Parameter estimation of aircraft using extreme learning machine and Gauss-Newton algorithm

Verma, Hari Om; Peyada, N. K.

doi:10.1017/aer.2019.123

Cited by 7 publications

(9 citation statements)

References 35 publications

(49 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Once the network trained, the aerodynamic coefficients C Y , C l ,C n in the network input are replaced by the estimated values ĈY ; Ĉl ; Ĉn of the aerodynamic coefficient model (31), and then the output of the network is obtained. Furthermore, LM algorithm based on PSO is used to estimate the parameter vector H. For the three kinds of networks, the initial guess of H can be selected randomly, which is different from the constraint of parameter selection in the vicinity of EEM estimation results proposed by Verma and Peyada (2020) . The estimates of parameter vector H based on EEM, FEM and NN-LM are shown in Table 4.…”

Section: Extracting Longitudinal Parameters From Simulation Datamentioning

confidence: 99%

“…Extreme learning machine (ELM) was used to substitute for the traditional FFNN in the nonlinear modeling of the aircraft’s dynamic model owing to its robustness and noniterative training way, and GN algorithm combined with ELM was used to estimate aerodynamic parameters of stable and unstable aircraft (Verma and Peyada, 2017, 2018, 2020). Although these researches focused on the improvement of neural network, all of them used GN algorithm in the process of parameter estimation.…”

Section: Introductionmentioning

confidence: 99%

“…In these literatures, the prior information of the parameters to be identified is often required when the GN method is used to estimate the parameters, otherwise the GN algorithm may converge to the local minimum or diverge. For example, Verma and Peyada (2020) used the estimated results of EEM to initialize ELM-GN method. Verma and Peyada (2017) mentioned that when the number of hidden layer nodes of neural network is large, the GN method may fail to estimate parameters.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Aerodynamic coefficients modeling using Levenberg–Marquardt algorithm and network

Wang

et al. 2021

AEAT

View full text Add to dashboard Cite

Purpose The purpose of this paper is to propose a new aerodynamic parameter estimation methodology based on neural network and output error method, while the output error method is improved based on particle swarm algorithm. Design/methodology/approach Firstly, the algorithm approximates the dynamic characteristics of aircraft based on feedforward neural network. Neural network is trained by extreme learning machine, and the trained network can predict the aircraft response at (k + 1)th instant given the measured flight data at kth instant. Secondly, particle swarm optimization is used to enhance the convergence of Levenberg–Marquardt (LM) algorithm, and the improved LM method is used to substitute for the Gauss Newton algorithm in output error method. Finally, the trained neural network is combined with the improved output error method to estimate aerodynamic derivatives. Findings Neither depending on the initial guess of the parameters to be estimated nor requiring numerical integration of the aircraft motion equation, the proposed algorithm can be used for unstable aircraft and is successfully applied to extract aerodynamic derivatives from both simulated and real flight data. Research limitations/implications The proposed method requires iterative calculation and can only identify parameters offline. Practical implications The proposed method is successfully applied to estimate aircraft aerodynamic parameters and can also be used as a new algorithm for other optimization problems. Originality/value In this study, the output error method is improved to reduce the dependence on the initial value of parameters and expand its application scope. It is applied in aircraft aerodynamic parameter identification together with neural network.

show abstract

Section: Extracting Longitudinal Parameters From Simulation Datamentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Aerodynamic coefficients modeling using Levenberg–Marquardt algorithm and network

Wang

et al. 2021

AEAT

View full text Add to dashboard Cite

show abstract

“…Two major techniques that have been widely used and well developed are the time-domain identification and the frequency-domain identification (10) . The Maximum Likelihood (ML) method is by far the most commonly used time-domain technique for estimating parameters from dynamic flight data (14,(16)(17)(18) . Previously, Saderla et al (12,17) efficiently applied the ML together with Gauss Newton (GN) method to estimate longitudinal and lateral-directional parameter for Unmanned Aerial Vehicle (UAV).…”

Section: Introductionmentioning

confidence: 99%

“…Previously, Saderla et al (12,17) efficiently applied the ML together with Gauss Newton (GN) method to estimate longitudinal and lateral-directional parameter for Unmanned Aerial Vehicle (UAV). Verma and Peyada (18) utilized the extreme learning machine based to extract the stability and control derivatives of the all composites HANSA-3 aircraft using the real flight test data.…”

Section: Introductionmentioning

confidence: 99%

Enhancement of light aircraft 6 DOF simulation using flight test data in longitudinal motion

et al. 2021

View full text Add to dashboard Cite

This paper proposes a procedure to improve the accuracy of the light aircraft 6 DOF simulation model by implementing model tuning and aerodynamic database correction using flight test data. In this study, the full-scale flight testing of a 2-seater aircraft has been performed in specific longitudinal manoeuver for model enhancement and simulation validation purposes. The baseline simulation model database is constructed using multi-fidelity analysis methods such as wind tunnel (W/T) test, computational fluid dynamic (CFD) and empirical calculation. The enhancement process starts with identifying longitudinal equations of motion for sensitivity analysis, where the effect of crucial parameters is analysed and then adjusted using the model tuning technique. Next, the classical Maximum Likelihood (ML) estimation method is applied to calculate aerodynamic derivatives from flight test data, these parameters are utilised to correct the initial aerodynamic table. A simulation validation process is introduced to evaluate the accuracy of the enhanced 6 DOF simulation model. The presented results demonstrate that the applied enhancement procedure has improved the simulation accuracy in longitudinal motion. The discrepancy between the simulation and flight test response showed significant improvement, which satisfies the regulation tolerance.

show abstract

Estimation of aerodynamic parameters near stall using maximum likelihood and extreme learning machine-based methods

Verma

Peyada

2020

Aeronaut. j.

View full text Add to dashboard Cite

The stability and control derivatives are essential parameters in the flight operation of aircraft, and their determination is a routine task using classical parameter estimation methods based on maximum likelihood and least-squares principles. At high angle-of-attack, the unsteady aerodynamics may pose difficulty in aerodynamic structure determination, hence data-driven methods based on artificial neural networks could be an alternative choice for building models to characterise the behaviour of the system based on the measured motion and control variables. This research paper investigates the feasibility of using a recurrent neural model based on an extreme learning machine network in the modelling of the aircraft dynamics in a restricted sense for identification of the aerodynamic parameters. The recurrent extreme learning machine network is combined with the Gauss–Newton method to optimise the unknowns of the postulated aerodynamic model. The efficacy of the proposed estimation algorithm is studied using real flight data from a quasi-steady stall manoeuvre. Furthermore, the estimates are validated against the parameters estimated using the maximum likelihood method. The standard deviations of the estimates demonstrate the effectiveness of the proposed algorithm. Finally, the quantities regenerated using the estimates present good agreement with their corresponding measured values, confirming that a qualitative estimation can be obtained using the proposed estimation algorithm.

show abstract

Parameter estimation of aircraft using extreme learning machine and Gauss-Newton algorithm

Cited by 7 publications

References 35 publications

Aerodynamic coefficients modeling using Levenberg–Marquardt algorithm and network

Aerodynamic coefficients modeling using Levenberg–Marquardt algorithm and network

Enhancement of light aircraft 6 DOF simulation using flight test data in longitudinal motion

Estimation of aerodynamic parameters near stall using maximum likelihood and extreme learning machine-based methods

Contact Info

Product

Resources

About