Fail-Safe Vehicle Rendezvous Using Passive Safety Constraints
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current chaser spacecraft rendezvous techniques are inadequate for ensuring safe collisions avoidance with targets in the event of a complete loss of thruster control, particularly on elliptic orbits, due to limitations in existing navigation and control algorithms.
Innovation Solution
A fail-safe control policy utilizing backwards reachable sets and model predictive control (MPC) is implemented to compute unsafe regions around the target, formulating these as passive safety constraints for online trajectory generation, ensuring the chaser spacecraft avoids collision trajectories even in the absence of thruster control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional open-loop trajectory generation with suboptimal closed-loop tracking is used, then the rendezvous process can be completed, but the system lacks fail-safe capability and cannot avoid collision in the event of complete thruster control loss
Solution Approach 1:
The backwards reachable sets are pre-computed offline to identify all unsafe states that would lead to collision if thruster control is lost. These pre-computed unsafe regions are then incorporated as constraints in the MPC controller, allowing the system to guarantee fail-safe capability without requiring complex real-time computation during actual operation.
Solution Approach 2:
The unsafe regions act as an intermediary between the system dynamics and the control input. By formulating these unsafe regions as passive safety constraints in the MPC controller, the system automatically avoids collision trajectories without requiring active intervention or complex real-time safety monitoring systems.
2Reliability
If backwards reachable sets are computed and incorporated as passive safety constraints in MPC, then collision-free trajectories are guaranteed in the event of thruster failure, but the computational burden increases
Solution Approach 1:
The computationally intensive backwards reachable set computations are performed offline before the actual rendezvous operation. The results are stored and then efficiently queried and applied as constraints during real-time MPC control, separating the heavy computational burden from the time-critical control loop.
Solution Approach 2:
The MPC controller dynamically adjusts the trajectory by incorporating the pre-computed unsafe regions as time-varying constraints. The controller optimizes the control input at each time step while respecting the safety boundaries, allowing adaptive response to changing states without re-computing the entire safety analysis.
3Reliability
If passive safety constraints are enforced to avoid unsafe regions, then the chaser spacecraft maintains safe separation from target, but the trajectory optimization becomes more restrictive
Solution Approach 1:
The unsafe regions serve as intermediary constraints that mediate between the control objectives and safety requirements. By formulating these as passive safety constraints in the MPC optimization problem, the trajectory generator automatically finds optimal trajectories that satisfy both performance and safety requirements without requiring separate safety verification steps.
Solution Approach 2:
The MPC controller changes the trajectory parameters dynamically to maintain safe separation from the target while still achieving the rendezvous objective. The passive safety constraints define the boundaries of acceptable parameter values, and the optimizer adjusts the trajectory within these boundaries to optimize performance.
Data Source
AI summary
A system for controlling an operation of a vehicle to rendezvous with a target over a finite time horizon, wherein the vehicle and the target form a multi-object celestial system. A processor to formulate passive unsafe regions as passive safety constraints. The passive unsafe regions represents regions of space around the target guaranteeing collision trajectories with the target, in an event of total thruster failure. Update a controller having a model of dynamics of the vehicle with received data, and subject the updated controller to the passive safety constraints to generate control commands that produce a collision free rendezvous trajectory which avoids unsafe regions for the specified time period, guaranteeing a collision free trajectory with respect to the target in the event of the total vehicle thruster failure, so the vehicle does not collide with the target. Output the control commands to activate or not activate thrusters of the vehicle.