SBIR-STTR Award

Ballistic missiles pose formidable challenges for target tracking and intercept because of their spiraling dynamics and significant maneuverability potential as they re-enter the atmosphere at high speeds. Recent studies have shown that conventional methods to estimation, guidance and control design are unable to guarantee sufficient accuracy against such highly maneuvering targets. This is not unexpected since the coupling between the individual elements is not accounted for completely using the conventional design paradigm, and the separation principle which is implicit in such designs can break down in engagements that require high maneuverability. To meet these challenges we propose developing i) adaptive integrated guidance and control (G&C) designs for increased performance and robustness to uncertain interceptor dynamics, ii) adaptive target-state estimation designs for increased performance and robustness against maneuvering targets, iii) an integrated design method for combining adaptive target-state estimation and adaptive G&C designs for improving overall system performance, and iv) collision avoidance methods to avoid mutual kill-vehicle collisions. The adaptive methods are based on flight proven Neural-Network based adaptive control algorithms. We expect that our strategy for integrated adaptive designs can produce never before achieved levels of lethality for the intercept of future maneuvering targets.

Keywords:
Adaptive Guidance, Adaptive Control, Adaptive Estimation, Missile Defense, Target Tracking, Integrated Estimation, Guidance And Control, Missile Intercept

Existing approaches to guidance and estimation for intercept of maneuvering targets are often dependent on assumed models of maneuvering target behavior. We have shown in Phase I that intercept lethality can be dramatically reduced when an interceptor guidance strategy optimized for a specific maneuvering target type encounters a target whose behavior differs from that of the assumed model. We then demonstrated in simulation the ability to augment a baseline combination of guidance and estimation with an adaptive element. Through rapid adaptation, the augmented system performs well against a variety of maneuvering targets. The adaptive approach appears to be especially beneficial when the inceptor does not enjoy a large acceleration advantage over the target. In partnership with Lockheed Martin, the phase II program shall (1) refine the adaptive approach developed in Phase I; (2) provide quantitative demonstration of improved or same Agile Kill Vehicle lethality compared to traditional guidance laws pitted against uncertain maneuvering threats in high fidelity simulation, and do so without dependence on models of target behavior; and (3) provide qualitative demonstration of the potential to further optimize overall system design by achieving a reduction in interceptor acceleration capability required to maintain required interceptor lethality against uncertain maneuvering targets.

Keywords:
Adaptive Guidance, Neural Networks, Output Feedback Adaptive Control, Agile Kill Vehicle, Maneuvering Targets, Ballistic Missile Defense