrylic box Electronic Design MICHAEL KURNIAWAN 5 Study Case Figure 5: Turbidity Sensor Wiring Design The SENO189 sensor detects water clarity and particles. It uses a 5V input, connected to the ESP32 GPIO. The output controls the 12VDC pump via the L298N driver. Turbidity data can be monitored online via the Blynk application. Figure 6: Overall Electronic Design Table 1 explains the placement of ESP32 pins for all sensors and actuators. TABLE 1: OVERALL ELECTRONIC DESIGN PINS MICHAEL KURNIAWAN 6 Study Case Fuzzy Logic Design a) Error Variable Figure 7: Membership function of the error variable: b) PWM Variable Membership function of the PWM output: c) Rule Base Table 2: Fuzzy rule base: d) Inference Using the min-max method. The minimum value is sought from the input membership function, then the maximum of the minimum values is used to determine the output. Example of inference calculation: MICHAEL KURNIAWAN 7 Study Case e) Defuzzification The Weighted Average method is used because it is computationally efficient. Each membership function is weighted by its maximum value. The PWM output value is calculated by: Here, ∑ is algebraic summation, and xi is the element with maximum membership. The defuzzification result determines the speed of the pump motor. RESULTS AND DISCUSSION Testing the SENO189 Turbidity Sensor TABLE 3: Turbidity Sensor Testing The calibration testing of the SENO189 sensor was performed using a HACH 2100Q turbidity meter, a device that measures water turbidity in NTU units with a maximum capacity of 800 NTU. This device package includes NTU standard solutions from the manufacturer to ensure accurate calibration. The calibration results showed an average error of 1.18% for the SENO189 sensor, indicating that the sensor works optimally and can be used for aquascape turbidity measurements. MICHAEL KURNIAWAN 8 Study Case Mamdani Fuzzy Testing TABLE 4: Mamdani Fuzzy Testing The testing was conducted by comparing the results of the ESP32 microcontroller processing with manual calculations. The average error obtained was 0.1%, indicating that the system was almost perfect. MICHAEL KURNIAWAN 9 Study Case Figure 8 : Water Condition 200 NTU The initial water contained a mixture of black sand and fine sand, resulting in turbidity reaching 200 NTU. Figure 9: Water Condition 0.48 NTU After continuous filtration, the water became clear and safe for aquascape plants. Some plants appeared slightly wilted due to the initial condition of the water being very turbid. Figure 13: Fuzzy Control Response Graph The curve shows more consistent and stable changes in PWM values (motor speed) when using Mamdani fuzzy control, which is able to adjust to changes in water turbidity in real time. MICHAEL KURNIAWAN 10 Study Case Figure 14: Response Graph Without Fuzzy Control Without fuzzy control, motor PWM changes are more rigid and do not meet requirements, resulting in a less than optimal response to changes in water turbidity. CONCLUSION This study adds several supporting components to maintain the survival of biota in aquascapes, including temperature sensors, ultrasonic sensors, RTC, LED lights, and heaters, all of which are controlled according to their respective functions. The main control for water turbidity uses the Mamdani fuzzy method, with a comparison between microcontroller and manual calculations showing an error of 0.1%, almost approaching 0%, although some plants appear slightly wilted due to the initial poor water conditions. A comparison of pump speed graphs shows that the fuzzy system produces more regular and smoother PWM changes, unlike the non-fuzzy system, which is more rigid and unresponsive every second. Thus, Mamdani fuzzy control has proven to be effective in maintaining water turbidity stability and supporting the health of aquascape plants. MICHAEL KURNIAWAN 11 Study Case REFERENCE MICHAEL KURNIAWAN 12