Rigid-body attitude stabilization with attitude and angular rate constraints

Reinforcement Learning-Based Approximate Optimal Control for Attitude Reorientation Under State Constraints

“…Remark 1: It is noteworthy that, M ij in (6) a constant matrix. This is distinct from the relevant results in [9] and [10]; and it is achieved through fully utilizing the invariance property of α b i and β r j . Moreover, this property can greatly benefit the design of the subsequent approximate optimal controller, since it renders V a to be a function with the unique argument q e .…”

“…By employing artificial potentials (APs), Lee and Mesbahi [9] designed a feedback attitude reorientation control law subject to multiple attitude constraint zones. Shen et al [10] extended this result by considering additional angular-velocity limits. Some other constrained feedback controllers for six-degreeof-freedom problems were given in [15], [16].…”

“…2) The proposed controller has significant advantages over the non-RL feedback control methods for constrained attitude reorientation problems [9], [10]. It has the ability to make a trade-off between control cost and performance.…”

“…Here, mainly two types of state constraints are considered. First, special payloads, such as infrared telescopes, must be kept away from direct exposure to the sunlight or other bright celestial objects [9], [10], which forms a set of attitude forbidden zones. Second, per safety concerns, the angular velocities need to be restricted.…”

“…Second, per safety concerns, the angular velocities need to be restricted. For example, NASA's X-ray Timing Explorer (XTE) requires that the angular velocities should always be within the saturation limit of its rate gyros during reorientation actions [10], [11].…”