Model-based optimal tracking control architecture for ground-based telescopes

This paper focuses on the designing of tracking control strategies for ground-based telescopes by also comparing model-based solutions with more classical alternatives. Within this framework, we synthesize a double-layer control architecture consisting of: i) a position control layer, which combines a Kalman filter observer and Linear-Quadratic-Gaussian-Proportional-Integral (LQG-PI) controller to compute the appropriate speed profile guaranteeing a reliable tracking of a given telescope position trajectories; ii) a speed control layer, which ensures the optimal tracking of the computed speed profile by driving the torque of the telescope. Moreover, a trapezoidal speed pre-processor is embedded in our control architecture with the aim of computing the appropriate telescope axes position trajectories: this ensures that all the telescope physical constraints, in terms of speed and acceleration, are not always violated. Virtual simulations, carried out via an ad-hoc simulation platform, implemented in Matalb&Simulink and tailored for the specific case study Telescopio Nazionale Galileo (TNG) located at La Palma island, disclose the effectiveness of the hierarchical control architecture for a representative set of star trajectories. Validation phase also considers several realistic conditions and takes into account input disturbance such as the Von-Karman wind disturbance model. Finally, a comparison analysis with a PID-based control architecture is provided to discuss about the advantages and benefits of the proposed optimal control solution.