During this PHD, we have studied two technological solutions being involved for solving energy, environmental and economic problems related with transport area. The first solution treats a new concept based on the development of a powertrain plant with on board Fuel Cell and on board reformer. The second solution is based on the evolution of the classic engine. Thus, it takes advantage of several previous developments. In this study, we propose control laws and diagnosis in order to optimize the efficiency of the powertrain plant under autonomy, pollution and cost constraints. This led us to develop reduced models of physical and chemical phenomena involved on powertrain components. When studying the two systems, we found several structural similarities which have enabled us to formalize a common control problem. Than we proposed a general almost optimal control law under the constraints set above. The different models are validated thanks to experimental tests. Algorithms of control and diagnosis are validated in simulation and using prototyping tools. Key words Automatic : Modelling, Model reduction, Control, Diagnosis Automotive : Powertrain, Fuel Cell, Reformer, Condenser, Oxidation catalysis, Particulate filter Others : Electrochemistry, Thermal, Catalysis, Combustion, Condensation, Energy management
