Line of Sight controller tuning using Bayesian optimization of a high-level optronic criterion

Sophie Frasnedo (1, 2) , Julien Bect (3, 2) , Cédric Chapuis (1) , Gilles Duc (2) , Philippe Feyel (1, 2) , Guillaume Sandou (4, 2)

1 Sagem Defence and Security (Safran)
2 L2S - Laboratoire des signaux et systèmes
3 GdR MASCOT-NUM - Méthodes d'Analyse Stochastique des Codes et Traitements Numériques
4 DISCO - Dynamical Interconnected Systems in COmplex Environments

A method to globally optimize the parameters of the controller of an inertially stabilized platform is presented. This platform carries an electro-optical system. The quality of the produced image is obviously influenced by the capacity of the controller to compensate for the unwanted motion of the platform. The motion Modulation Transfer Function (motion MTF) measures the amount of blur brought into the image by those parasite movements. The controller is tuned by minimizing a criterion which includes the motion MTF. However, evaluating this criterion is time-consuming. Using an optimization method that needs numerous evaluations of the criterion is not compatible with industrial constraints. Bayesian optimization methods consist in combining prior information about the criterion and previous evaluation results in order to choose efficiently new evaluation points and reach the global minimizer within reasonable time. In this paper, a Bayesian approach is used to optimize the motion MTF-based criterion. The results are compared with a local optimization of the same MTF-based criterion, initialized with an acceptable point. Similar performances are achieved by the proposed methodology, without requiring an initialization point.

Mots clés: Inertial Platform; Global optimization; Image motion compensation; Expensive to evaluate functions; Modulation Transfer Function; Parameter optimization

Domaines: Sciences de l'ingénieur [physics]; Automatique / Robotique

Sophie Frasnedo, Julien Bect, Cédric Chapuis, Gilles Duc, Philippe Feyel, et al.. Line of Sight controller tuning using Bayesian optimization of a high-level optronic criterion. 16th IFAC Workshop on Control Applications of Optimization, Oct 2015, Garmisch-Partenkirchen, Andorra. pp.56-61, ⟨10.1016/j.ifacol.2015.11.059⟩. ⟨hal-01259423⟩