Hybrid methods for determining time-optimal, constrained spacecraft reorientation maneuvers

Maximum-likelihood estimation optimizer for constrained, time-optimal satellite reorientation

“…Consequently, the direct-method solver must then expend some computational effort to find a feasible solution. As determined in previous work [9,12], this is not always successful or can take considerable time (on the order of 1-12 h).…”

“…While linear penalty functions serve well for handling constraints in conjunction with the heuristic methods [12], the CMA-ES method performs much better with penalties that are quadratic in the amount of the constraint violation. For equality constraints of the form…”

“…These methods are not always fully automatic and may require an intelligent initial guess. Hybrid methods, using heuristic algorithms such as Particle Swarm Optimisation (PSO) [10] and Bacteria Foraging Optimisation (BFO) [11] to generate initial guesses for the states and controls, have been shown to significantly reduce overall computation time [12].…”

Three-axis active magnetic attitude control asymptotical study

“…Fluid rings [1] can be noted among novel actuators. Control may be based on optimal or boundary value problem solutions [2], however this approach is hardly available for small satellite onboard computer. Present paper overcomes the underactuation issue without additional actuators and using rather simple control algorithm.…”

Performance evaluation of the inverse dynamics method for optimal spacecraft reorientation

“…In particular, Bilimoria and Wie investigated the time optimal rest to rest reorientation of a symmetric spacecraft, showing that bang-bang control is optimal and the resulting motion has a significant nutational component [6].methods are based on the numerical integration of the differential equations of motion and provide accurate solutions, but may converge slowly [8,9]. On the contrary, in inverse dynamics the trajectory is approximated by analytical functions.…”