Centre for Sensors, Instruments and Systems Development

Universitat Politècnica de Catalunya

Shaping light to your needs

Foto Foto Foto Foto



Misalignments calculation method in optical systems. Application using neural networks.

On September 27th Esther Oteo Lozano defended his doctoral thesis entitled "Calculation method Misalignments in optical systems. Application using neural networks."

The thesis was directed by Professor Josep Arasa from CD6 (UPC).


In this thesis a new method to determine element misalignments in optical systems is presented. The method starts with the adjustment of Zernike polynomials to the wavefront of the optical system, and sets a transformation function between the Zernike coefficients and the misalignment variables.
Herein we present the mathematical development of the method and three transformation functions: a system of linear equations, a system of nonlinear equations and artificial neural networks. The method is validated by simulation by applying the three transformation functions to an optical system composed of three lenses in which one of them is misaligned. We present and compare the results of applying each function, showing that for simple misalignments the three functions work well, however, for complex misalignments, it has been required to apply the method with the artificial neural networks in two steps, one for decentering variables and one for tilt variables. The method has been generalized for more than one element, showing satisfactory results in the calculation of misalignment for two elements of the triplet system.
CD6 Centre for Sensors, Instruments and Systems Development
Rambla de Sant Nebridi, 10  ·  08222  ·  Terrassa (Barcelona)