Nonlinear System Identification Using MLP Dr.-Ing. Sudchai Boonto

Size: px
Start display at page:

Download "Nonlinear System Identification Using MLP Dr.-Ing. Sudchai Boonto"

Transcription

1 Dr-Ing Sudchai Boonto Department of Control System and Instrumentation Engineering King Mongkut s Unniversity of Technology Thonburi Thailand

2 Nonlinear System Identification Given a data set Z N = {y(k), φ(k); k = 1,, N}, generated from a nonlinear function y = g(φ) where the regressor vector φ(k) contains past input and output data A nonlinear regressor model is of the form y(k) = g(φ(k), θ) + e(k) y(k) represents measured system outputs the regressor vector φ(k) R r contains samples of measured input and output signals taken prior to sampling instant k θ R n p is a vector whose elements are the weights and biases of the MLP network to be trained 2/31

3 Nonlinear System Identification cont The nonlinear function g( ) describes the mapping of neural network inputs into outputs The perturbation v(k) is already transforms into the form where e(k) is a white noise process v(k) = e(k) + past values of v the disturbance acting on the measured plant output may be non-white, a noise model taking account of this is absorbed into the function g( ) and is represented together with the plant dynamics by the parameter vector θ the prediction error is white when θ takes its optimal value The predictor model ŷ(k) = g(φ(k), θ) 3/31

4 NNARX Model Structure The simplest neural network based model structure is the NNARX structure, where ] T φ(k) = [y k 1 y k n u k d u k d m φ 1 (k) φ r (k) NN θ y(k) ŷ(k) ε(k) 4/31

5 NNARX Model Structure cont a static feedforward network is used to represent a dynamic nonlinear system as predictor model in training mode, such a predictor model is used to find the weights and biases that give the best fit between predicted and measured output applied the data to the network, and the network output ŷ(k) obtained with weights and biases according to θ(l) are compared with the measured outputs y(k) the resulting values of ŷ(k) and ε(k) are used in Levenberg-Marquardt backpropagation to compute a search direction and update θ(l) 5/31

6 NNARMAX Model Structure The regressor vector for an NNARMAX model is ] T φ(k, θ) = [y k 1 y k n u k d u k d m ε k 1 ε k n The predictor model includes feedback from network output to inputs the regressor vector depends on the network parameters θ, thus instead of ŷ(k) = g(φ(θ), θ) we have ŷ(k) = g(φ(k, θ), θ) 6/31

7 NNARMAX Model Structure cont { φ(k) ε k n NN y(k) z 1 ŷ(k) z 1 ε k 1 θ ε(k) 7/31

8 NNARMAX Model Structure cont Since the predicted output is not a function of independent variables, the gradient of the cost is determined by the total derivative dŷ(k) dθ = ŷ(k) θ + ŷ(k) dε(k 1) + + ŷ(k) dε(k n) ε(k 1) dθ ε(k n) dθ Denote ψ(k) = ϕ(k) ŷ(k) ψ(k 1) ŷ(k) ψ(k n) ε(k 1) ε(k n) 8/31

9 NNARMAX Model Structure backpropagation Computing ψ k requires also the partial derivatives of predicted outputs with respect to past prediction errors The problem simplifies if the NNARMAX structure is modified such that the disturbance model is linear Such a regressor model is ŷ(k) = g( φ(k), θ) + (C(z 1 ) 1)ε(k) where φ and θ represent the regressor vector and the network parameters, respectively 9/31

10 NNARMAX Model Structure backpropagation cont The polynomial C(z 1 ) = 1 + c 1 z c n z n describes the noise characteristics as in linear ARMAX models NN φ(k) θ ε(k) C(z 1 ) 1 ŷ(k) y(k) 10/31

11 NNARMAX Model Structure backpropagation cont With the modification, the total derivative becomes or ψ(k) = ϕ(k) c 1 ϕ(k 1) c n ϕ(k n) where ψ(k) = 1 C(z 1 ) ϕ(k) θ c θ = 1 c n ŷ(k) θ ε(k 1) and ϕ(k) = ε(k n) 11/31

12 NNARMAX Model Structure backpropagation cont Backpropagation can be used to compute ŷ(k) θ ψ(k) is then computed by forming ϕ(k) and filtering it through 1 C(z 1 ) the coefficients of C(z 1 ) are contained in θ and updated in each iteration the stability of 1 C(z 1 ) must be checked at each step If C is unstable the spectral factorization theorem can be used to replace it with a stable polynomial 12/31

13 Practical Issues Befor a two-layer MLP network is to be trained for a control application, the following choices have to be made Sampling time Dynamic system order Number of hidden neurons Training signal 13/31

14 ph Plant plant Figure: ph neutralisation process 14/31

15 ph Plant plant u 1 : the base (NaOH) flow rate u 2 : the buffer (NaHCO 3 ) flow rate u 3 : the acid (HNO 3 ) flow rate y : the ph of the effluent solution Figure: ph neutralisation process 14/31

16 ph Plant plant Figure: ph neutralisation process u 1 : the base (NaOH) flow rate u 2 : the buffer (NaHCO 3 ) flow rate u 3 : the acid (HNO 3 ) flow rate y : the ph of the effluent solution u 3 and volume (V ) of the tank are assumed to be constant 14/31

17 NNARX y k 1 u k d Neural Network y k n u k d m ŷ k Figure: NNARX model structure 15/31

18 NNARX-Lag space Dynamic system order (lag space) y(k) = g 0 [ϕ(k), θ)] ϕ T (k) = [y(k 1) y(k n), u(k d) u(k d m)] Too small implies that essential dynamics will not be modelled Use He and Asada (1993) method by using function lipschit but sometime it doesn t tell anything Trial and error, start from small value Recommenced value: n = 4, m = 4, d = 1 16/31

19 NNARX-Sampling Time Sampling Time (T s ) A reasonable choice of T s is between 5-10 samples in the rise time 17/31

20 NNARX-Sampling Time Sampling Time (T s ) A reasonable choice of T s is between 5-10 samples in the rise time T r = 50 sec Step test 9 ph values In this case, we choose T s = 5 s Time (sec) 17/31

