Statistical approach to systems engineering for the Thirty MeterTelescope

摘要

Core components of systems engineering are the proper understanding of the top level system requirements, their allocation to the subsystems, and then the verification of the system built against these requirements. System performance, ultimately relevant to all three of these components, is inherently a statistical variable, depending on random processes influencing even the otherwise deterministic components of performance, through their input conditions. The paper outlines the Stochastic Framework facilitating both the definition and estimate of system performance in a consistent way. The environmental constraints at the site of the observatory are significant design drivers and can be derived from the Stochastic Framework, as well. The paper explains the control architecture capable of achieving the overall system performance as well as its allocation to subsystems. An accounting for the error and disturbance sources, as well as their dependence on environmental and operational parameters is included. The most current simulations results validating the architecture and providing early verification of the preliminary TMT design are also summarized.