Comparisons of direct and adaptative optimal controls for interior permanent magnet synchronous integrated starter generator

Integrated starter generator (ISG) applications impose very high constraints. So, classic control laws, based on linear models become inadequate. This paper describes two optimal controls (total losses minimisation) based on a numerical, nonlinear constrained optimisation routine. Optimal computed currents are applied as current references to a machine with various temperatures. This paper shows that adaptative optimal control (AOC) is required to maintain performances of ISG. Without the AOC, efficiencies are reduced but the main problem is a drastic reduction of the maximal mechanical power output in the motor mode.