Objectif In many applications, systems include both continuous and discrete components, discrete characteristics being also often introduced by the control system or its specifications. Such systems are commonly referred to as 'hybrid systems', and their understanding is rather limited at present. In practice, the control of hybrid systems is left to schemes based on heuristics rules deduced from practical plant operations. The aim of this work is to investigate the field of 'Mixed Logic Dynamical' systems, which seem an appropriate tool to model hybrid systems, then to examine related problems such as control, estimation and verification. Some current ideas for control and estimation are based on receding horizon predictive control and moving horizon state estimation. The 'three repositories system' was chosen M an implementation for the hybrid systems theory and the 'Mixed Logical Dynamical' models, because this case study can provide the necessary complexity for a simulation purpose. Starting from the theoretical results in Part 1, the aim of this work was to modify the simulations in order to feet a real case, a case of agricultural water management with three irrigation lakes. The objective wag to verify the opportunity of a control philosophy for the levels of the lakes based on the hybrid model. This can be further extended to more complex systems in water resources modeling, giving the chance of using flexible control laws in the management process, Taking into account the high computational load, a special attention was given to the optimization routine with the objective in finding a competitive methodology of solving 'Mixed Integer Quadratic Problems', the type of problem that is encountered In the predictive law for hybrid system in MLD form. Champ scientifique engineering and technologyelectrical engineering, electronic engineering, information engineeringelectronic engineeringcontrol systemsnatural sciencesearth and related environmental scienceshydrologyengineering and technologyenvironmental engineeringnatural resources managementwater managementnatural sciencescomputer and information sciencesartificial intelligenceheuristic programmingsocial scienceslaw Programme(s) FP5-HUMAN POTENTIAL - Programme for research, technological development and demonstration on "Improving the human research potential and the socio-economic knowledge base" (1998-2002) Thème(s) Data not available Appel à propositions Data not available Régime de financement ACM - Preparatory, accompanying and support measures Coordinateur OLARU SORIN Contribution de l’UE Aucune donnée Adresse c/o Supelec (Ecole Supérieure d'Eléctricité), Serv 91192 GIF SUR YVETTE France Voir sur la carte Coût total Aucune donnée