21 NNARX-Hidden neurons Number of hidden neurons Too small: the network will not have sufficient degrees of freedom to adequately represent the mapping Too large: more local minima, overfit, slow Function Approximation Output Input Start for a large value, ie 10, and reduce it to the smallest value that can map the data 18/31

22 Training Signal The input signal should : Excite all the frequencies of interest (persistently excitation) Excite the process over the whole of the required operation region For nonlinear systems: It is well known now that PRBS signal is not enough Multi-level PRBS Multisine You should know: System bandwidth Input ranges 19/31

23 Training Signal Multi-Level PRBS Multi-Level PRBS: u(k 1) u(k) = e(k) with probability α with probability 1 α where e(k) is a normal distribute random noise Note: The minimum hold time should be long enough so that the systems output has time to approach the new set point 20/31

24 Training Signal Multi-Level PRBS cont 30 Estimation Validation 20 u x y(ph) time (sec) x 10 4 Figure: Estimation (first points) and validation (remaining points) input-output data Multisine 21/31

25 Training Signal Multi-Level PRBS cont 21/31

26 Training Signal Multisine Multisine where u(k) = N = ω max ω min η + 1 N A cos(2πω l k + ϕ l ) l=1 ϕ is random phase (more robust than Schroeder s phased) A is a required overall amplitude ω is a required frequency Normally start from 0 to 3 times bandwidth of open-loop system bandwidth Use command idinput in System Identification toolbox 22/31

27 Training Signal Multisine cont First trick N 1 u(k) = A cos(2πω l k + ϕ l ) + γ l=1 N 2 m=1 A cos(2πω m k + ϕ m ) where the first term start from 0 to 3 times bandwidth and the second term start from 3 times bandwidth to half of sampling frequency γ is very small, ie 03 Second Validation signal should be 10 times slower than training signal 23/31

28 Training Signal Multisine cont u time (sec) y(ph) time (sec) Figure: An example of multisine signal 24/31

29 Training Results 15 Output (solid) and one step ahead prediction (dashed) time (samples) Prediction error (y yhat) time (samples) 25/31

30 Training Results cont 1 Auto correlation function of prediction error lag Cross correlation coef of u1 and prediction error lag 26/31

31 Training Results cont 11 Output (solid) and 10 step ahead prediction (dashed) time (samples) 27/31

32 MATLAB NN Toolbox NNARX 28/31

33 MATLAB NN Toolbox cont 29/31

34 Reference 1 Lecture note on Neural and Genetic Computing for Control Engineering, Werner, H, TUHH 2 System Identification: Theory for the user, Ljung, L,1999, Prentice Hall 3 Neural Networks for Modelling and Control of Dynamic Systems, Norgaard, M Ravn, O Poulsen, N K and Hansen, L K 30/31

Lecture 9: Input Disturbance A Design Example Dr.-Ing. Sudchai Boonto

Lecture 9: Input Disturbance A Design Example Dr.-Ing. Sudchai Boonto Dr-Ing Sudchai Boonto Department of Control System and Instrumentation Engineering King Mongkuts Unniversity of Technology Thonburi Thailand d u g r e u K G y The sensitivity S is the transfer function

More information

EL1820 Modeling of Dynamical Systems

EL1820 Modeling of Dynamical Systems EL1820 Modeling of Dynamical Systems Lecture 10 - System identification as a model building tool Experiment design Examination and prefiltering of data Model structure selection Model validation Lecture

More information

Linear Least Square Problems Dr.-Ing. Sudchai Boonto

Linear Least Square Problems Dr.-Ing. Sudchai Boonto Dr-Ing Sudchai Boonto Department of Control System and Instrumentation Engineering King Mongkuts Unniversity of Technology Thonburi Thailand Linear Least-Squares Problems Given y, measurement signal, find

More information

Control Systems Lab - SC4070 System Identification and Linearization

Control Systems Lab - SC4070 System Identification and Linearization Control Systems Lab - SC4070 System Identification and Linearization Dr. Manuel Mazo Jr. Delft Center for Systems and Control (TU Delft) m.mazo@tudelft.nl Tel.:015-2788131 TU Delft, February 13, 2015 (slides

More information

Recurrent Neural Network Identification: Comparative Study on Nonlinear Process M.Rajalakshmi #1, Dr.S.Jeyadevi #2, C.Karthik * 3

Recurrent Neural Network Identification: Comparative Study on Nonlinear Process M.Rajalakshmi #1, Dr.S.Jeyadevi #2, C.Karthik * 3 ISSN (Online) : 2319-8753 ISSN (Print) : 2347-6710 International Journal of Innovative Research in Science, Engineering and Technology Volume 3, Special Issue 3, March 2014 2014 International Conference

More information

Advanced Process Control Tutorial Problem Set 2 Development of Control Relevant Models through System Identification

Advanced Process Control Tutorial Problem Set 2 Development of Control Relevant Models through System Identification Advanced Process Control Tutorial Problem Set 2 Development of Control Relevant Models through System Identification 1. Consider the time series x(k) = β 1 + β 2 k + w(k) where β 1 and β 2 are known constants

More information

Identification of ARX, OE, FIR models with the least squares method

Identification of ARX, OE, FIR models with the least squares method Identification of ARX, OE, FIR models with the least squares method CHEM-E7145 Advanced Process Control Methods Lecture 2 Contents Identification of ARX model with the least squares minimizing the equation

More information

J. Sjöberg et al. (1995):Non-linear Black-Box Modeling in System Identification: a Unified Overview, Automatica, Vol. 31, 12, str

J. Sjöberg et al. (1995):Non-linear Black-Box Modeling in System Identification: a Unified Overview, Automatica, Vol. 31, 12, str Dynamic Systems Identification Part - Nonlinear systems Reference: J. Sjöberg et al. (995):Non-linear Black-Box Modeling in System Identification: a Unified Overview, Automatica, Vol. 3,, str. 69-74. Systems

