This article presents a novel identification approach which can deal with nonlinear and time-varying characteristics of complex dynamic systems, especially an aerial vehicle in the entire flight envelope. A set of local sub-models are first developed at different operating points of the system, and subsequently a self-organizing multi-model ensemble is introduced to aggregate the outputs of the local models as a single model. The number of employed local models in the proposed multi-model ensemble is optimized using a novel self-organizing approach. Also, wavelet neural networks, which combine both the universal approximation property of neural networks and the wavelet decomposition capability, are used as the local models of the proposed method. In addition, a generalized online sequential extreme learning machine is adopted in the introduced approach to determine the optimal validity function of the local models at each time step. Finally, the introduced self-organizing multi-model ensemble is applied to the NASA Generic Transport Model as a complex nonlinear system to demonstrate the effectiveness of the proposed identification approach. Furthermore, the results obtained from the conventional artificial neural networks are carefully compared with those from the wavelet neural networks, which are employed as the local models of the introduced multi-model ensemble. The simulation results suggest that the introduced wavelet neural network–based self-organizing multi-model ensemble can be used satisfactorily as the prediction model of model-based control systems for long prediction horizons.
ChenLTulsyanAHuangB, et al. Multiple model approach to nonlinear system identification with an uncertain scheduling variable using EM algorithm. J Proces Control2013; 23(10): 1480–1496.
2.
BillingsSA (ed.) Nonlinear system identification: NARMAX methods in the time, frequency, and spatio-tempora domains. 1st ed. Hoboken, NJ: John Wiley & Sons, 2013.
3.
TischlerMBRempleRK.Aircraft & rotorcraft system identification: engineering methods with flight-test examples. Blacksburg, VI: AIAA, 2006.
4.
ShraimHAwadaAYounessR.A survey on quadrotors: configurations, modeling and identification, control, collision avoidance, fault diagnosis and tolerant control. IEEE Aerosp Electr Syst Mag2018; 33(7): 14–33.
BanazadehABehrooM.Development, instrumentation, and dynamics identification of a coanda air vehicle. IEEE Aerosp Electr Syst Mag2015; 30(10): 4–12.
7.
EmamiSABanazadehA.Control oriented modeling and identification of nonlinear systems. Appl Mech Mater2016; 841: 330–337.
8.
NellesO.Nonlinear system identification: from classical approaches to neural networks and fuzzy models. New York: Springer, 2001.
9.
BoutalisYKottasTTheodoridisD, et al. System identification and adaptive control: theory and applications of the neurofuzzy and fuzzy cognitive network models. New York: Springer, 2014.
10.
Pérez-CruzJHChairezIde Jesús RubioJ, et al. Identification and control of class of non-linear systems with non-symmetric deadzone using recurrent neural networks. IET Control Theory Appl2014; 8(3): 183–192.
11.
RoudbariASaghafiF.Intelligent modeling and identification of aircraft nonlinear flight dynamics. Chin J Aeronaut2014; 27(4): 759–771.
12.
YuWPachecoM.Impact of random weights on nonlinear system identification using convolutional neural networks. Inform Sci2019; 477: 1–14.
13.
KangYChenSWangX, et al. Deep convolutional identifier for dynamic modeling and adaptive control of unmanned helicopter. IEEE T Neural Netw Learn Syst2019; 30(2): 524–538.
14.
JategaonkarRV.Flight vehicle system identification: a time-domain methodology, 2nd ed.Reston, VA: American Institute of Aeronautics and Astronautics, 2015.
15.
EsenHOzgenFEsenM, et al. Artificial neural network and wavelet neural network approaches for modelling of a solar air heater. Expert Syst Appl2009; 36(8): 11240–11248.
16.
AlexandridisAKZapranisAD.Wavelet neural networks: a practical guide. Neural Netw2013; 42: 1–27.
17.
HashemiSMAHaji KazemiHKaramodinA.Localized genetically optimized wavelet neural network for semi-active control of buildings subjected to earthquake. Struct Control Health Monit2016; 23(8): 1074–1087.
18.
Pérez-SánchezBFontenla-RomeroOGuijarro-BerdiñasB.A review of adaptive online learning for artificial neural networks. Artif Intel Rev2018; 49(2): 281–299.
19.
RoudbariASaghafiF.Generalization of ANN-based aircraft dynamics identification techniques into the entire flight envelope. IEEE Trans Aerosp Electr Syst2016; 52(4): 1866–1880.
20.
LuYHuangB.Robust multiple-model LPV approach to nonlinear process identification using mixture t distributions. J Proces Control2014; 24(9): 1472–1488.
21.
AdeniranAAEl FerikS.Modeling and identification of nonlinear systems: a review of the multimodel approach—Part 1. IEEE T Syst Man Cyb: Syst2017; 47(7): 1149–1159.
22.
EmamiSARoudbariA.Multimodel ELM-based identification of an aircraft dynamics in the entire flight envelope. IEEE T Aerosp Electr Syst2019; 55(5): 2181–2194.
