A. El Kamel

Assignment and Scheduling in Flexible Job-Shops by Hierarchical Optimization

In this paper, we propose a new hierarchical method for the flexible job-shop scheduling problem (FJSP). This approach is mainly adapted to a job-shop problem (JSP) with high flexibility and is based on the decomposition of the problem in an assignment subproblem and a sequencing subproblem. For the first subproblem, we propose two methods: the first one is based successively on a heuristic approach and a local search; the second one, however, is based on a branch-and-bound algorithm. The quality of the assignment is evaluated by a lower bound. For the second subproblem we apply a hybrid genetic algorithm to deal with the sequencing problem. Computational tests are finally presented.

Fuzzy controller for lateral guidance of busses

A knowledge-based system involving fuzzy control for lateral guidance of buses during final approach is presented. A nonlinear dynamical model with three degrees of freedom is developed and completed by a model derived from expertise which expresses the behavior of the conductor while berthing alongside the quay. The purpose of the study is to improve the accessibility of people with reduced mobility to classical buses by implementing real-time control strategies.

Multimodel multicontrol decision making in system automation

The purpose of this paper is to take advantage of the multimodel multicontrol approaches in system automation. A comparative study, based on a nonlinear system, is presented and results are discussed. It is shown, in particular, that the choice of control strategies is a decision making oriented problem.

A Reinforcement Learning Method of Obstacle Avoidance for Industrial Mobile Vehicles in Unknown Environments Using Neural Network

This paper presents a reinforcement learning method for a mobile vehicle to navigate autonomously in an unknown environment. Q-learning algorithm is a model-free reinforcement learning technique and is applied to realize the robot self-learning ability. The state-action Q-values are traditionally stored in a Q table, which will decrease the learning speed when large storage memory is needed. The neural network has a strong ability to deal with large-scale state spaces. Therefore, the neural network is introduced to work with Q-learning to ensure the self-learning efficiency of avoiding obstacles for industrial vehicles in unpredictable environments. Experiment results show that an autonomous mobile vehicle using the proposed method can successfully navigate to the target place without colliding with obstacles, and hence prove the self-learning ability of navigation in an unknown environment.

On the residue formulation in multimodel control

The evaluation of validities is an important step in the multimodel control generation procedure. These trust coefficients are generally derived from residues comparing the behavior of the different base models toward the processed one. The present paper introduces new types of residues using either the horizon notion or the state estimation or both at the same time. Results of their use in controlling a flexible transmission an discussed. Another more complex system is considered in order to show the performance limitation of some of these residues.

Multimodel control using fuzzy fusion

The use of the multimodel approach in modelling, analysis and control of non-linear complex and/or ill-defined systems has been advocated by many researchers. This approach supposes the definition of a set of local models valid in a given region or domain. Different strategies exist in the literature and are generally based on a partitioning of the non-linear systems full range of operation into multiple smaller operating regimes each of which is associated with a locally valid model or controller. However, most of these strategies, which suppose the determination of these local models as well as their validity domain, remain arbitrary and are generally fixed thanks to a certain a priori knowledge of the system whatever its order. We recently proposed a new approach to derive a multimodel basis which allows us to limit the number of models in the basis to almost four models. Meanwhile, the transition problem between the different models, which may use either a simple commutation or a fusion technique, may still arise. In this paper, a fuzzy fusion technique is presented with the following main advantages: use of a fuzzy partitioning in order to determine the validity of each model which enhances the robustness of the solution; introduction of, besides the four extreme models, another model, called the average model, determined as an average of the boundary models. The theoretical study is validated by experimental results.

Multi-agent based architecture for virtual reality intelligent simulation system of vehicles

Most existed researches on traffic simulation with virtual reality are mainly concentrated on the visualization or the real-time simulation to create immersion experiences in virtual world, but little attention has been paid on system modeling and formal specification of traffic simulation. Based on multi-agent technology, Virtual Reality Intelligent Simulation System of Vehicles (VR-ISSV) is proposed to investigate the modeling method of vehicles simulation system, with the advantages of reusability, scalability and flexibility. Firstly, the framework modeling is introduced to depict the general intelligent simulation system. Then, intelligent vehicle agent and environment agent are presented for the simulation of interactions among the vehicles, the traffic situation and the environment. Finally, the application results of the proposed system are realized.

Total Tardiness in a Flexible Job-shop

We propose in this paper different approaches for the flexible job shop scheduling problem. The chosen criterion is the total tardiness due to the importance of respecting the due dates in production process and in the whole supply chain. The first approach is based on the decomposition of the problem in an assignment problem and a sequencing problem. The first part is ensured by an assignment technique based on a heuristic approach and the definition of an intermediary criteria. The second part consists in applying a genetic algorithm to deal with the sequencing problem. The second approach is rather an integrated one based on genetic algorithm and the generalization of a coding used for the classic job-shop

Design of multicontrollers for smooth transition between multiple operating regimes

A multiple-model based controller i s designed for transition control when a plant owns several nominal worlting points on a static curve. The multiple-model controller ensure a smooth transition between two neigbboring working points, using am adequate combination of the local controls. When the working points are distant. a supervisor will switch between multi-model controllers so that the trajectory will pass close to intermediate nominal points and gradually reach the final set-point. The local and transition controls validity domains are computed using a method of comparison systems and Vector Norms.

Multimodel analysis and control of multivariable systems

The use of the multimodel approach in the modeling, analysis and control of nonlinear complex and/or ill-defined systems was advocated by many researchers. This approach supposes the definition of a set of local models valid in a given region or domain. Different strategies exist in the literature and are generally based on a portioning of the nonlinear systems full range of operation into multiple smaller operating regimes each of which is associated with a locally valid model or controller. The global control may be a fusion between the partials laws. The purpose of the paper is to pay attention to interesting aspects of the multimodel approach thus revealing some answers to problems not yet resolved for multivariable systems.