Nonlinear stochastic dynamics of detuned bladed-disks with uncertain mistuning and detuning optimization using a probabilistic machine learning tool

A state-of-the-art review on uncertainty analysis of rotor systems

“…The application and enhancement of non-probabilistic approaches should be further exploited for practical situations when data does not support known probability law and avoid introducing too many human assumptions. In addition, recently emerged machine learning methods [296,297] may possibly contribute to the state-of-the-art. However, the development of such methods must go hand in hand with a mastered vision of the most important physical phenomena involved in rotordynamics.…”

Updating Nonlinear Stochastic Dynamics of an Uncertain Nozzle Model Using Probabilistic Learning With Partial Observability and Incomplete Dataset

“…The PLoM algorithm is particularly well-suited for scenarios involving small training datasets, and its efficiency has been demonstrated across various domains. Examples include non-convex optimization under uncertainty [41,42,43,44], model-form uncertainties using random bases [45], and the updating, design, and control of dynamical systems [46,47,48]. The statistical surrogate model will be based on conditional statistics for given control parameter, using the learned realizations obtained from PLoM under the constraints defined by the target.…”

Detuning Optimization of Nonlinear Mistuned Bladed Disks Using a Probabilistic Learning Tool

“…A full data basis is constructed by using a finite element model of a bladed disk with cyclic order 12 [1,2], which allows the random responses of all the possible detuning configurations to be identified [3]. Such a detuning optimization requires to solve an high-dimensional combinatorial optimization problem for which the cost function is evaluated from a nonlinear stochastic reduced computational model (High-Fidelity Computational Model (HFCM)), that has previously been constructed [3,4]. In practical situations, only a small data training set, issued from the HFCM and which does not a priori include any optima, is available.…”

“…It is thus interpreted as an highly nonlinear function of w c, that is to say q c, = f HFCM (w c, ). An available data basis [3] is constructed using the finite element model of the blisk described in [1], yielding 352 detuning configurations that are restricted to the set C c ⊂ N c of the n c =216 detuning configurations having a majority of blades with type 0. The detuning optimization consists in solving the combinatorial optimization problem such that optimum w c,opt un is defined by…”