Neural Networks based control of nonlinear systems

Slides:

Advertisements

Similar presentations

1 The 2-to-4 decoder is a block which decodes the 2-bit binary inputs and produces four output All but one outputs are zero One output corresponding to.

Advertisements

Part IV ：Enhancements to Single-Loop PID Feedback Control

Nonlinear & Neural Networks LAB. CHAPTER 11 LATCHES AND FLIP-FLOPS 11.1Introduction 11.2Set-Reset Latch 11.3Gated D Latch 11.4Edge-Triggered D Flip-Flop.

Scaled Helicopter Mathematical Model and Hovering Controller Brajtman Michal & Sharabani Yaki Supervisor : Dr. Rotstein Hector.

Nonlinear & Neural Networks LAB. CHAPTER 19 State Machine Design with SM charts 19.1 State Machine Charts 19.2 Derivation of SM Charts 19.3 Realization.

S. Mandayam/ ANN/ECE Dept./Rowan University Artificial Neural Networks ECE /ECE Fall 2010 Shreekanth Mandayam ECE Department Rowan University.

Neural Network Based Control Dan Simon Cleveland State University 1.

I welcome you all to this presentation On: Neural Network Applications Systems Engineering Dept. KFUPM Imran Nadeem & Naveed R. Butt &

Non Isothermal Chemical Reactor ChE 479 Alexander Couzis.

Presenting: Itai Avron Supervisor: Chen Koren Characterization Presentation Spring 2005 Implementation of Artificial Intelligence System on FPGA.

Process Control Computer Laboratory Dr. Emad M. