Résumé
The development, in the last decades, of technologies for precision agriculture allows the acquisition of crop data with a high spatial resolution. This offers possibilities for innovative control and raises new logistics issues that may be solved using discrete event models. In vineyards, some technologies make it possible to define zones with different qualities of grapes and sort the grapes at harvest to make different vintages. In this context, the Differential Harvest Problem (DHP) consists in finding a trajectory of the harvesting machine in the field in order to obtain at least a given quantity of higher quality grapes and minimising working time. In available literature, the DHP has been solved using Constraint Programming. In this paper, we investigate if it is possible to solve the DHP using the Cost Optimal Reachability Analysis feature of a model-checking tool such as UPPAAL-CORA. A model named DHP_PTA has been designed based on the priced timed automata formalism and the UPPAAL-CORA tool. The method made it possible to obtain the optimal trajectory of a harvesting machine for a vine plot composed of up to 14 rows. The study is based on real vineyard data. This paper is an extended version of a communication presented at WODES 2018 (Saddem-Yagoubi IFAC-PapersOnLine 51(7):57-63, 2018).