Thierry Galinho

Data representation layer in a MultiAgent decision support system

This article presents a system designed to help deciders manage cases of crisis. The system represents, characterises and interprets the dynamic evolution of a given situation and displays the results of its analysis. The core of the system is made of three multiagent systems (MAS): the data representation MAS for the static and dynamic representation of the current situation, the characterisation MAS for dynamically regrouping sets of agents of the data representation MAS and the interpretation MAS for matching results between the characterisation MAS and previous scenarios stored in the persistent memory of the system in order to have a deeper analysis of the situation. The case based reasoning of the interpretation MAS sends its results to the user as a view of the current situation linked to some views of similar situations. This article will focus on the data representation MAS.

Scheduling of lifting vehicles with time windows in an automated port container terminal

A port container terminal is generally exploited with three types of equipments: Quay Cranes (QCs), vehicles and Yard Cranes (YCs). In this paper, we present a particular type of vehicles which is used at terminal of Normandy in Le Havre port in France. In this terminal, containers are transported from the marshalling yard to ship by the lifting vehicles (LV), capable of raising and of depositing a single container at various heights. This study discusses how to schedule LVs in loading or unloading operation by using information about pickup and delivery locations. Our goal is to organize the transfer of containers in order to minimize the total travel time of all LV, respecting the time window of every LV for a given number of containers. In order to solve this problem we propose a new modeling and a genetic algorithm adapted to the studied problem called “Genetic Algorithm for Loading or Unloading Operation GALUO”. We applied special types of crossover and mutation inspired from various data of our problem. GALUO is compared with a simulated annealing algorithm (SA) inspired from the same problem. Computational results on real instances show the good quality of the solutions obtained by the proposed method.

Ant Colony Optimization for aircraft landings

We address the problem of determining the arrival sequencing and scheduling for aircrafts in air traffic control system. The overall objective of this problem is the minimization of delay cost. The problem is NP-hard for this reason solving the problem optimally is limited to instances of small sizes. In this paper we treat the static case of ASS problem where all data are known in advance. We begin by proposed a mathematical quadratic integer model, with the constraint of limited time window. After we propose a change of variable in order to make linear the initial quadratic constraints. The second target of this work is to resolve the model by an ant colony algorithm. Numerical simulations are applied to real data to show the effectiveness of our new formulation and our approach. They are presented at the end of this article.

Industrial Virtual Environments and Learning Process

Today, we are in the fourth industrial revolution. In this revolution, we include the Industry 4.0. The connectivity of all the objects, in our life or in the industry with the sensors and the actuators connected on the industrial network, create the Industry 4.0.

Dynamic Representation and Interpretation in a Multiagent 3D Tutoring System

In this paper we present an intelligent tutoring system which aims at decreasing students’ dropout rate by offering the possibility of a personalized follow-up. We address the specific problem of the evolution of the large amount of data to be processed and interpreted in an intelligent tutoring system. In this regard we detail the architecture and experimental results of our decision support system used as the core of the intelligent tutor—which could be applied to a variety of teaching fields. The first part presents an overview of the characteristics of intelligent tutors, the chosen data organization—composed of a composite factual semantic feature descriptive representation associated to a multiagent system—and two examples used to illustrate the architecture of our prototype. The second and last part describes all the components of the prototype: student interface, dynamic representation layer, characterization, and interpretation layers. First, for the student interface, the system shows our 3D virtual campus named GE3D to be connected to the intelligent tutor. Then we explain how the agents of the first layer represent the evolution of the situation being analyzed. Next, we specify the use of the characterization layer to cluster the agents of representation layer and to compute compound parameters. Finally, we expose how—using compound parameters—the third layer can measure similarity between current target case and past cases to constitute an interpretation of cases according to a case-based reasoning paradigm.

A linear integer program to reduce air traffic delay in enroute airspace

Due to fast growing of sector of air transportation, air traffic management becomes more and more complex. Therefore, the ability of systems to manage air traffic presents difficulties. In this paper, we address the Air Traffic Flow Management (ATFM) problem, as a new Linear Integer Program (LIP). It takes into account all flights phases, i.e., taking-off, cruising and landing. The model also allows rerouting decisions. All constraints of our model and objective function are linear, differently from the model of Bertsimas et al. [2] which contains non-linear constraints. Finally, numerical simulations are presented at the end of this article showing the effectiveness of our new formulation.

New modelling and exact method for aircraft Arrival Sequencing and Scheduling

In this work, the first target is to model and effectively resolve the Arrival Sequencing and Scheduling problem (ASS). The problem is NP-hard and exists in the Air Traffic Control (ATC) for a long time. It has been the hottest topic in transportation research due to the constantly increasing aviation traffic volume around the world. This paper attempts to design a mathematical integer linear model, with the constraints of limited time windows. We begin by modeling the (ASS) problem by a quadratic integer program and propose a change of variables to linearize the problem, i.e. transforming the quadratic constraints to linear constraints. The second target is to resolve the model by an exact method based on the algorithm of branch and bound. Numerical simulations are applied to real data to show the effectiveness of our new formulation and our approach. They are presented at the end of this article.