Study of terahertz ellipsoidal lens assembly method based on surface fitting

The assembly and measurement of quasi-optical feed network has the characteristics of small size, too many components, compact and complex structure. The measurement and assembly process of quasi-optical feed network has the requirements of high accuracy and high reliability. In this paper, focalizing on problems such as great error in the terahertz quasi ellipsoid lens optical feed networks, slow speed and poor reliability in the traditional assembly accuracy. Putting forward a kind of least-squares surface fitting method with parameter constrains, getting the ellipsoidal mirror, long half shaft and other characteristic parameters, the relative error of fitting characteristic parameters is within 0.03%. We adjust the position of the terahertz quasi ellipsoid from lens optical feed networks based on these characteristic parameters. The experimental results show that, by adjusting the proposed surface fitting algorithm, the assembly efficiency of the terahertz quasi-optical feed network is significantly improved compared with the traditional manual adjustment method, and the assembly error of the terahertz quasi-optical network can be effectively reduced. By the end, the high precision assembly of the ellipsoidal mirror of the quasi-optical feed network can be realized.