Andre Antunes de Sa, Kayhan Space; Matthew Shouppe, Kayhan Space; Shota Takahashi, Kayhan Space; Felix Portillo, Kayhan Space; Masafumi Isaji, Kayhan Space; Siamak Hesar, Kayhan Space Corp; Sita Sonty, Boston Consulting Group
Keywords: SSA, STM, optimal collision avoidance, coordinated avoidance maneuver
Abstract:
The continued increase of resident space objects (RSOs) poses a serious space congestion problem, driving an aggressive growth of on-orbit collision likelihood and endangering the potential of tomorrow’s space economy – one that is forecast to become a trillion dollar industry by 2040. Yet, the methodologies currently employed by satellite operators for collision risk mitigation are often inadequate, non-uniform, uncoordinated and are increasingly vulnerable to being overcome by events. Operators have not reached a consensus on collision avoidance risk metrics and actionable thresholds. There are no explicit norms of behavior for risk mitigation and coordination between operators, and space traffic management (STM) authority, regulation, and policy remain in their infancy. STM needs have grown to a global scale, involving multiple nation’s assets in orbit. Such high entropy in the collision avoidance mitigation processes is often the main cause of costly inefficiencies at best, and potential damage to in-orbit assets at worst. Spacecraft tracking uncertainties and inaccuracies associated with collision avoidance risk metrics have been connected to higher avoidance maneuver fuel usage and greater number of avoidance maneuvers. Such increased resource usage bears financial cost to each operator, not to mention the operators whose assets might be rendered vulnerable by others’. Operational uncertainty of the conjuncting (secondary) object, due to uncoordinated avoidance or orbit maintenance maneuvers, can have a similar effect. Most concerning, sub-optimal and uncoordinated maneuvers between two operator’s assets can increase the risk of collisions, and they also tend to increase encounter rates and the need for further avoidance maneuvers during a spacecraft’s lifetimes. Lack of process automation significantly increases response time of collision avoidance and further complicates coordination, which can grow quickly in a scenario involving multiple organizations. The cost of these inefficiencies includes increased fuel expenditure, increased regulatory fees, the potential cost of litigation/mediation, decreased revenues from utilization of vulnerable assets, increased mission downtime as more time is spent performing avoidance maneuvers, and decreased mission lifetime due to faster fuel depletion. Collision risk mitigation will soon become too costly, unless the risk metric uncertainties and deficiencies in avoidance maneuver selection, including lack of coordination, are simultaneously addressed. To help address the latter, we investigate a novel optimal avoidance maneuver framework.
Optimal avoidance maneuver planning is a complex problem, especially when automation and spacecraft coordination are required. The current diversity of collision avoidance processes employed by satellite operators is driven by their different mission requirements, long-time risk exposure, and corporate and cultural considerations, among others. Thus, any viable avoidance maneuver framework must be flexible and customizable, considering the mission and/or organization needs of each stakeholder. In this context, an optimal avoidance maneuver solution is one that best fulfills the interests of an organization, or of multiple organizations when coordination is considered. The framework must also level the playing field with respect to the scale of financial asset loss/risk encumbered by each respective party through means of pricing/financial risk assessment.
Kayhan Space has developed, and continues to update, an optimal avoidance maneuver framework to address the needs identified above. It combines a user-facing platform for asset automation configuration with an expansive maneuver suggestion engine. This maneuver suggestion engine is a critical part of the avoidance maneuver framework, and it is the main focus of this paper. Brief commentary on the Kayhan Space avoidance maneuver framework as part of a broader solution to the current STM needs of the space industry, including incorporation of rules of the road and data sharing, are also provided. The maneuver suggestion engine has a modular design built to be highly flexible, expandable, efficient, and parallelizable. It generates maneuver tradespaces for a conjunction event based on variety of available settings and optimization algorithms. These algorithms all accept a polymorphic object-oriented optimization problem model as guiding input, featuring plug-and-play metrics that can be used as objective functions or as constraints, e.g., miss distance, probability of collision (PC), maneuver direction, shared-burden between spacecraft, etc. The combination of these metrics allows for capturing the operational constraints and interests of spacecraft owners and producing optimal avoidance maneuver suggestions. A variety of optimization algorithms are supported, including nonlinear programming (NLP) using the Interior Point Optimizer (IPOPT) solver, global grid search, heuristic and metaheuristic methods such as ant colony optimization, gradient descent with Nesterov-momentum, and a trivial loop-back fixed solution. Some of these metrics and algorithms are described in this paper, and their performance for different optimization problem case studies are characterized both in terms of processing speed and optimality. The case studies include a combination of PC, maneuver magnitude and direction, and shared-burden minimization and constraint targeting. Two lower-fidelity modelling techniques commonly employed in avoidance maneuver generation are also characterized, specifically (1) the use of linearized dynamics to compute spacecraft relative position changes at time of closest approach (TCA) from a preceding maneuver, as well as (2) the use of risk metrics computed at the original TCA as a proxy for metrics corresponding to the new TCA, which is affected by the avoidance maneuver. Planned future research efforts include: characterizing the entire avoidance maneuver framework, furthering its automation capabilities, supporting multi-objective function in the maneuver suggestion engine, and expanding the supported risk metrics and algorithms.
Date of Conference: September 19-22, 2023
Track: Conjunction/RPO