Sim-optimization hybrid approach for scheduling randomly deteriorating treatment tasks in horticulture

Abstract

In this paper, we study the problem of scheduling robotized tasks in the context of Agriculture 4.0. The objective is to optimize the treatment tasks of plants against an evolving disease (mildew) within a greenhouse. The treatment is performed using a type-C ultraviolet radiation (UV-C) by an UV-Robot. We propose a semi-dynamic simulation-optimization approach based on a Markovian model of the disease behavior in the greenhouse. Two variants of simulation-optimization hybridization are tested and analyed.