Flutter speed prediction by using deep learning

Deep learning technology has been widely used in various field in recent years. This study intends to use deep learning algorithms to analyze the aeroelastic phenomenon and compare the differences between Deep Neural Network (DNN) and Long Short-term Memory (LSTM) applied on the flutter speed prediction. In this present work, DNN and LSTM are used to address complex aeroelastic systems by superimposing multi-layer Artificial Neural Network. Under such an architecture, the neurons in neural network can extract features from various flight data. Instead of time-consuming high-fidelity computational fluid dynamics (CFD) method, this study uses the K method to build the aeroelastic flutter speed big data for different flight conditions. The flutter speeds for various flight conditions are predicted by the deep learning methods and verified by the K method. The detailed physical meaning of aerodynamics and aeroelasticity of the prediction results are studied. The LSTM model has a cyclic architecture, which enables it to store information and update it with the latest information at the same time. Although the training of the model is more time-consuming than DNN, this method can increase the memory space. The results of this work show that the LSTM model established in this study can provide more accurate flutter speed prediction than the DNN algorithm.

Multimodal Estimation of Sine Dwell Vibrational Responses from Aeroelastic Flutter Flight Tests

Aircraft envelope expansion during new underwing stores installation is a challenging problem, mainly related to the aeroelastic flutter phenomenon. Aeroelastic models are usually very hard to model, and therefore flight tests are usually required to validate the aeroelastic model predictions, which given the catastrophic consequences of reaching the flutter point pose an important problem. This constraint favors using short time excitations like Sine Dwell to perform the flight tests, so that the aircraft stays close to the flutter point as little time as possible, but short time data implies a poor spectrum resolution and therefore leads to inaccurate and non repetitive results. The present paper will address the problem related to processing Sine Dwell signals from aeroelastic Flutter Flight Tests, characterized by very short data length (less than 5 s) and low frequency (less than 10 Hz) and used to identify the natural modes associated with the structure. In particular, a new robust technique, the PRESTO algorithm, will be presented and compared to a Matching Pursuit estimation based on Laplace Wavelet. Both techniques have demonstrated to be very accurate and robust procedures on very short time (Sine Dwell) signals, with the particularity that the Laplace Wavelet estimation has already been validated over F-18 real Flutter Flight Test data as described in different papers. However, the PRESTO algorithm improves the performance and accuracy of the Laplace Wavelet processing while keeping its robustness, both on real and simulated data.

Aeroelastic Flutter and Sliding Mode Control of Wind Turbine Blade

Flutter is an important form of wind turbine blade failure. Based on damping analysis, synthetically considering aeroelastic vibration instability of the blade and using the parameter fitting method, the aeroelastic flutter model of the pretwisted blade is built, with the simulation and emulation of flap and lead-lag directions flutter of the 2D dangerous cross section realized. Through the construction of two controllers, modular combinatorial sliding mode controller and sliding mode controller based on LMI for parameterized design suppress blade aeroelastic flutter. The results show that a better control effect can be achieved on the premise of the design of the precise parameters of the controller: the proposed sliding mode control algorithm based on LMI can effectively act on the aeroelastic system of the blade, significantly reduce the vibration frequency, and make the aeroelastic system converge to an acceptable static difference in a short time, which proves the effectiveness of sliding mode control in suppressing high-frequency vibration under high wind speed.