More information

EECE Adaptive Control

EECE Adaptive Control EECE 574 - Adaptive Control Basics of System Identification Guy Dumont Department of Electrical and Computer Engineering University of British Columbia January 2010 Guy Dumont (UBC) EECE574 - Basics of

More information

Lecture 7 : Generalized Plant and LFT form Dr.-Ing. Sudchai Boonto Assistant Professor

Lecture 7 : Generalized Plant and LFT form Dr.-Ing. Sudchai Boonto Assistant Professor Dr.-Ing. Sudchai Boonto Assistant Professor Department of Control System and Instrumentation Engineering King Mongkuts Unniversity of Technology Thonburi Thailand Linear Quadratic Gaussian The state space

More information

CSTR CONTROL USING MULTIPLE MODELS

CSTR CONTROL USING MULTIPLE MODELS CSTR CONTROL USING MULTIPLE MODELS J. Novák, V. Bobál Univerzita Tomáše Bati, Fakulta aplikované informatiky Mostní 39, Zlín INTRODUCTION Almost every real process exhibits nonlinear behavior in a full

More information

Lecture 10. Neural networks and optimization. Machine Learning and Data Mining November Nando de Freitas UBC. Nonlinear Supervised Learning

Lecture 10. Neural networks and optimization. Machine Learning and Data Mining November Nando de Freitas UBC. Nonlinear Supervised Learning Lecture 0 Neural networks and optimization Machine Learning and Data Mining November 2009 UBC Gradient Searching for a good solution can be interpreted as looking for a minimum of some error (loss) function

More information

A recursive algorithm based on the extended Kalman filter for the training of feedforward neural models. Isabelle Rivals and Léon Personnaz

A recursive algorithm based on the extended Kalman filter for the training of feedforward neural models. Isabelle Rivals and Léon Personnaz In Neurocomputing 2(-3): 279-294 (998). A recursive algorithm based on the extended Kalman filter for the training of feedforward neural models Isabelle Rivals and Léon Personnaz Laboratoire d'électronique,

More information

( t) Identification and Control of a Nonlinear Bioreactor Plant Using Classical and Dynamical Neural Networks

( t) Identification and Control of a Nonlinear Bioreactor Plant Using Classical and Dynamical Neural Networks Identification and Control of a Nonlinear Bioreactor Plant Using Classical and Dynamical Neural Networks Mehmet Önder Efe Electrical and Electronics Engineering Boðaziçi University, Bebek 80815, Istanbul,

More information

Stochastic Process II Dr.-Ing. Sudchai Boonto

Stochastic Process II Dr.-Ing. Sudchai Boonto Dr-Ing Sudchai Boonto Department of Control System and Instrumentation Engineering King Mongkuts Unniversity of Technology Thonburi Thailand Random process Consider a random experiment specified by the

More information

A STATE-SPACE NEURAL NETWORK FOR MODELING DYNAMICAL NONLINEAR SYSTEMS

A STATE-SPACE NEURAL NETWORK FOR MODELING DYNAMICAL NONLINEAR SYSTEMS A STATE-SPACE NEURAL NETWORK FOR MODELING DYNAMICAL NONLINEAR SYSTEMS Karima Amoura Patrice Wira and Said Djennoune Laboratoire CCSP Université Mouloud Mammeri Tizi Ouzou Algeria Laboratoire MIPS Université

More information

NONLINEAR GAS TURBINE MODELING USING FEEDFORWARD NEURAL NETWORKS. Neophytos Chiras, Ceri Evans and David Rees

NONLINEAR GAS TURBINE MODELING USING FEEDFORWARD NEURAL NETWORKS. Neophytos Chiras, Ceri Evans and David Rees Proceedings of ASME TURBO EXPO 22 June 3-6, 22, Amsterdam, The etherlands GT-22-335 OLIEAR GAS TURBIE MODELIG USIG FEEDFORWARD EURAL ETWORKS eophytos Chiras, Ceri Evans and David Rees University of Glamorgan,

More information

Modeling and Identification of Dynamic Systems (vimmd312, 2018)

Modeling and Identification of Dynamic Systems (vimmd312, 2018) Modeling and Identification of Dynamic Systems (vimmd312, 2018) Textbook background of the curriculum taken. In parenthesis: material not reviewed due to time shortage, but which is suggested to be read

More information

EL1820 Modeling of Dynamical Systems

EL1820 Modeling of Dynamical Systems EL1820 Modeling of Dynamical Systems Lecture 9 - Parameter estimation in linear models Model structures Parameter estimation via prediction error minimization Properties of the estimate: bias and variance

More information

Modeling and Analysis of Dynamic Systems

Modeling and Analysis of Dynamic Systems Modeling and Analysis of Dynamic Systems by Dr. Guillaume Ducard c Fall 2016 Institute for Dynamic Systems and Control ETH Zurich, Switzerland G. Ducard c 1 Outline 1 Lecture 9: Model Parametrization 2

More information

Neural Network Control

Neural Network Control Neural Network Control Daniel Eggert 24th February 2003 Technical University of Denmark Informatics and Mathematical Modelling Building 321, DK-2800 Lyngby, Denmark Phone +45 4525 3351, Fax +45 4588 2673

More information

Lecture 7: Discrete-time Models. Modeling of Physical Systems. Preprocessing Experimental Data.