23.
EmamiSARezaeizadehA.Adaptive model predictive control-based attitude and trajectory tracking of a VTOL aircraft. IET Control Theory Appl2018; 12(15): 2031–2042.
24.
EmamiSABanazadehA.Intelligent trajectory tracking of an aircraft in the presence of internal and external disturbances. Int J Robust Nonl Control2019; 29(16): 5820–5844.
25.
QinPNishiiRYangZJ.Selection of NARX models estimated using weighted least squares method via GIC-based method and l 1-norm regularization methods. Nonlin Dynam2012; 70(3): 1831–1846.
26.
OktayTArikSTurkmenI, et al. Neural network based redesign of morphing UAV for simultaneous improvement of roll stability and maximum lift/drag ratio. Aircraft Eng Aerosp Tech2018; 90(8): 1203–1212.
27.
BrewickPTMasriSF.An evaluation of data-driven identification strategies for complex nonlinear dynamic systems. Nonlin Dynam2016; 85(2): 1297–1318.
28.
da Costa LopesFWatanabeEHRolimLGB. A control-oriented model of a PEM fuel cell stack based on NARX and NOE neural networks. IEEE T Ind Electr2015; 62(8): 5155–5163.
29.
AkpanVAHassapisGD.Nonlinear model identification and adaptive model predictive control using neural networks. ISA Trans2011; 50(2): 177–194.
30.
PatanKKorbiczJ.Nonlinear model predictive control of a boiler unit: a fault tolerant control study. Int J Appl Math Comput Sci2012; 22(1): 225–237.
31.
PatanK.Neural network-based model predictive control: fault tolerance and stability. IEEE Trans Control Syst Tech2015; 23(3): 1147–1155.
32.
DreyfusG.Neural networks: methodology and applications, vol. 24. Berlin: Springer, 2005.
33.
ZhangQ.Nonlinear system identification with output error model through stabilized simulation. IFAC Proc Vol2004; 37(13): 501–506.
34.
LinCJChinCC.Prediction and identification using wavelet-based recurrent fuzzy neural networks. IEEE T Syst Man Cyb, Part B2004; 34(5): 2144–2154.
35.
OussarYRivalsIPersonnazL, et al. Training wavelet networks for nonlinear dynamic input–output modeling. Neurocomputing1998; 20(1–3): 173–188.
36.
ZhangQ.Using wavelet network in nonparametric estimation. IEEE Transactions on Neural Networks1997; 8(2): 227–236 DOI:10.1109/72.557660.
37.
RoudbariASaghafiF.Modeling and identification of highly maneuverable fighter aircraft dynamics using block-oriented nonlinear models. Proc IMechE, Part G: J Aerospace Engineering2016; 231(7): 1293–1311.
38.
El FerikSAdeniranAA. Modeling and identification of nonlinear systems: a review of the multimodel approach—part 2. IEEE T Syst Man Cyb: Syst2017; 47(7): 1160–1168.
39.
HuangJJiGZhuY, et al. Identification of multi-model LPV models with two scheduling variables. J Process Control2012; 22(7): 1198–1208.
40.
EmamiSABanazadehA.Online identification of aircraft dynamics in the presence of actuator faults. J Intel Robot Syst2019; 96(3-4): 541–553.
41.
ÅströmKJWittenmarkB.Adaptive control(Dover books on engineering). Mineola, NY: Dover Publications, 2008.
42.
SoaresSGAraújoR.An adaptive ensemble of on-line extreme learning machines with variable forgetting factor for dynamic system prediction. Neurocomputing2016; 171: 693–707.
43.
HanHGZhangLHouY, et al. Nonlinear model predictive control based on a self-organizing recurrent neural network. IEEE T Neural Networ Learn Syst2016; 27(2): 402–415.
44.
BoškoviJDJacksonJA.Adaptive flight control under actuation and sensor failures and slow-engine dynamics. J Guid Control Dynam2017; 40(4): 905–919.
45.
AnsariABernsteinDS.Retrospective cost adaptive control of generic transport model under uncertainty and failure. J Aerospa Inform Syst2017; 14(3): 123–174.
46.
Al-JibooryAKZhuGSweiSSM, et al. LPV modeling of a flexible wing aircraft using modal alignment and adaptive gridding methods. Aerosp Sci Technol2017; 66: 92–102.
47.
Dan ForeseeFHaganMT. Gauss-Newton approximation to Bayesian learning. In: Proceedings of international conference on neural networks (ICNN’97), Houston, TX, 9–12 June 1997, pp.1930–1935. New York: IEEE.
48.
ZainuddinZPaulineO.Modified wavelet neural network in function approximation and its application in prediction of time-series pollution data. Appl Soft Comput2011; 11(8): 4866–4874.
49.
de Jesus RubioJ. SOFMLS: online self-organizing fuzzy modified least-squares network. IEEE T Fuzzy Syst2009; 17(6): 1296–1309.
