Fast Position Control of a Large Radio Telescope's Secondary Mirror Using Iterative Learning

In this paper, we address a complex motion control problem of fast and precise positioning of the secondary mirror in a large radio telescope, which exhibits strong nonlinear behavior and is greatly influenced by gravity and wind. We optimally integrate cascade feedback control, kinematic feedforwards, and iterative learning control to achieve the required positioning accuracy in minimum time. The kinematic feedforwards are generated by a modified online jerk-limited trajectory generator with high numerical accuracy. This novel triple combination marks a first in telescopes motion control, yielding exceptional positioning performance as confirmed by experimental results.