Lecture 7: Discrete-time Models. Modeling of Physical Systems. Preprocessing Experimental Data. ISS0031 Modeling and Identification Lecture 7: Discrete-time Models. Modeling of Physical Systems. Preprocessing Experimental Data. Aleksei Tepljakov, Ph.D. October 21, 2015 Discrete-time Transfer Functions

More information

CBE507 LECTURE III Controller Design Using State-space Methods. Professor Dae Ryook Yang

CBE507 LECTURE III Controller Design Using State-space Methods. Professor Dae Ryook Yang CBE507 LECTURE III Controller Design Using State-space Methods Professor Dae Ryook Yang Fall 2013 Dept. of Chemical and Biological Engineering Korea University Korea University III -1 Overview States What

More information

Research Note INTELLIGENT FORECASTING OF RAINFALL AND TEMPERATURE OF SHIRAZ CITY USING NEURAL NETWORKS *

Research Note INTELLIGENT FORECASTING OF RAINFALL AND TEMPERATURE OF SHIRAZ CITY USING NEURAL NETWORKS * Iranian Journal of Science & Technology, Transaction B, Vol. 28, No. B1 Printed in Islamic Republic of Iran, 24 Shiraz University Research Note INTELLIGENT FORECASTING OF RAINFALL AND TEMPERATURE OF SHIRAZ

More information

ECE Introduction to Artificial Neural Network and Fuzzy Systems

ECE Introduction to Artificial Neural Network and Fuzzy Systems ECE 39 - Introduction to Artificial Neural Network and Fuzzy Systems Wavelet Neural Network control of two Continuous Stirred Tank Reactors in Series using MATLAB Tariq Ahamed Abstract. With the rapid

More information

Neural Network Control in a Wastewater Treatment Plant

Neural Network Control in a Wastewater Treatment Plant Neural Network Control in a Wastewater Treatment Plant Miguel A. Jaramillo 1 ; Juan C. Peguero 2, Enrique Martínez de Salazar 1, Montserrat García del alle 1 ( 1 )Escuela de Ingenierías Industriales. (

More information

INC 693, 481 Dynamics System and Modelling: Introduction to Modelling Dr.-Ing. Sudchai Boonto Assistant Professor

INC 693, 481 Dynamics System and Modelling: Introduction to Modelling Dr.-Ing. Sudchai Boonto Assistant Professor INC 693, 481 Dynamics System and Modelling: Introduction to Modelling Dr.-Ing. Sudchai Boonto Assistant Professor Department of Control System and Instrumentation Engineering King Mongkut s Unniversity

More information

Chapter 4 Neural Networks in System Identification

Chapter 4 Neural Networks in System Identification Chapter 4 Neural Networks in System Identification Gábor HORVÁTH Department of Measurement and Information Systems Budapest University of Technology and Economics Magyar tudósok körútja 2, 52 Budapest,

More information

Analysis of Fast Input Selection: Application in Time Series Prediction

Analysis of Fast Input Selection: Application in Time Series Prediction Analysis of Fast Input Selection: Application in Time Series Prediction Jarkko Tikka, Amaury Lendasse, and Jaakko Hollmén Helsinki University of Technology, Laboratory of Computer and Information Science,

More information

Identification in closed-loop, MISO identification, practical issues of identification

Identification in closed-loop, MISO identification, practical issues of identification Identification in closed-loop, MISO identification, practical issues of identification CHEM-E7145 Advanced Process Control Methods Lecture 4 Contents Identification in practice Identification in closed-loop

More information

The Short Sampling Period in Adaptive Control

The Short Sampling Period in Adaptive Control Proceedings of the World Congress on Engineering and Computer cience 9 Vol II WCEC 9, October -, 9, an Francisco, UA he hort ampling Period in Adaptive Control P. Pivoňka, V. Veleba, M. Šeda, P. Ošmera,

More information

Further Results on Model Structure Validation for Closed Loop System Identification

Further Results on Model Structure Validation for Closed Loop System Identification Advances in Wireless Communications and etworks 7; 3(5: 57-66 http://www.sciencepublishinggroup.com/j/awcn doi:.648/j.awcn.735. Further esults on Model Structure Validation for Closed Loop System Identification

More information

y(x n, w) t n 2. (1)

y(x n, w) t n 2. (1) Network training: Training a neural network involves determining the weight parameter vector w that minimizes a cost function. Given a training set comprising a set of input vector {x n }, n = 1,...N,

More information

Neural Network Control of Robot Manipulators and Nonlinear Systems

Neural Network Control of Robot Manipulators and Nonlinear Systems Neural Network Control of Robot Manipulators and Nonlinear Systems F.L. LEWIS Automation and Robotics Research Institute The University of Texas at Arlington S. JAG ANNATHAN Systems and Controls Research

More information

Pattern Classification

Pattern Classification Pattern Classification All materials in these slides were taen from Pattern Classification (2nd ed) by R. O. Duda,, P. E. Hart and D. G. Stor, John Wiley & Sons, 2000 with the permission of the authors

More information

PROPORTIONAL-Integral-Derivative (PID) controllers

PROPORTIONAL-Integral-Derivative (PID) controllers Multiple Model and Neural based Adaptive Multi-loop PID Controller for a CSTR Process R.Vinodha S. Abraham Lincoln and J. Prakash Abstract Multi-loop (De-centralized) Proportional-Integral- Derivative

More information

Thermal error compensation for a high precision lathe

Thermal error compensation for a high precision lathe IWMF214, 9 th INTERNATIONAL WORKSHOP ON MICROFACTORIES OCTOBER 5-8, 214, HONOLULU, U.S.A. / 1 Thermal error compensation for a high precision lathe Byung-Sub Kim # and Jong-Kweon Park Department of Ultra

More information

NONLINEAR IDENTIFICATION ON BASED RBF NEURAL NETWORK

NONLINEAR IDENTIFICATION ON BASED RBF NEURAL NETWORK DAAAM INTERNATIONAL SCIENTIFIC BOOK 2011 pp. 547-554 CHAPTER 44 NONLINEAR IDENTIFICATION ON BASED RBF NEURAL NETWORK BURLAK, V. & PIVONKA, P. Abstract: This article is focused on the off-line identification

More information

RELAY CONTROL WITH PARALLEL COMPENSATOR FOR NONMINIMUM PHASE PLANTS. Ryszard Gessing

RELAY CONTROL WITH PARALLEL COMPENSATOR FOR NONMINIMUM PHASE PLANTS. Ryszard Gessing RELAY CONTROL WITH PARALLEL COMPENSATOR FOR NONMINIMUM PHASE PLANTS Ryszard Gessing Politechnika Śl aska Instytut Automatyki, ul. Akademicka 16, 44-101 Gliwice, Poland, fax: +4832 372127, email: gessing@ia.gliwice.edu.pl

