|OZKAN, H.ERDAL - The Ohio State University|
Submitted to: Computers and Electronics in Agriculture
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 11/16/2010
Publication Date: 1/2/2011
Citation: Jeon, H., Zhu, H., Derksen, R.C., Ozkan, H., Krause, C.R. 2011. Evaluation of ultrasonic sensors for the variable rate tree liner sprayer development. Computers and Electronics in Agriculture. 75(1):213-221. DOI:10.10161j.compag.2010.11.007.
Interpretive Summary: Even though ultrasonic sensors have been used for sensor-controlled variable-rate sprayers to reduce agrochemical use, extensive tests of the sensors under field conditions have not been attempted yet. Excluding sensor test under field conditions may lead to malfunctions of sensor-controlled sprayers in field operation. Durability and detection stability of ultrasonic sensors were investigated under simulated field conditions for the development of variable-rate nursery sprayers. In addition, potential issues in detecting targets with multiple synchronized sensors were investigated by integrating them into a prototype sprayer system operated under typical field application conditions. The investigation demonstrated that the sensor accuracy could be increased by optimizing the sensor and nozzle configuration strategy. The information from this study will assist to develop new robust sensor-controlled field sprayers with significantly increased detection accuracy.
Technical Abstract: Sensors functioning reliably under harsh field conditions are needed for the development of variable-rate sprayers to apply pest control agents for tree liners in ornamental nurseries. Two ultrasonic sensors were tested to determine how their durability and detection stability would be influenced by the changes in temperature, wind, dust, travel speed and spray cloud. One of the sensors did not perform satisfactorily. It had a 0.1% change in root mean square (RMS) error of the detecting distance before and after exposing to the outdoor cold weather conditions. Mean RMS error was 8 % under the dusty condition, 1.5–1.8 % under the windy condition, and 12.3 % to 23 % for the travel speed ranging from 0.8 to 3.0 m/s. It also showed that increasing ambient temperature from 16.7 to 41.6 °C reduced the detection distance by 4.0 %. Detecting through spray cloud caused the RMS error to vary from 1.2 to 61.4 %, the largest measurement error among the six variables tested. To reduce the RMS errors, optimal configurations between the sensor and spray nozzles were identified. In addition, strategies of synchronized multiple sensors were tested for improving the measurement stability and accuracy of the sensor while detecting targets.