Abort-safe vehicle rendezvous in case of partial control failure

Patent metadata

Field	Value
Publication number	US-11807404-B2
Application number	US-202016914386-A
Country	US
Kind code	B2
Filing date	Jun 28, 2020
Priority date	Jun 28, 2020
Publication date	Nov 7, 2023
Grant date	Nov 7, 2023

Abstract

Systems and methods controlling an operation of a vehicle in real time to rendezvous the vehicle with a target over a finite time horizon having multiple specified time periods. Select a set of unsafe regions from stored unsafe regions, the set of unsafe regions represents regions of space around the target in which any operation of the PSNO thrusters does not avoid collision with the target, guaranteeing collision trajectories with the target. Formulating the set of unsafe regions as safety constraints, and updating a controller having a model of dynamics of the vehicle with the accepted data. Generating control commands by subjecting the updated controller to the safety constraints to produce a rendezvous trajectory that avoids the set of unsafe regions, guaranteeing an operation of at least the PSNO thrusters, in the event of partial vehicle thruster failure results in a trajectory that does not collide with the target.

First claim

What is claimed is:

1. A system for controlling an operation of a vehicle in real time to rendezvous the vehicle with a target over a finite time horizon, the system in communication with a transceiver that accepts data in real time including values of vehicle states and target states in a multi object celestial system, and a predetermined subset of a number of operational thrusters that is less than a total number of operational thrusters of the vehicle, at a specified time period within the finite time horizon, comprising:
a memory having unsafe regions, the memory configured to store executable instructions; and
a processor configured to execute the executable instructions, at the specified time period to:
identify a target orbit location from the accepted data in real time, access the memory having unsafe regions, to select a set of unsafe regions corresponding to the target orbit location and the predetermined subset of the number of operational thrusters of the vehicle, and wherein the set of unsafe regions represents regions of space around the target in which any operation of the predetermined subset of the number of operational thrusters does not avoid collision with the target, guaranteeing collision trajectories with the target, wherein the set of unsafe regions are determined by computing robust backwards reachable sets of a region around the target;
formulate the set of unsafe regions as safety constraints, and update a controller having a model of dynamics of the vehicle with the accepted data;
generate control commands by subjecting the updated controller to the safety constraints to produce a rendezvous trajectory that avoids the set of unsafe regions, guaranteeing an operation of at least the predetermined subset of the number of operational thrusters, that, in the event of partial vehicle thruster failure, results in a trajectory that does not collide with the target; and
output the control commands to activate or not activate one or more thrusters of the vehicle for the specified time period based on the control commands.

2. The system of claim 1 , wherein a guidance and control computer (GCC) of the controller is in communication with the transceiver and the memory, such that the target orbit is determined based on uploaded ephemeris from a ground station, based on ground data obtained in satellite tracking databases, or estimated from onboard sensor measurements on the vehicle obtained from the accepted data.

3. The system of claim 1 , wherein the target is one of a spacecraft, a celestial body or orbital debris, and a region around the target is one of an approach of an ellipsoid (AE) region or a keep-out sphere (KOS) region or an over-approximation of the target's physical geometry.

4. The system of claim 1 , wherein a region around the target is one of an approach of a polytope (AP) region or a keep-out polytope (KOP) region or an over-approximation of a target's physical geometry.

5. The system of claim 1 , wherein the target is a spacecraft, a celestial body or orbital debris, and the region around the target is one of an over approximation of a physical geometry of the target, or an approach ellipsoid (AE) region, or a keep-out ellipsoid region.

6. The system of claim 1 , wherein the robust backwards reachable sets are computed backwards-in-time from the region around the target, as regions of state-space under which any operation of the predetermined subset of the number of operational thrusters does not avoid collision with the target region.

7. The system of claim 1 , wherein the robust backwards reachable sets are polytopes or zonotopes.

8. The system of claim 1 , wherein the computations of the robust backwards reachable sets of the region around the target are performed offline and stored in memory.

9. The system of claim 1 , wherein the computations of the robust backwards reachable sets of the region around the target are performed online, and in real time based on an estimated position of the target from onboard sensor measurements on the vehicle and stored in memory.

10. The system of claim 1 , wherein the region around the target is time-varying as the target moves along the target orbit such that the robust backwards reachable sets are computed for multiple target positions and target region positions along the target orbit.

11. The system of claim 1 , wherein the controller is a model predictive controller (MPC) that uses a local convexification of unsafe regions to formulate linear safety constraints that are only satisfied when a vehicle state is not inside the set of unsafe regions.

12. The system of claim 11 , wherein the local convexification of the set of unsafe regions is achieved by computing a half space constraint that approximates an unsafe region boundary, such that the computing of the half-space covers a local region of unsafe sets that represents a safety constraint for an online trajectory generation process, whereby enforcing one or more half-space constraints provides safety so that the vehicle state remains in a safe set of safe regions and outside an unsafe set of unsafe regions.

13. The system of claim 12 , wherein the half space constraint is formulated as a chance constraint which requires that the half space constraint be satisfied with at least a priori specified probability level due to an uncertainty regarding a position of the vehicle or the target, and/or an uncertainty of a thruster magnitude or a direction.

14. The system of claim 1 , wherein the updated controller is subjected to the safety constraints by formulating an optimal control problem that includes the safety constraints so that when optimized over a set of admissible control inputs, an optimizer generates the control commands.

15. The system of claim 1 , wherein the control commands are generated as a solution to a model predictive control policy that produces the control commands by optimizing a cost function over a receding horizon.

16. The system of claim 1 , wherein the control commands are generated for each specified time period of multiple specified time periods in the finite time horizon, or generated iteratively over a receding time-horizon, such that at least one iteration includes updating one or combination of the components of the cost function, and weights of the components of the cost function and safety constraints based on a change of a desired operation of the spacecraft.

17. The system of claim 16 , wherein for each iteration at a next sequential specified time period, there are different sets of unsafe regions.

18. The system of claim 1 , wherein the vehicle states and the target states in the multi-object celestial system includes one or combination of positions, orientations, and translational and angular velocities of the vehicle and the target, and perturbations acting on the multi-object celestial system, wherein the vehicle and the target form the multi-object celestial system.

19. The system of claim 18 , wherein the perturbations acting on the multi-object celestial system are natural orbital forces such as solar and lunar gravitational perturbations, anisotropic gravitational perturbations due to a central body's non sphericity, solar radiation pressure, and air drag.

20. The system of claim 19 , wherein the other celestial objects include a primary body such as Earth around which the target orbits, or a primary body such as Earth and a secondary body such as a Moon, so that the target is in a halo orbit, a periodic three-dimensional orbit near one of a L1 Lagrange point

Assignees

• Mitsubishi Electric Res Laboratories Inc

Inventors

• Weiss Avishai
• Di Cairano Stefano
• Marsillach Daniel Aguilar
• Kalabic Uros

Classifications

• B64G1/242 Primary

  Orbits and trajectories · CPC title

• B64G1/2422

  using Lagrange points, e.g. halo orbits · CPC title

• B64G1/244

  Spacecraft control systems · CPC title

• B64G1/646 Primary

  Docking or rendezvous systems (refuelling in space B64G1/4024) · CPC title

• B64G1/1081

  for debris removal · CPC title