Mobile Robot Based Autonomous Selection of Fuzzy-PID Behavior and Visual Odometry for Navigation and Avoiding Barriers in the Plant Environment

Abstract

This study aims to assess navigation and obstacle avoidance by autonomous mobile robots in plant environments using Fuzzy-PID behavior selection algorithms and visual odometry. The research variables are height, width, and distance. The PID set point is 80, the proportional constant value is 0.5, the integral constant value is 0.45, and the derivative constant value is 0.6. Mobile robots were tested using model 1 and model 2 designs on different patterns with as many as 9 obstacles. The result of the ultrasonic sensor testing is that the overall accuracy was 99.71%. For each of the category, the Pixy camera testing results in block 1 having an accuracy of 77.44%, block 2 97.88%, and block 3 85.88%. Mobile robot testing on the same model shows an increase in accuracy after the distance input value on the fuzzy is enlarged. Future research can use machine learning methods.