More information

Multilayer Neural Networks

Multilayer Neural Networks Multilayer Neural Networks Multilayer Neural Networks Discriminant function flexibility NON-Linear But with sets of linear parameters at each layer Provably general function approximators for sufficient

More information

Closed loop Identification of Four Tank Set up Using Direct Method

Closed loop Identification of Four Tank Set up Using Direct Method Closed loop Identification of Four Tan Set up Using Direct Method Mrs. Mugdha M. Salvi*, Dr.(Mrs) J. M. Nair** *(Department of Instrumentation Engg., Vidyavardhini s College of Engg. Tech., Vasai, Maharashtra,

More information

INC 341 Feedback Control Systems: Lecture 3 Transfer Function of Dynamic Systems II

INC 341 Feedback Control Systems: Lecture 3 Transfer Function of Dynamic Systems II INC 341 Feedback Control Systems: Lecture 3 Transfer Function of Dynamic Systems II Asst. Prof. Dr.-Ing. Sudchai Boonto Department of Control Systems and Instrumentation Engineering King Mongkut s University

More information

NEURAL NETWORK ADAPTIVE SEMI-EMPIRICAL MODELS FOR AIRCRAFT CONTROLLED MOTION

NEURAL NETWORK ADAPTIVE SEMI-EMPIRICAL MODELS FOR AIRCRAFT CONTROLLED MOTION NEURAL NETWORK ADAPTIVE SEMI-EMPIRICAL MODELS FOR AIRCRAFT CONTROLLED MOTION Mikhail V. Egorchev, Dmitry S. Kozlov, Yury V. Tiumentsev Moscow Aviation Institute (MAI), Moscow, Russia Keywords: aircraft,

More information

Adaptive Inverse Control

Adaptive Inverse Control TA1-8:30 Adaptive nverse Control Bernard Widrow Michel Bilello Stanford University Department of Electrical Engineering, Stanford, CA 94305-4055 Abstract A plant can track an input command signal if it

More information

COGS Q250 Fall Homework 7: Learning in Neural Networks Due: 9:00am, Friday 2nd November.

COGS Q250 Fall Homework 7: Learning in Neural Networks Due: 9:00am, Friday 2nd November. COGS Q250 Fall 2012 Homework 7: Learning in Neural Networks Due: 9:00am, Friday 2nd November. For the first two questions of the homework you will need to understand the learning algorithm using the delta

More information

Direct Method for Training Feed-forward Neural Networks using Batch Extended Kalman Filter for Multi- Step-Ahead Predictions

Direct Method for Training Feed-forward Neural Networks using Batch Extended Kalman Filter for Multi- Step-Ahead Predictions Direct Method for Training Feed-forward Neural Networks using Batch Extended Kalman Filter for Multi- Step-Ahead Predictions Artem Chernodub, Institute of Mathematical Machines and Systems NASU, Neurotechnologies

More information

NEURAL NETWORKS BASED SYSTEM IDENTIFICATION TECHNIQUES FOR MODEL BASED FAULT DETECTION OF NONLINEAR SYSTEMS. Afef Fekih

NEURAL NETWORKS BASED SYSTEM IDENTIFICATION TECHNIQUES FOR MODEL BASED FAULT DETECTION OF NONLINEAR SYSTEMS. Afef Fekih International Journal of Innovative Computing, Information and Control ICIC International c 2007 ISSN 1349-4198 Volume x, Number 0x, x 2005 pp. 0 0 NEURAL NETWORKS BASED SYSTEM IDENTIFICATION TECHNIQUES

More information

Lecture 9: Time-Domain Analysis of Discrete-Time Systems Dr.-Ing. Sudchai Boonto

Lecture 9: Time-Domain Analysis of Discrete-Time Systems Dr.-Ing. Sudchai Boonto Lecture 9: Time-Domain Analysis of Discrete-Time Systems Dr-Ing Sudchai Boonto Department of Control System and Instrumentation Engineering King Mongkut s Unniversity of Technology Thonburi Thailand Outline

More information

NONLINEAR PLANT IDENTIFICATION BY WAVELETS

NONLINEAR PLANT IDENTIFICATION BY WAVELETS NONLINEAR PLANT IDENTIFICATION BY WAVELETS Edison Righeto UNESP Ilha Solteira, Department of Mathematics, Av. Brasil 56, 5385000, Ilha Solteira, SP, Brazil righeto@fqm.feis.unesp.br Luiz Henrique M. Grassi

More information

Lecture 13 - Handling Nonlinearity

Lecture 13 - Handling Nonlinearity Lecture 3 - Handling Nonlinearit Nonlinearit issues in control practice Setpoint scheduling/feedforward path planning repla - linear interpolation Nonlinear maps B-splines Multivariable interpolation:

More information

Neural Network Based System Identification TOOLBOX Version 2

Neural Network Based System Identification TOOLBOX Version 2 eural etwork Based System Identification OOLBOX Version 2 For Use with MALAB ^ y(t 2) ^ y(t ) ^ y(t) u(t 2) u(t ) Magnus ørgaard Department of Automation Department of Mathematical Modelling echnical Report

More information

Study on the use of neural networks in control systems

Study on the use of neural networks in control systems Study on the use of neural networks in control systems F. Rinaldi Politecnico di Torino & Italian Air Force, Italy Keywords: neural networks, automatic control Abstract The purpose of this report is to

More information

Analysis of Discrete-Time Systems

Analysis of Discrete-Time Systems TU Berlin Discrete-Time Control Systems TU Berlin Discrete-Time Control Systems 2 Stability Definitions We define stability first with respect to changes in the initial conditions Analysis of Discrete-Time

More information

Input-output data sets for development and benchmarking in nonlinear identification

Input-output data sets for development and benchmarking in nonlinear identification Input-output data sets for development and benchmarking in nonlinear identification Torbjörn Wigren Division of Systems and Control, Department of Information Technology, Uppsala University, PO box 337,

More information

Using Neural Networks for Identification and Control of Systems

Using Neural Networks for Identification and Control of Systems Using Neural Networks for Identification and Control of Systems Jhonatam Cordeiro Department of Industrial and Systems Engineering North Carolina A&T State University, Greensboro, NC 27411 jcrodrig@aggies.ncat.edu

More information

Lecture 5: Recurrent Neural Networks

Lecture 5: Recurrent Neural Networks 1/25 Lecture 5: Recurrent Neural Networks Nima Mohajerin University of Waterloo WAVE Lab nima.mohajerin@uwaterloo.ca July 4, 2017 2/25 Overview 1 Recap 2 RNN Architectures for Learning Long Term Dependencies

More information

Artificial Neural Networks" and Nonparametric Methods" CMPSCI 383 Nov 17, 2011!

