Laurent Deroussi

Combining Optimization Methods and Discrete Event Simulation: A Case Study in Flexible Manufacturing Systems

The studied problem is an extension of the well known job shop scheduling problem, in which the material handling system is considered as an overall part of the scheduling. This problem has many applications in flexible manufacturing systems. In these systems, jobs are transported from any machine to any other machine by automated guided vehicles. This problem has both an algorithmic and a structural complexity. We propose to take into account this double complexity by combining metaheuristic (for the optimization process), and discrete event simulation (for the evaluation of a solution). The results obtained show the efficiency of the presented approach, 11 new upper bounds are found on a benchmark of 40 instances of the literature. Realistic constraints like traffic congestion or collision could be included by a more sophisticated evaluation model

Coupling local search methods and simulated annealing to the job shop scheduling problem with transportation

This paper addresses the job and device scheduling problems in flexible manufacturing systems (FMS) using an automated guided vehicle system (AGV) by simultaneously dealing with material processing and transportation functions. The problem is solved using a two stage iterative approach which includes optimization and computer simulation. An iterative procedure is developed. At the first stage, a meta-heuristic determines the AGV schedule, i.e., the order in which the job transfers by AGVs are made. At the second stage, a discrete event simulation model is used to evaluate the makespan depending on the AGVs schedule. This evaluation is used by the meta-heuristic to improve the initial AGV schedule. Finally, the iterative procedure determines jobs and AGV schedules which minimize the makespan (the schedule length). The investigated meta-heuristics are based on the iterated local search method and simulated annealing. An efficient neighboring system used inside meta-heuristic schemes is proposed. Our approach is numerically tested under different experimental conditions.

An iterative two-step heuristic for the parallel drone scheduling traveling salesman problem

A recent evolution in urban logistics involves the usage of drones. In this article, we address a heuristic solution of the parallel drone scheduling traveling salesman problem, recently introduced by Murray and Chu. In this problem, deliveries are split between a vehicle and drones. The vehicle performs a classical delivery tour, while the drones are constrained to perform back and forth trips. The objective is to minimize completion time. We propose an iterative two-step heuristic, composed of: a coding step that transforms a solution into a customer sequence, and a decoding step that decomposes the customer sequence into a tour for the vehicle and trips for the drones. Decoding is expressed as a bicriteria shortest path problem and is carried out by dynamic programming. Experiments conducted on benchmark instances confirm the efficiency of the approach and give some insights on this drone delivery system.

Optimization of Logistics Systems Using Metaheuristic-Based Hybridization Techniques

In the postwar years, the development of operational research provided companies with tools to deal with their logistical problems in a quantitative way. For a long time, these problems were split into unrelated subproblems, each subproblem often being tackled separately. This is mainly due to the fact that the subproblems considered, such as the localization problem, planning problem, scheduling problem, and transportation problem, are generally NP-hard problems and their computational complexity remains a significant issue for many researchers. Nevertheless, in an increasingly competitive industrial environment, companies continue to have a strong demand for decision aid tools to provide a global view of their organization.

Notation and classification for logistic network design models

This paper presents a notation and a classification system for the design models of logistic network. Our notation consists of three fields (analogous to Graham’s α | β | γ notation for scheduling problems). The proposed notation is applied for several articles from the literature. We focus on multi-period models with deterministic and stochastic demands. The proposed notation is based on three criteria corresponding to the main characteristics of the logistic networks: the structure (field α ), the management rules (field β ) and the performance criteria (field γ ). A description of solution methods, datasets and results is also provided. Most articles deal with deterministic, multi-level models and only few of them include the international aspect of logistics, lead-times or subcontracting. Datasets used to test the methods are randomly generated by the authors and have different sizes. The heuristic methods are most commonly used.

Un système de voisinage efficace pour le problème du job-shop avec transport

This paper is devoted to the study of an extension of the job-shop, in which transport of the parts between the machines is taken into account. The objective is then to simultaneously schedule the machines and the vehicles, in order to minimize the makespan. Both problems are known to be NP-hard. To deal with this problem, we propose a new representation of the solutions space, and an efficient neighbouring system. Three different metaheuristics (iterated local search, simulated annealing and their hybridization) have been implemented. The results obtained are better than those in the literature, and this independently from the method used. New upper bounds are proposed for 11 of the 40 instances which compose our benchmark test.