• Corpus ID: 238259272

Efficient Modeling of Morphing Wing Flight Using Neural Networks and Cubature Rules

  title={Efficient Modeling of Morphing Wing Flight Using Neural Networks and Cubature Rules},
  author={Paul Ghanem and Yunus Bicer and Deniz Erdoğmuş and Alireza Ramezani},
Fluidic locomotion of flapping Micro Aerial Vehicles (MAVs) can be very complex, particularly when the rules from insect flight dynamics (fast flapping dynamics and light wings) are not applicable. In these situations, widely used averaging techniques can fail quickly. The primary motivation is to find efficient models for complex forms of aerial locomotion where wings constitute a large part of body mass (i.e., dominant inertial effects) and deform in multiple directions (i.e., morphing wing… 

Figures from this paper



Computational Structure Design of a Bio-Inspired Armwing Mechanism

This work demonstrates that several biologically meaningful degrees of freedom can be interconnected to one another by mechanical intelligence and, as a result, the responsibility of feedback-driven components is subsumed under computational morphology.

Trajectory Optimization With Implicit Hard Contacts

A contact invariant trajectory optimization formulation to synthesize motions for legged robotic systems using concepts from bilevel optimization to find gradients of the system dynamics including the constraint forces and solve the optimal control problem with the unconstrained iLQR algorithm.

State-Space Adaptation of Unsteady Lifting Line Theory: Twisting/Flapping Wings of Finite Span

In this paper, a low-order state-space adaptation of the unsteady lifting line model has been analytically derived for a wing of finite aspect ratio, suitable for use in real-time control of

Bayesian Optimization Using Domain Knowledge on the ATRIAS Biped

The results show that this feature transform captures important aspects of walking and accelerates learning on hardware and simulation, as compared to traditional BO.

Cubature Kalman Filters

A third-degree spherical-radial cubature rule is derived that provides a set of cubature points scaling linearly with the state-vector dimension that may provide a systematic solution for high-dimensional nonlinear filtering problems.

Predict time series using extended, unscented, and cubature Kalman filters based on feed-forward neural network algorithm

The extended, unscented, and cubature Kalman filters is used for training the feed-forward neural network (FNN) and to evaluate the proposed method, these techniques have been used to forecast Mackey-Glass time series.

Robust nonlinear system identification using neural-network models

A neural-network version of an H(infinity)-based identification algorithm from Didinsky et al is presented and it is shown how this algorithm leads to satisfaction of a relevant persistency of excitation condition, and thereby to robust identification of the system nonlinearity.

Convergence Study in Extended Kalman Filter-Based Training of Recurrent Neural Networks

The results show that the R adaption law can effectively avoid the divergence problem and ensure the training convergence, whereas the Q ada adaptation law helps improve theTraining convergence speed.

A Bayesian approach to problems in stochastic estimation and control

In this paper, a general class of stochastic estimation and control problems is formulated from the Bayesian Decision-Theoretic viewpoint. A discussion as to how these problems can be solved step by