A Sensor Fusion-Based Cutting Device Attitude Control to Improve the Accuracy of Korean Cabbage Harvesting

Abstract

Highlights Developed the attitude control system of the cutting device to improve the accuracy of Korean cabbage harvesting. The cutting device provides attitude control, allowing precise harvesting. The attitude control algorithm was designed based on sensor fusion. Field experiments were conducted to compare attitude maintenance and harvest success rates as the attitude control system of cutting device was with/without. Abstract. Harvesting Korean cabbage requires precise cutting since the cutting method determines its quality. However, Korean cabbage fields have uneven surfaces and severe slopes, making it challenging to automatically harvest the crops. This unstructured environment hindered precise cutting, delaying the commercialization of the Korean cabbage harvester. Therefore, we designed an attitude control system for the cutting machine to improve the accuracy of cabbage harvesting. The proposed system controls the cutting level, angle, and height according to the field to improve the harvest performance. This study applied the Kalman filter-based sensor fusion to measure the cutting angle and height for pose estimation and attitude control. The attitude of the cutting device is estimated even in soil and mud by fusing a gyroscope, accelerometer, and linear potentiometer. Subsequently, the attitude is controlled by two hydraulic cylinders so that the cutting machine has the desired cutting angle and height. For the validation of the attitude control, the root mean square error (RMSE) of the cutting angle and height was calculated through experiments in a sloped environment. As a result, the proposed system reduced the RMSE for the cutting angle and height by 92% and 72%, respectively. Furthermore, field tests were performed on farmland with a Korean cabbage harvester to evaluate the harvesting performance of the attitude control system. The harvesting performance was quantitatively computed by scoring the cutting surface of the cabbage, and the performance improved from 56 to 89 points when the attitude control was applied to the cutting machine. Keywords: Attitude control, Automatic cabbage harvesting, Kalman filter, Korean cabbage harvester, Sensor fusion.