Optimization, design and modeling of electrotechnical systems

This document is organized around three main parts. The first part focuses on the problem of optimization and technical-economical modeling of electrical networks, with presentations of the collaborations with the economic world through thesis co-supervision; then, the massive arrival of renewable energies has led the author to study the incorporation of uncertainties in models and optimizations. In a second part, the optimization of motorization systems is considered; after an introduction to the problem and the presentation of the methodology, two examples are developed: an industrial optimization of a range of electric motors, with new techniques to reduce manufacturing and usage costs; this optimization approach for the motorisation chain is based on the complete consideration of the motorization system (motor, gearbox, speed, motion and power converters, energy source). Each component having to be modeled accordingly, a modeling and a first optimization of a power converter is proposed with a view to its incorporation in the overall optimization of the motorization chain. However, to make optimization, models of behavior must be developed, whether for the purposes of minimization as for the constraints. In fact, optimizing is very often positioning itself at the system reliability limits, so it is important to understand the physical and ensure its robustness. This point is the subject of the last part. In a first example, the inflammation of electronic cards submitted to an important stress is considered, coupling experimental results with an electrothermal modeling of the mechanisms leading to the inflammation of printed circuits. Recent advances in discharge physics modeling are presented