Neural network based online estimation of maneuvering steady states and control limits

Download
2010
Gürsoy, Gönenç
This thesis concerns the design and development of neural network based predictive algorithms to predict approaching aircraft limits. Therefore, approximate dynamics of flight envelope parameters such as angle of attack and load factor are constructed using neural network augmented dynamic models. Then, constructed models are used to predict steady state responses. By inverting the models and solving for critical controls at the known envelope limits, critical control inputs are calculated as well. The performance of the predictor algorithm is then evaluated with a different neural network online adaptation law which uses a stack of recorded data. It is shown that using a stack of recorded data online, constructed models become much more representative of limit parameter dynamics compared to adaptation using instantaneous measured data only. The benefits of recording data online and using it for weight adaptation are presented in the scope of dynamic trim and control limit predictions.

Suggestions

Attitude estimation and magnetometer calibration using reconfigurable TRIAD plus filtering approach
Söken, Halil Ersin (Elsevier BV, 2020-04-01)
This paper proposes using TRIAD and Unscented Kalman Filter (UKF) algorithms in a sequential architecture as a part of a small satellite attitude estimation algorithm. This TRIAD+UKF approach can both provide accurate attitude estimates for the satellite and calibrate the magnetometers in real-time. A complete calibration model for the magnetometers, considering bias, scale factor, soft iron and nonorthogonality errors, is assumed. In the algorithm's first stage, the TRIAD uses the available vector measurem...
Representing temporal knowledge in connectionist expert systems
Alpaslan, Ferda Nur (1996-09-27)
This paper introduces a new temporal neural networks model which can be used in connectionist expert systems. Also, a Variation of backpropagation algorithm, called the temporal feedforward backpropagation algorithm is introduced as a method for training the neural network. The algorithm was tested using training examples extracted from a medical expert system. A series of experiments were carried out using the temporal model and the temporal backpropagation algorithm. The experiments indicated that the alg...
Neural network based beamforming for linear and cylindrical array applications
Güreken, Murat; Dural Ünver, Mevlüde Gülbin; Department of Electrical and Electronics Engineering (2009)
In this thesis, a Neural Network (NN) based beamforming algorithm is proposed for real time target tracking problem. The algorithm is performed for two applications, linear and cylindrical arrays. The linear array application is implemented with equispaced omnidirectional sources. The influence of the number of antenna elements and the angular seperation between the incoming signals on the performance of the beamformer in the linear array beamformer is studied, and it is observed that the algorithm improves...
Prediction of Nonlinear Drift Demands for Buildings with Recurrent Neural Networks
Kocamaz, Korhan; Binici, Barış; Tuncay, Kağan (2021-09-08)
Application of deep learning algorithms to the problems of structural engineering is an emerging research field. Inthis study, a deep learning algorithm, namely recurrent neural network (RNN), is applied to tackle a problemrelated to the assessment of reinforced concrete buildings. Inter-storey drift ratio profile of a structure is a quiteimportant parameter while conducting assessment procedures. In general, procedures require a series of timeconsuming nonlinear dynamic analysis. In this study, an extensiv...
Neural identification of dynamic systems on FPGA with improved PSO learning
Cavuslu, Mehmet Ali; KARAKUZU, CİHAN; KARAKAYA, FUAT (2012-09-01)
This work introduces hardware implementation of artificial neural networks (ANNs) with learning ability on field programmable gate array (FPGA) for dynamic system identification. The learning phase is accomplished by using the improved particle swarm optimization (PSO). The improved PSO is obtained by modifying the velocity update function. Adding an extra term to the velocity update function reduced the possibility of stucking in a local minimum. The results indicates that ANN, trained using improved PSO a...
Citation Formats
G. Gürsoy, “Neural network based online estimation of maneuvering steady states and control limits,” M.S. - Master of Science, Middle East Technical University, 2010.