DESIGN AND IMPLEMENTATION OF HARDWARE, CONTROL POSITION OF STEPPER MOTOR USING PROFILE ACCELERATION AND DECELERATION IN REAL-TIME, AND SUPERVISED MACHINE LEARNING MODEL ON SYSTEM DEFICIENCY DETECTION NITROGEN OF LETTUCE HYDROPONIC NUTRIENT FILM TECHNIQUE

• Main Author: Alfarisy, Muhamad
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/49278
• Institution: Institut Teknologi Bandung

Nitrogen as an important nutrient in plants, whose levels are currently obtained in the laboratory for a long time, ranging from 3 days to 1 month. In this final project, a system is designed to obtain information on Nitrogen levels faster by using an automation system that performs plant image acquisition on hydroponic racks to obtain mean hue information, and performs Nitrogen predictions using linear regression model correlation and the help of standard DRIS tables so that able to further classify plant status, whether the plant is deficient in Nitrogen or not. The final result, obtained by modeling the R-squared value of 0.686 in the test data and 0.656 in the training data with the linear regression equation model y = 0.187x - 3.01 with variable x in the form of mean hue and y values in the form of prediction of Nitrogen levels, and error evaluation parameters. Mean Absolute Error (MAE) of 0.259, Mean Squared Error (MSE) of 0.51, Root Mean Squared Error (RMSE) of 0.714, and the results of evaluation of classification models on 7 test data and verification of plant status whether Nitrogen deficiency or not has been successful done right.