Accelerated formulation of optimal control law for adaptive cycle engines: A novel design methodology

Abstract

Adaptive cycle engines (ACEs) represent a sophisticated class of power units, embodying numerous variable geometry components, marking them as pivotal for the advancement of next-generation high-speed civil and military engines. Nonetheless, this variability introduces substantial complexity into the engineering of control laws. Contemporary research predominantly relies on either manual methodologies or global optimization algorithms. The former rarely achieves peak performance, while the latter incurs substantial computational expenses. This research introduces an accelerated methodology by using variable-step gradient with boundary constraints (VGB) algorithm tailored for the nuanced optimization challenges posed by multi-variable, multi-constraint nonlinearities inherent in ACEs. Recognizing that ACE's peak performance is typically boundary-constrained, the algorithm strategically manipulates variable geometries to widen the operational gap from the most disadvantageous boundary (MDB), thereby inching closer to the engine's ideal operational nexus. This strategy culminates in a decease in computational overhead, diverging significantly from conventional global optimization approaches. Empirical findings represent the method's efficacy, with a maximal performance deviation of just 3.78 % across three steady-state missions. Post transient control law refinement, engine's starting time contracted by 5.93 s via VGB algorithm, opposed to baseline design point configurations. Collectively, this approach serves as a versatile and efficient blueprint for designing control laws.