Reactive navigation system based on H∞ control system and LiDAR readings on corn crops

Severe working conditions, the cost reduction need and production growth justify the adoption of technological techniques in agricultural production. Currently, Global Navigation Satellite System (GNSS) based navigation systems are very popular in agriculture. However, relying only on GNSS data and the availability of crop maps that also need to be updated may become a problem. One may face lack of GNSS satellite signal during the path and the navigation system may fail. As an alternative, this paper presents the modeling, development and validation of a reactive navigation system for agricultural fields based on a frontal light detection and ranging (LiDAR) sensor and a H infinity robust controller. A small-scale mobile robot (Helvis3) was used to validate the controllers and carry out experiments in both controlled and farm scenarios. According to the experimental results, the proposed navigation system is capable of controlling the robot displacement between crop rows, keeping it in the middle of the corresponding path with minimum error, despite environmental disturbances.