Artificial Neural Networks and Nonparametric Methods CMPSCI 383 Nov 17, 2011! Artificial Neural Networks" and Nonparametric Methods" CMPSCI 383 Nov 17, 2011! 1 Todayʼs lecture" How the brain works (!)! Artificial neural networks! Perceptrons! Multilayer feed-forward networks! Error

More information

Lecture 4: Perceptrons and Multilayer Perceptrons

Lecture 4: Perceptrons and Multilayer Perceptrons Lecture 4: Perceptrons and Multilayer Perceptrons Cognitive Systems II - Machine Learning SS 2005 Part I: Basic Approaches of Concept Learning Perceptrons, Artificial Neuronal Networks Lecture 4: Perceptrons

More information

TIME DELAY TEMPERATURE CONTROL WITH IMC AND CLOSED-LOOP IDENTIFICATION

TIME DELAY TEMPERATURE CONTROL WITH IMC AND CLOSED-LOOP IDENTIFICATION TIME DELAY TEMPERATURE CONTROL WITH IMC AND CLOSED-LOOP IDENTIFICATION Naoto ABE, Katsuyoshi KARAKAWA and Hiroyuki ICHIHARA Department of Mechanical Engineering Informatics, Meiji University Kawasaki 4-857,

More information

Analysis of Discrete-Time Systems

Analysis of Discrete-Time Systems TU Berlin Discrete-Time Control Systems 1 Analysis of Discrete-Time Systems Overview Stability Sensitivity and Robustness Controllability, Reachability, Observability, and Detectabiliy TU Berlin Discrete-Time

More information

Online Identification And Control of A PV-Supplied DC Motor Using Universal Learning Networks

Online Identification And Control of A PV-Supplied DC Motor Using Universal Learning Networks Online Identification And Control of A PV-Supplied DC Motor Using Universal Learning Networks Ahmed Hussein * Kotaro Hirasawa ** Jinglu Hu ** * Graduate School of Information Science & Electrical Eng.,

More information

DESIGN OF AN ON-LINE TITRATOR FOR NONLINEAR ph CONTROL

DESIGN OF AN ON-LINE TITRATOR FOR NONLINEAR ph CONTROL DESIGN OF AN ON-LINE TITRATOR FOR NONLINEAR CONTROL Alex D. Kalafatis Liuping Wang William R. Cluett AspenTech, Toronto, Canada School of Electrical & Computer Engineering, RMIT University, Melbourne,

More information

You submitted this quiz on Wed 16 Apr :18 PM IST. You got a score of 5.00 out of 5.00.

You submitted this quiz on Wed 16 Apr :18 PM IST. You got a score of 5.00 out of 5.00. Feedback IX. Neural Networks: Learning Help You submitted this quiz on Wed 16 Apr 2014 10:18 PM IST. You got a score of 5.00 out of 5.00. Question 1 You are training a three layer neural network and would

More information

Time domain identification, frequency domain identification. Equivalencies! Differences?

Time domain identification, frequency domain identification. Equivalencies! Differences? Time domain identification, frequency domain identification. Equivalencies! Differences? J. Schoukens, R. Pintelon, and Y. Rolain Vrije Universiteit Brussel, Department ELEC, Pleinlaan, B5 Brussels, Belgium

More information

Combining Regressive and Auto-Regressive Models for Spatial-Temporal Prediction

Combining Regressive and Auto-Regressive Models for Spatial-Temporal Prediction Combining Regressive and Auto-Regressive Models for Spatial-Temporal Prediction Dragoljub Pokrajac DPOKRAJA@EECS.WSU.EDU Zoran Obradovic ZORAN@EECS.WSU.EDU School of Electrical Engineering and Computer

More information

POWER SYSTEM DYNAMIC SECURITY ASSESSMENT CLASSICAL TO MODERN APPROACH

POWER SYSTEM DYNAMIC SECURITY ASSESSMENT CLASSICAL TO MODERN APPROACH Abstract POWER SYSTEM DYNAMIC SECURITY ASSESSMENT CLASSICAL TO MODERN APPROACH A.H.M.A.Rahim S.K.Chakravarthy Department of Electrical Engineering K.F. University of Petroleum and Minerals Dhahran. Dynamic

More information

ADAPTIVE NEURO-FUZZY INFERENCE SYSTEMS

ADAPTIVE NEURO-FUZZY INFERENCE SYSTEMS ADAPTIVE NEURO-FUZZY INFERENCE SYSTEMS RBFN and TS systems Equivalent if the following hold: Both RBFN and TS use same aggregation method for output (weighted sum or weighted average) Number of basis functions

More information

4.0 Update Algorithms For Linear Closed-Loop Systems

4.0 Update Algorithms For Linear Closed-Loop Systems 4. Update Algorithms For Linear Closed-Loop Systems A controller design methodology has been developed that combines an adaptive finite impulse response (FIR) filter with feedback. FIR filters are used

More information

An Improved Differential Evolution Trained Neural Network Scheme for Nonlinear System Identification

An Improved Differential Evolution Trained Neural Network Scheme for Nonlinear System Identification International Journal of Automation and Computing 6(2), May 2009, 137-144 DOI: 10.1007/s11633-009-0137-0 An Improved Differential Evolution Trained Neural Network Scheme for Nonlinear System Identification

