Real-Time Neural Network Based Online Identification Technique for a UAV Platform

Puttige, Vishwas; Anavatti, Sreenatha G.

doi:10.1109/cimca.2006.170

Cited by 17 publications

(9 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the ARX-NN model, the network retains information about the past dynamics to predict the next output. The predicted output of a nonlinear ARX model is given by [18], [19], [20],…”

Section: Neural Network Identificationmentioning

confidence: 99%

A UKF-NN Framework for System Identification of Small Unmanned Aerial Vehicles

Kallapur

Samal

Puttige

et al. 2008

2008 11th International IEEE Conference on Intelligent Transportation Systems

Self Cite

View full text Add to dashboard Cite

This paper presents a novel system identification framework for small unmanned aerial vehicles (UAVs) by combining an unscented Kalman filter (UKF) estimator with a neural network (NN) identifier. The method is effective for systems with low-cost, erroneous sensors where the sensor outputs cannot be used directly for system identification and control. The UKF state estimator computes error-compensated attitude and velocities by integrating sensor data from an inertial measurement unit (IMU) and a global positioning system (GPS). The NN identifier approximates the nonlinear dynamics of the UAV from the UKF estimated states, hence identifying the system. As an illustration, the UKF-NN system identification framework is applied to fixed-wing as well as rotary-wing 6-DOF multi-input-multi-output (MIMO) nonlinear UAV models.

show abstract

“…In the ARX-NN model, the network retains information about the past dynamics to predict the next output. The predicted output of a nonlinear ARX model is given by [18], [19], [20],…”

Section: Neural Network Identificationmentioning

confidence: 99%

A UKF-NN Framework for System Identification of Small Unmanned Aerial Vehicles

Kallapur

Samal

Puttige

et al. 2008

2008 11th International IEEE Conference on Intelligent Transportation Systems

Self Cite

View full text Add to dashboard Cite

show abstract

“…This provides equivalent retention capabilities of the dynamics of the UAV by the network. The predicted output of a nonlinear model can be obtained as [9], [15] y(t|θ) = g(a 1 y(t − 1) + a 2 y(t − 2) + ..…”

Section: The Neural Network Modelsmentioning

confidence: 99%

“…The models are based on Autoregressive technique with Exogenous inputs (ARX) introduced by Ljung in [9]. A novel training method is adapted for the online model where the network is trained with small batches of data and the weights from the previous batch are retained in memory [15]. The retraining of the network online is carried out only when the error in prediction is beyond a certain threshold.…”

Section: Introductionmentioning

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

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

“…The auto-regressor employs the past inputs and outputs as inputs to the neural network. The major disadvantage of this technique is however the computational time which increases with an increase in the regressor size [11].…”

Section: Introductionmentioning

confidence: 99%

A computationally efficient approach for NN based system identification of a rotary wing UAV

Samal

Anavatti

Ray

et al. 2010

Int. J. Control Autom. Syst.

Self Cite

View full text Add to dashboard Cite

Neural Network (NN) models based on autoregressive structures have long been used for nonlinear system identification problems. Their application for on-line implementations, however require them to be trained within a prescribed time span, which is often related to the sampling time of the system. In this paper, we introduce a NN model that is embedded with a dimensionality reduction mechanism in order to reduce the size of the network. The dimensionality reduction is based on Principal Component Analysis (PCA) and the resulting smaller NN trains faster. The longitudinal and lateral dynamics of a rotary wing Unmanned Aerial Vehicle (UAV) is modelled using flight test data. The results of system identification, error statistics and training times are provided to highlight the benefits of the proposed approach for NN based system identification models.

show abstract

Real-Time Neural Network Based Online Identification Technique for a UAV Platform

Cited by 17 publications

References 13 publications

A UKF-NN Framework for System Identification of Small Unmanned Aerial Vehicles

A UKF-NN Framework for System Identification of Small Unmanned Aerial Vehicles

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

A computationally efficient approach for NN based system identification of a rotary wing UAV

Contact Info

Product

Resources

About