High-level simulation of JWST event-driven operations

摘要

The James Webb Space Telescope (JWST) has an event-driven architecture: an onboard Observation Plan Executive(OPE) executes an Observation Plan (OP) consisting of a sequence of observing units (visits). During normal operations,ground action to update the OP is only expected to be necessary about once a week. This architecture is designed totolerate uncertainty in visit duration, and occasional visit failures due to inability to acquire guide stars, without creatinggaps in the observing timeline. The operations concept is complicated by the need for occasional scheduling of timecriticalscience and engineering visits that cannot tolerate much slippage without inducing gaps, and also by onboardmomentum management. A prototype Python tool called the JWST Observation Plan Execution Simulator (JOPES) hasrecently been developed to simulate OP execution at a high level and analyze the response of the Observatory and OPEto both nominal and contingency scenarios. Incorporating both deterministic and stochastic behavior, JOPES haspotential to be a powerful tool for several purposes: requirements analysis, system verification, systems engineeringstudies, and test data generation. It has already been successfully applied to a study of overhead estimation bias: whetherto use conservative or average-case estimates for timing components that are inherently uncertain, such as thoseinvolving guide-star acquisition. JOPES is being enhanced to support interfaces to the operational Proposal PlanningSubsystem (PPS) now being developed, with the objective of "closing the loop" between testing and simulation byfeeding simulated event logs back into the PPS.© (2012) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.