More information

Predictive Control Design Based on Neural Model of a Non-linear System

Predictive Control Design Based on Neural Model of a Non-linear System Acta Polytechnica Hungarica Vol. 5, No. 4, 008 Predictive Control Design Based on Neural Model of a Non-linear System Anna Jadlovská, Nikola Kabakov, Ján Sarnovský Department of Cybernetics and Artificial

More information

A NEURO-FUZZY MODEL PREDICTIVE CONTROLLER APPLIED TO A PH-NEUTRALIZATION PROCESS. Jonas B. Waller and Hannu T. Toivonen

A NEURO-FUZZY MODEL PREDICTIVE CONTROLLER APPLIED TO A PH-NEUTRALIZATION PROCESS. Jonas B. Waller and Hannu T. Toivonen Copyright 22 IFAC 15th Triennial World Congress, Barcelona, Spain A NEURO-FUZZY MODEL PREDICTIVE CONTROLLER APPLIED TO A PH-NEUTRALIZATION PROCESS Jonas B. Waller and Hannu T. Toivonen Department of Chemical

More information

Application of Neural Networks for Control of Inverted Pendulum

Application of Neural Networks for Control of Inverted Pendulum Application of Neural Networks for Control of Inverted Pendulum VALERI MLADENOV Department of Theoretical Electrical Engineering Technical University of Sofia Sofia, Kliment Ohridski blvd. 8; BULARIA valerim@tu-sofia.bg

More information

Keywords- Source coding, Huffman encoding, Artificial neural network, Multilayer perceptron, Backpropagation algorithm

Keywords- Source coding, Huffman encoding, Artificial neural network, Multilayer perceptron, Backpropagation algorithm Volume 4, Issue 5, May 2014 ISSN: 2277 128X International Journal of Advanced Research in Computer Science and Software Engineering Research Paper Available online at: www.ijarcsse.com Huffman Encoding

More information

Neural Network Modeling and Identification of Naturally Ventilated Tropical Greenhouse Climates

Neural Network Modeling and Identification of Naturally Ventilated Tropical Greenhouse Climates Neural Network Modeling and Identification of Naturally Ventilated Tropical Greenhouse Climates ROBIAH AHMAD, MD.NORAZLAN MD.LAZIN, SAIFUL FARHAN MOHD SAMSURI 3 UTM RAZAK School of Engineering and Advanced

More information

Neural Networks Lecture 10: Fault Detection and Isolation (FDI) Using Neural Networks

Neural Networks Lecture 10: Fault Detection and Isolation (FDI) Using Neural Networks Neural Networks Lecture 10: Fault Detection and Isolation (FDI) Using Neural Networks H.A. Talebi Farzaneh Abdollahi Department of Electrical Engineering Amirkabir University of Technology Winter 2011.

More information

A Hybrid Time-delay Prediction Method for Networked Control System

A Hybrid Time-delay Prediction Method for Networked Control System International Journal of Automation and Computing 11(1), February 2014, 19-24 DOI: 10.1007/s11633-014-0761-1 A Hybrid Time-delay Prediction Method for Networked Control System Zhong-Da Tian Xian-Wen Gao

More information

SAMPLE SOLUTION TO EXAM in MAS501 Control Systems 2 Autumn 2015

SAMPLE SOLUTION TO EXAM in MAS501 Control Systems 2 Autumn 2015 FACULTY OF ENGINEERING AND SCIENCE SAMPLE SOLUTION TO EXAM in MAS501 Control Systems 2 Autumn 2015 Lecturer: Michael Ruderman Problem 1: Frequency-domain analysis and control design (15 pt) Given is a

More information

8.6 Bayesian neural networks (BNN) [Book, Sect. 6.7]

8.6 Bayesian neural networks (BNN) [Book, Sect. 6.7] 8.6 Bayesian neural networks (BNN) [Book, Sect. 6.7] While cross-validation allows one to find the weight penalty parameters which would give the model good generalization capability, the separation of

More information

CMU-Q Lecture 24:

CMU-Q Lecture 24: CMU-Q 15-381 Lecture 24: Supervised Learning 2 Teacher: Gianni A. Di Caro SUPERVISED LEARNING Hypotheses space Hypothesis function Labeled Given Errors Performance criteria Given a collection of input

More information

Forecasting River Flow in the USA: A Comparison between Auto-Regression and Neural Network Non-Parametric Models

Forecasting River Flow in the USA: A Comparison between Auto-Regression and Neural Network Non-Parametric Models Journal of Computer Science 2 (10): 775-780, 2006 ISSN 1549-3644 2006 Science Publications Forecasting River Flow in the USA: A Comparison between Auto-Regression and Neural Network Non-Parametric Models

More information

Designing Dynamic Neural Network for Non-Linear System Identification

Designing Dynamic Neural Network for Non-Linear System Identification Designing Dynamic Neural Network for Non-Linear System Identification Chandradeo Prasad Assistant Professor, Department of CSE, RGIT,Koderma Abstract : System identification deals with many subtleties

More information

Robust and Optimal Control, Spring A: SISO Feedback Control A.1 Internal Stability and Youla Parameterization

Robust and Optimal Control, Spring A: SISO Feedback Control A.1 Internal Stability and Youla Parameterization Robust and Optimal Control, Spring 2015 Instructor: Prof. Masayuki Fujita (S5-303B) A: SISO Feedback Control A.1 Internal Stability and Youla Parameterization A.2 Sensitivity and Feedback Performance A.3

More information

Multilayer Perceptrons (MLPs)

Multilayer Perceptrons (MLPs) CSE 5526: Introduction to Neural Networks Multilayer Perceptrons (MLPs) 1 Motivation Multilayer networks are more powerful than singlelayer nets Example: XOR problem x 2 1 AND x o x 1 x 2 +1-1 o x x 1-1

