The Performance Analysis of Tactile and Ultrasonic Sensors for Planting, Fertilizing, and Cultivating Cover Crops

摘要

Navigation and guidance are becoming more crucial in agricultural processes for various reasons including to maximize the ground coverage and to minimize redundant applications and skipped areas, especially for those operations, which need the passageof agricultural machines after planting and germination of plants, such as cover crops. For this purpose, two guidance systems, Ultrasonic and TACTILE of Reichhardt Electronic Innovations, have been tested in laboratory conditions. A test bench purposely has been designed in order to capture the performances of the both sensors at the same conditions viz. different patterns ofplant rows and seven different tractor speeds. Signals from the sensors are collected and processed The output data of sensors were too noisy and thus, several digital filters for both sensors have been tested and analyzed. According to the studies, the Low-Pass Butterworth filter is best for the data collected with the Tactile Sensor and Gaussian filter is most appropriate for filtering data collected by the Ultrasonic sensor. Comparing the observed patterns (after filtering) with sensors and the actual patterns of plants in laboratory conditions, the accuracies of both sensors are compared. The results showed that in higher speeds undesired noise levels reduced significantly. The accuracy of TACTILE sensor was higher than Ultrasonic sensor in all cases. Comparative studies of technical capabilities of both sensors indicate that each sensor has its own advantages and disadvantages. They should be chosen based on the nature of the plants and operation. Moreover, based on the known kinematic bicycle moid of the vehicle, the discretized control model equations are derived and simulated numerically in MATLAB environment Finally,the holistic view of the control algorithm of the whole system - autonomously self-guiding system has been developed.