Wireless positioning and path tracking for a mobile platform in greenhouse

In order to realize intelligent greenhouse, an automatic navigation method for a mobile platform based on ultra-wideband (UWB) positioning technology was proposed and validated in this study. The time difference of arrival (TDOA) approach was used to monitor and track the UWB positioning to obtain the localization information of the mobile platform working in a greenhouse. After applying polynomial fitting for positioning error correction, the system accuracy was within 5 mm. A fuzzy controller model was constructed by incorporating the lateral and heading deviations as input variables and the steering angle of front wheel as the output variable. A fuzzy rule was established based on domain knowledge, as well as the steering angle of front wheel offline query table, which was applied to alleviate the calculative load of the controller. Experimental results confirmed that the automatic navigation method proposed in this study performed satisfactorily, with a steady-state error ranging from 41 mm to 79 mm when tracking straight line, and an average error of 185 mm and an average maximum error of 532 mm when tracking polygon. In addition, the maximum error occurred at the polygonal corner which could meet the needs of driving on the narrow road in the greenhouse. The method proposed in this study provides a new systematic approach for the research of greenhouse automatic navigation.