Comparison of Fuzzy Mamdani and Fuzzy Tsukamoto Methods on Automatic Spray Fan

Simon Petrus Silalahi (Mechatronics Engineering Study Program, Electrical Engineering Department, Politeknik Negeri Batam, Batam, Indonesia)
Agnes I Nababan (Mechatronics Engineering Study Program, Electrical Engineering Department, Politeknik Negeri Batam, Batam, Indonesia)
Daniel Sutopo Pamungkas (Mechatronics Engineering Study Program, Electrical Engineering Department, Politeknik Negeri Batam, Batam, Indonesia)
Adlian Jefiza (Mechatronics Engineering Study Program, Electrical Engineering Department, Politeknik Negeri Batam, Batam, Indonesia)

Journal of Applied Electrical Engineering
Vol. 10 No. 1 (2026): JAEE, June 2026

This study evaluates the implementation of Fuzzy Mamdani and Fuzzy Tsukamoto inference methods in an automated spray fan system designed to mitigate thermal discomfort in tropical climates. Utilizing an Arduino UNO integrated with DHT11 and PIR sensors, the system regulates fan speed and a water pump based on ambient temperature and human presence. Experimental results from 22 data points (25,8°C to 33,3°C) indicate that the Mamdani method exhibits higher sensitivity, initiating fan speed increases at 26,1°C and spray activation at 29,5°C. In contrast, the Tsukamoto method provides a more stable and gradual response, with fan transitions occurring at 28,4°C and spray activation at 30,5°C. While Mamdani is suitable for rapid cooling requirements, Tsukamoto demonstrates superior operational stability and energy efficiency through monotonic output functions. Manual calculations for both methods showed 100% consistency with experimental hardware outputs.