More information

MODEL PREDICTIVE CONTROL and optimization

MODEL PREDICTIVE CONTROL and optimization MODEL PREDICTIVE CONTROL and optimization Lecture notes Model Predictive Control PhD., Associate professor David Di Ruscio System and Control Engineering Department of Technology Telemark University College

More information

Multilayer Perceptron = FeedForward Neural Network

Multilayer Perceptron = FeedForward Neural Network Multilayer Perceptron = FeedForward Neural Networ History Definition Classification = feedforward operation Learning = bacpropagation = local optimization in the space of weights Pattern Classification

More information

Adaptive Dual Control

Adaptive Dual Control Adaptive Dual Control Björn Wittenmark Department of Automatic Control, Lund Institute of Technology Box 118, S-221 00 Lund, Sweden email: bjorn@control.lth.se Keywords: Dual control, stochastic control,

More information

New Recursive-Least-Squares Algorithms for Nonlinear Active Control of Sound and Vibration Using Neural Networks

New Recursive-Least-Squares Algorithms for Nonlinear Active Control of Sound and Vibration Using Neural Networks IEEE TRANSACTIONS ON NEURAL NETWORKS, VOL. 12, NO. 1, JANUARY 2001 135 New Recursive-Least-Squares Algorithms for Nonlinear Active Control of Sound and Vibration Using Neural Networks Martin Bouchard,

More information

This document is downloaded from DR-NTU, Nanyang Technological University Library, Singapore.

This document is downloaded from DR-NTU, Nanyang Technological University Library, Singapore. This document is downloaded from DR-NTU, Nanyang Technological University Library, Singapore. Title A robust recurrent simultaneous perturbation stochastic approximation training algorithm for recurrent

More information

CSC242: Intro to AI. Lecture 21

CSC242: Intro to AI. Lecture 21 CSC242: Intro to AI Lecture 21 Administrivia Project 4 (homeworks 18 & 19) due Mon Apr 16 11:59PM Posters Apr 24 and 26 You need an idea! You need to present it nicely on 2-wide by 4-high landscape pages

More information

Temporal Backpropagation for FIR Neural Networks

Temporal Backpropagation for FIR Neural Networks Temporal Backpropagation for FIR Neural Networks Eric A. Wan Stanford University Department of Electrical Engineering, Stanford, CA 94305-4055 Abstract The traditional feedforward neural network is a static

More information

Neuro-Fuzzy Methods for Modeling and Identification

Neuro-Fuzzy Methods for Modeling and Identification Neuro-Fuzzy Methods for Modeling and Identification Robert Babuška Delft University of Technology,Faculty of Information Technology and Systems Control Systems Engineering Group, P.O.Box 5031, 2600 GA

More information

Control System Design

Control System Design ELEC ENG 4CL4: Control System Design Notes for Lecture #36 Dr. Ian C. Bruce Room: CRL-229 Phone ext.: 26984 Email: ibruce@mail.ece.mcmaster.ca Friday, April 4, 2003 3. Cascade Control Next we turn to an

More information

DESIGNING A KALMAN FILTER WHEN NO NOISE COVARIANCE INFORMATION IS AVAILABLE. Robert Bos,1 Xavier Bombois Paul M. J. Van den Hof

DESIGNING A KALMAN FILTER WHEN NO NOISE COVARIANCE INFORMATION IS AVAILABLE. Robert Bos,1 Xavier Bombois Paul M. J. Van den Hof DESIGNING A KALMAN FILTER WHEN NO NOISE COVARIANCE INFORMATION IS AVAILABLE Robert Bos,1 Xavier Bombois Paul M. J. Van den Hof Delft Center for Systems and Control, Delft University of Technology, Mekelweg

More information

Iterative Feedback Tuning

Iterative Feedback Tuning Iterative Feedback Tuning Michel Gevers CESAME - UCL Louvain-la-Neuve Belgium Collaboration : H. Hjalmarsson, S. Gunnarsson, O. Lequin, E. Bosmans, L. Triest, M. Mossberg Outline Problem formulation Iterative

More information

inear Adaptive Inverse Control

inear Adaptive Inverse Control Proceedings of the 36th Conference on Decision & Control San Diego, California USA December 1997 inear Adaptive nverse Control WM15 1:50 Bernard Widrow and Gregory L. Plett Department of Electrical Engineering,

More information

ADAPTIVE NEURAL NETWORK MODEL PREDICTIVE CONTROL. Ramdane Hedjar. Received January 2012; revised May 2012

ADAPTIVE NEURAL NETWORK MODEL PREDICTIVE CONTROL. Ramdane Hedjar. Received January 2012; revised May 2012 International Journal of Innovative Computing, Information and Control ICIC International c 13 ISSN 1349-4198 Volume 9, Number 3, March 13 pp. 145 157 ADAPTIVE NEURAL NETWORK MODEL PREDICTIVE CONTROL Ramdane

More information

EECE Adaptive Control

EECE Adaptive Control EECE 574 - Adaptive Control Overview Guy Dumont Department of Electrical and Computer Engineering University of British Columbia Lectures: Thursday 09h00-12h00 Location: PPC 101 Guy Dumont (UBC) EECE 574

More information

Nonlinear Identification of Backlash in Robot Transmissions

Nonlinear Identification of Backlash in Robot Transmissions Nonlinear Identification of Backlash in Robot Transmissions G. Hovland, S. Hanssen, S. Moberg, T. Brogårdh, S. Gunnarsson, M. Isaksson ABB Corporate Research, Control Systems Group, Switzerland ABB Automation

More information

An Intelligent Nonlinear System Identification Method with an Application to Condition Monitoring

An Intelligent Nonlinear System Identification Method with an Application to Condition Monitoring Rochester Institute of Technology RIT Scholar Works Theses Thesis/Dissertation Collections 4-215 An Intelligent Nonlinear System Identification Method with an Application to Condition Monitoring Clara

More information

FRTN 15 Predictive Control

FRTN 15 Predictive Control Department of AUTOMATIC CONTROL FRTN 5 Predictive Control Final Exam March 4, 27, 8am - 3pm General Instructions This is an open book exam. You may use any book you want, including the slides from the

More information