Modeling and analysis of spacecraft power systems

Abstract

A comprehensive large-scale power system modeling is developed to facilitate the design and analysis of present and future spacecraft power systems. A two-port coupling method is utilized to provide a modularity in model building and analysis of the system. The modular approach allows the model to be flexible, verifiable and computationally efficient. A methodology for the system level analysis is presented with the ability to focus on the performance characteristics of an arbitrary component or subsystem. The system performance parameters are derived explicitly in terms of the two-port hybrid g-parameter representation of the component or subsystem, and impedances of its terminating subsystems. From this, the stability of the system is analytically determined and the subsystem interaction criteria is observed. Also presented is a model development from the empirical data employing the complex curve fitting technique. The technique is especially powerful for large scale system modeling and analysis where certain components and subsystems are viewed as black boxes with measurable terminal characteristics. The technique can also be used to realize a reduced order model of a complex subsystem.

The Direct Energy Transfer (DET) spacecraft power system is modeled to demonstrate the versatility of the comprehensive system model by performing various DC, small-signal and large-signal analyses. Of particular interest is the analysis of the large-signal behavior of the nonlinear solar array system by employing the state-plane method. The analysis of the solar array system operation focused on the transition mode between the shunt mode and the battery discharging mode is presented. The subsystem interaction problems in the local component and global system are illustrated. A methodology for the design and trouble-shooting of a system dealing with the interaction problems using the g-parameters is described. Finally, a system level analysis of the DET system using an empirical data modeling technique is performed.