Pursuit-evasion game theoretic uncertainty oriented sensor management for elusive space objects

摘要

A pursuit-evasion (PE) orbital game approach for space situational awareness (SSA) is presented to deal with imperfect measurements and informational uncertainties. In the two-sided optimization problem, a pursuer (an observer or sensor) will use the sensor resource to minimize the uncertainty (modeled by entropy) while an evader (a space object being tracked) will maximize it by performing space maneuvers. Since the cost and opportunity cost of sensor resources, pursuer will make decisions on the when to use these resources. The proposed PE approach provides a method to solve the SSA problem, where the evader will exploit the sensing and tracking model to confuse the opponent by corrupting their tracking estimates, while the pursuer wants to efficiently decrease the tracking uncertainties. A numerical simulation scenario with one space based space surveillance (SBSS) satellite as a pursuer and one geosynchronous (GEO) satellite as an evader is simulated to demonstrate the PE orbital game approach. The GEO applies the continuous low-thrust such as the Ion thrust in maneuvers. An add-on module is developed for SGP4/SDP4 algorithms to propagate the satellites with maneuvers. The GEO maneuvering strategies and on-off measurement controls for SBSS are obtained from the Nash equilibrium of the PE game.