A stochastic global identification framework for aerospace structures operating under varying flight states

Abstract

In this work, a novel data-based stochastic "global" identification framework is introduced for air vehicles operating under varying flight states and uncertainty. In this context, the term "global" refers to the identification of a model that is capable of representing the system dynamics under any admissible flight state based on data recorded from a sample of these states. The proposed framework is based on stochastic time-series models for representing the system dynamics and aeroelastic response under multiple flight states, with each state characterized by several variables, such as the airspeed, angle of attack, altitude, temperature, etc., forming a flight state vector. The method's cornerstone lies in the new class of Vectordependent Functionally Pooled (VFP) models which allow the explicit analytical inclusion of the flight state vector into the model parameters and, hence, system dynamics. This is achieved via the use of functional data pooling techniques for optimally treating -as a single entity-the data records corresponding to the various flight states. In this proof-of-concept study the flight state vector is defined by two variables, namely the airspeed and angle of attack of the vehicle. The experimental evaluation and assessment is based on a prototype bio-inspired self-sensing composite wing that is subjected to a series of wind tunnel experiments under multiple flight states. Distributed micro-sensors in the form of stretchable sensor networks are embedded in the composite layup of the wing in order to provide the sensing capabilities. Experimental data collected from piezoelectric sensors are employed for the identification of a stochastic "global" VFP model via appropriate parameter estimation and model structure selection methods. The estimated VFP model parameters constitute two-dimensional functions of the flight state vector defined by the airspeed and angle of attack. The identified model is able to successfully represent the wing's aeroelastic response under the admissible flight states via a minimum number of estimated parameters compared to standard identification approaches. The obtained results demonstrate the high accuracy and effectiveness of the proposed global identification framework, thus constituting a first step towards the next generation of "fly-by-feel" aerospace vehicles with state awareness capabilities.

Intelligent Flight State Identification of a Self-Sensing Wing through Neural Network Modelling

“…RBM can be regarded as an undirected graph including a visible layer for observed data and a hidden layer for feature detection. The energy function is defined as (6) where v, h indicate node values in the visible layer and the hidden layer, W is the weights between the visible layer and hidden layer, a,b denote the bias of the visible layer and the hidden layer. θ is the set of parameters including W, a and b.…”

“…PZT has both active and passive measurements. For active mode, it can be used for damage detection and structural health monitoring while in passive mode, the wing structural vibration during flying can be captured to reflect the air dynamic characteristics [6].…”

Gaussian Process Time-Series Models for Structures under Operational Variability

“…Typically, such a deterministic functional relationship is captured via a functional series expansion. Methods falling into this class include the regression/interpolation methods discussed in Worden et al (2002) and Sohn (2007), as well as the Functionally Pooled (FP) time-series models explained in Kopsaftopoulos et al (2018) and Sakellariou and Fassois (2016). These methods are particularly effective when a direct relationship exists between measurable input EOPs and the characteristic quantities of the time-series models.…”

Data-driven State Awareness for Fly-by-feel Aerial Vehicles: Experimental Assessment of a Non-parametric Probabilistic Stall Detection Approach

“…[1], (ii) sense their flight and aeroelastic state (airspeed, angle of attack, flutter, stall, aerodynamic loads, etc.) and F. Kopsaftopoulos internal structural condition (stresses, strains, damage) [2][3][4], and (iii) effectively interpret the sensing data to achieve real-time state awareness and health monitoring [5][6][7][8][9]. Towards this end, novel data-driven approaches are needed for the accurate interpretation of sensory data collected under varying flight states, structural conditions, and uncertainty in complex dynamic environments.…”

“…The method is based on the statistical analysis of the response signals recorded from piezoelectric sensors integrated with the aircraft structure and the subsequent use of decision making schemes to detect stall within a probabilistic framework. The experimental evaluation and assessment is based on a prototype bio-inspired self-sensing composite wing that is subjected to a series of wind tunnel experiments under multiple flight states [2,4].…”