Optimization of acceptance test plan for nanosatellites to minimize constellation lifecycle cost

• Main Author: Teo, Kah How
• Other Authors: Tai Kang
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Mechanical engineering
• Online Access: https://hdl.handle.net/10356/158990
• Institution: Nanyang Technological University

The successful demonstrations of nanosatellite technologies in recent years have brought increased attention to the benefits of nanosatellite constellations. However, it is unclear how the high volume production and testing of nanosatellites should be carried out such that the lifecycle cost of the constellation can be minimized. Burn-in can increase operational reliability and satellite lifetimes, yielding satellites that fail less frequently. However, subsequent replacements are more costly, with replacements also delayed if burn-in factors into the replacement duration. Hence there is a need to study the effects of burn-in on the lifecycle cost of the constellation. In this thesis, a burn-in cost minimization model has been developed with operational considerations for the whole constellation, to generate burn-in plans that are optimal for the mission duration of the constellation. This optimization model combines a “traditional” burn-in cost model (considering burn-in and warranty activities) with an availability model, to obtain a comprehensive model that describes the lifecycle costs of each satellite slot in the constellation. The model considers the nanosatellite as a system that is repairable during burn-in but subsequently is unrepairable in operation. The ``traditional” lifecycle cost effects of burn-in at both the component and system levels for these operationally unrepairable systems are studied. It is more advantageous to conduct system burn-in compared to component burn-in, for systems with many components. The reverse is true for systems with fewer components. In addition, there is a minimum system burn-in duration before cost savings are achieved. In general, there is no simple way to prescribe how burn-in should be carried out precisely. The availability model developed in this work treats outages from different sources as independent alternating renewal processes. Validation using operational data of the GPS 24 satellite constellation showed that the model is effective in modelling various important values such as constellation state probability and the number of satellite replacements needed over any given period of time. The proposed availability model avoids using the Markov assumption underlying most state-of-the-art methods and allows general reliability forms to be used as input. Finally, the burn-in cost model and the availability model are brought together for the optimization of mission lifecycle costs of a constellation. Different mission applications require different objectives to be minimized, so both single objective cost minimization problems and multi-objective cost and unavailability minimization problems are presented. The effects of system homogeneity and the use of sparing strategies are also investigated. Genetic algorithms were used as the optimization method, and the optimal solutions obtained for several example problems provided much insight into how burn-in should be carried out for minimizing total costs. The optimization is also applied to a case study of a proposed mission scenario and the constellation lifecycle costs of an optimized burn-in plan are compared to those of a typical “reliability target” burn-in plan. The optimization resulted in lifecycle cost savings of around 19% while significantly reducing the burn-in duration, showing much promise in the burn-in cost optimization approach.