Comparison of Real-time Online and Offline Neural Network Models for a UAV

Puttige, Vishwas; Anavatti, Sreenatha G.

doi:10.1109/ijcnn.2007.4370992

Cited by 19 publications

(5 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The system matrices A t , B t , Cand G t in model (13) are calculated by (14)(15)(16)(17)(18) for which knowledge of the working-point state predictionˆ( )( , , , ) w t jt j N + = − | 1 2 1 is required. It is possible to obtain knowledge about some processes being controlled.…”

Section: Multistep-ahead Predictive Model For Mpcmentioning

confidence: 99%

“…To overcome the disadvantages and restrictions of the physical modeling, some intelligent and advanced methods have been explored for modeling and controlling quad‐rotor vehicles. Puttige and Anavatti presented a comparison of real‐time online and offline NN models for a quad‐rotor UAV, the approximation property of NN is applied to learn the dynamics of the quad‐rotor UAV. In , the nonlinear dynamics of the quad‐rotor are described by a multiple input multiple output (MIMO) radial basis function (RBF)–autoregressive with exogenous variables (ARX) model, and a state feedback control law with a linear quadratic regulator (LQR) approach is proposed.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Quad‐Rotor Modeling and Attitude Control Using State‐Dependent ARX Type Model

Zeng¹,

Peng²,

Wu³

et al. 2014

Asian Journal of Control

View full text Add to dashboard Cite

A state-dependent autoregressive with exogenous variables (SD-ARX) model whose functional coefficients are approximated by sets of radial basis function (RBF) networks is proposed to describe the dynamic behavior of a quad-rotor in this paper. This model is identified offline and used as an internal predictor of a receding horizon predictive controller to address the quad-rotor's attitude control issue. In addition, the physical constraints of the system have been also taken into account during the controller design process. The results of real-time control on a quad-rotor aircraft illustrate satisfactory modeling accuracy in a large operating range and good performance of control approach proposed in this paper.

show abstract

Section: Multistep-ahead Predictive Model For Mpcmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Quad‐Rotor Modeling and Attitude Control Using State‐Dependent ARX Type Model

Zeng¹,

Peng²,

Wu³

et al. 2014

Asian Journal of Control

View full text Add to dashboard Cite

show abstract

“…Hence, it may not perform as expected in different flight conditions. Due to the constraints on the training (particularly in terms of the available training time) in the online model it may not provide the desired accuracy under all flight conditions [21]. As a solution to this issue the offline and the online models are combined in the form of a multi-network architecture and individual outputs are compared with respect to a particular criterion and dynamically selected.…”

Section: Multi-network Architecturementioning

confidence: 99%

Real-time System Identification of Unmanned Aerial Vehicles: A Multi-Network Approach

Puttige¹,

Anavatti²

2008

JCP

View full text Add to dashboard Cite

In this paper, real-time system identification of an unmanned aerial vehicle (UAV) based on multiple neural networks is presented. The UAV is a multi-input multi-output (MIMO) nonlinear system. Models for such MIMO system are expected to be adaptive to dynamic behaviour and robust to environmental variations. This task of accurate modelling has been achieved with a multi-network architecture. The multi-network with dynamic selection technique allows a combination of online and offline neural network models to be used in the architecture where the most suitable outputs are selected based on a given criterion. The neural network models are based on the autoregressive technique. The online network uses a novel training scheme with memory retention. Flight test validation results for online and offline models are presented. The multi-network dynamic selection technique has been validated on real-time hardware in the loop (HIL) simulation and the results show the superiority in performance compared to the individual models.

show abstract

“…Hence, it may not perform as expected in different flight conditions. Due to the constraints on the training (particularly in terms of the available training time) in the online model it may not provide the desired accuracy under all flight conditions [16]. As a solution to this issue the offline and the online models are compared with respect to a particular criterion and dynamically selected.…”

Section: Multi-network Architecturementioning

confidence: 99%

Real-time multi-network based identification with dynamic selection implemented for a low cost UAV

Puttige

Anavatti

2007

2007 IEEE International Conference on Systems, Man and Cybernetics

Self Cite

View full text Add to dashboard Cite

This paper describes a system identification technique based on dynamic selection of multiple neural networks for the Unmanned Aerial Vehicle (UAV). The UAV is a Multiinput Multi-output (MIMO) nonlinear system. The neural network models are based on the autoregressive technique. The multi-network dynamic selection method allows a combination of online and offline neural network models to be used in the architecture where the most suitable output is selected based on the given criteria. The online network uses a novel training scheme with memory retention. Flight test validation results for online and offline models are presented. Real-time Hardware in the loop (HIL) simulation results show that the multi-net dynamic selection technique performs better than the individual models.

show abstract

Comparison of Real-time Online and Offline Neural Network Models for a UAV

Cited by 19 publications

References 6 publications

Quad‐Rotor Modeling and Attitude Control Using State‐Dependent ARX Type Model

Quad‐Rotor Modeling and Attitude Control Using State‐Dependent ARX Type Model

Real-time System Identification of Unmanned Aerial Vehicles: A Multi-Network Approach

Real-time multi-network based identification with dynamic selection implemented for a low cost UAV

Contact Info

Product

Resources

About