Skip to Main content Skip to Navigation
Theses

Simulation et optimisation de la gestion dynamique de tâches évolutives sur des robots mobiles autonomes.Application sur le traitement UV-C robotisé dans l’horticulture.

Abstract : Task scheduling on complex stochastic systems is an activity that cannot be optimized with classical optimization methods. The main difficulty lies in the stochastic and dynamic behavior of these systems, which can modify the properties and/or the number of tasks and/or the resources necessary for their execution. Indeed, it is difficult to estimate the time needed to execute a task and to define an efficient scheduling that takes into account the behavior and evolution of the system.This thesis proposes to address this problem by using an approach based on the coupling between simulation and optimization in several situations. On one hand, simulation manages a temporal horizon, but it has a myopia in the state space. On the other hand, optimization manages the state space, but it has a myopia in time. The objective in coupling is to create an environment close to the real system with the simulation. Then, to integrate the optimization algorithms to improve the behavior of the system. We have applied this approach in the context of Agriculture 4.0, where we have automated the robotic treatment of blight in horticulture.We developed a simulator based on multi-agent systems, where we incorporated optimization engines based on exact and approximate methods. We modeled the disease behavior in a greenhouse based on the Markov chain formalism for each agent representing a plant.The simulation scenarios are based on different types of preventive treatments (conditional, predictive and calendar). The obtained results showed the efficiency of the approach and allowed to propose to horticulturists a tool combining simulation and optimization to help them define and parameterize the appropriate robotic treatment policy.Finally, this thesis proposes several solutions to optimize the application of treatment robots in horticulture. The developed simulator allows the user to choose several treatment strategies and to use the one that is best adapted to its environment.
Complete list of metadata

https://pastel.archives-ouvertes.fr/tel-03624997
Contributor : ABES STAR :  Contact
Submitted on : Wednesday, March 30, 2022 - 3:51:09 PM
Last modification on : Friday, August 5, 2022 - 2:54:01 PM

File

107866_MAZAR_2022_archivage.pd...
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-03624997, version 1

Collections

Citation

Merouane Mazar. Simulation et optimisation de la gestion dynamique de tâches évolutives sur des robots mobiles autonomes.Application sur le traitement UV-C robotisé dans l’horticulture.. Génie des procédés. HESAM Université, 2022. Français. ⟨NNT : 2022HESAE005⟩. ⟨tel-03624997⟩

Share

Metrics

Record views

85

Files downloads

12