Estimation and assessing the performance indexes of feedback mechanism in a high voltage weed controller system
Paper ID : 1141-NICAME1402
Authors:
Bahram Besharati *1, Ali Jafari1, Hossein Mousazadeh1, Hossein Navid2
1Department of Agricultural Machinery Engineering, University of Tehran, Karaj, Iran
2Department of Biosystems Engineering, Faculty of Agriculture, University of Tabriz, Iran
Abstract:
Removing and controlling weeds is necessary, as they can interfere with crops. There are different methods for weed control, but electrical energy is an ideal and environmental-friendly method. In this research, a high-voltage weed controller equipped with a feedback mechanism was evaluated. During tests, the weed controller was guided along a path and applied high voltage to annihilate weeds. The death rate of weeds measured by the feedback mechanism was compared with the actual death rates of them. To estimate and evaluate the performance indexes of the system in detecting the presence of weed plants (living plants) and their destruction (dead plants), the binary classification method was used. Due to the high value of the evaluation indexes (95% accuracy, 96% precision, 99% sensitivity, and 80% specificity), this feedback mechanism is a suitable method for monitoring and increasing the performance of high-voltage weed controllers. Therefore, the existence of a feedback mechanism for high-voltage weed controllers is a necessity.
Keywords:
Feedback, Performance indexes, High-voltage, Weed controller
Status : Paper Accepted (Oral Presentation)