Description: Advanced trajectory design and estimation capabilities in complex nonlinear dynamical regimes represent two of the greatest technical challenges of modern space flight. In addressing these challenges, DECISIVE ANALYTICS Corporation seeks to advance the capabilities of NASA's open source General Mission Analysis Tool to integrate the latest advances in trajectory path planning and estimation. This includes the development of an Advanced Path Planning (APP) plugin that leverages concepts from dynamical systems theory, multi-phase targeting, and visualization for trajectory design in regions where multi-body effects are significant, such as near the libration points. Parallel to that is the development of an Advanced Estimation (AE) plugin, which leverages the results of past studies done at DECISIVE analytics for the Missile Defense Agency and the US Air Force. The proposed AE plugin will be designed around a Hybrid Dynamic Bayesian Network framework, pioneered by DECISIVE ANALYTICS, which will enable advanced estimation capabilities including Unscented Kalman Filters and Gaussian Mixture Models. These two techniques, particularly Gaussian Mixture Models, offer enhanced predictive capabilities for the determination of the true probability density when nonlinearities significantly influence the estimation process. Phase 1 will focus on the software development, integration, testing, and validation of initial prototypes for both plugins.

Benefits: The Advanced Path Planning and Estimation plugins, proposed for integration into NASA's open source General Mission Analysis Tool (GMAT), will offer end-users enhanced mission design capabilities for analysis in libration point regimes, or in regions where multi-body effects become significant. This includes missions that explore the vicinity of small celestial bodies, such as asteroids. Parallel to this development, the Advanced Estimation plugin, also proposed for integration into GMAT, will offer end-users the ability to incorporate, into their analyses, techniques that adequately characterize the effects of nonlinearities in the true posterior probability density during the estimation process. GMAT's current capabilities are limited to simpler extended Kalman filter implementations, which represent the probability density in terms of the first two statistical moments. In contrast, the AE plugin will allow users to consider multi-modal probability density functions, which incorporates Unscented Kalman Filters and Gaussian Mixture Models, among others.

The capabilities provided by the Advanced Path Planning (APP) and Advanced Estimation (AE) plugins are generally applicable outside of NASA's mission. The underlying generalized multi-phase targeting framework within the APP plugin, for instance, is generally independent of the dynamical regime. Thus, the APP binary plugin can be transitioned into other applications related, for instance, to autonomous vehicle guidance. In fact, the initial motivation for the latest theoretical advancements in this framework originate in autonomous vehicle guidance. Furthermore, the AE plugin is built around the Guide and Adapt tools developed by DECISIVE ANALTYICS. These packages, previously developed under various contracts with the Missile Defense Agency and the US Air Force, offer generalized estimation capabilities that are also independent of the dynamical regime. Because we already have customers for the underlying tools, any enhanced capabilities developed under this contract will enable us to provide additional services to them as well.