A multi-agent guidance scheme for a cooperative inspection of an unknown space object

On-orbit servicing is a rapidly growing mission set concerned with extending the life of aging legacy space systems through refueling, replacing, or upgrading components. On-orbit inspection is a necessary pre-requisite to gain information on a Resident Space Object (RSO) before servicing. To take advantage of the emerging capabilities provided by using distributed systems for inspection, a multi-agent guidance algorithm must be designed with cooperation in mind, while still addressing traditional requirements of a single-agent inspection mission. In this paper, a pseudospectral method is used to design a multi-agent inspection mission with both equality and inequality constraints and a multi-objective cost function formulation. A novel method for evaluating information gained is proposed that incorporates viewing angle and lighting conditions to evaluate the quality of the information that could be gleaned from the images obtained. In addition, multiple phases are used to sequentially plan the agents to yield a more tractable Optimal Control Problem (OCP). An example problem is posed to demonstrate the efficacy of this method on a multi-agent inspection mission where each agent is only equipped with passive optical sensors and thus are reliant on favorable lighting conditions based on the sun vector. A range of initial guesses are explored to understand convergence characteristics of the proposed OCP formulation and methodology. The resulting trajectory is shown to adhere to all constraints, regardless of implementation as a hard or soft constraint. Values for the performance metrics of the problem, fuel use and information gained, are highly dependent on several weighting factors that are influenced by real world constraints such as sensor parameters and mission duration. Ultimately, the proposed method was shown to be an effective method for guidance of multiple agents inspecting a static RSO considering lighting constraints.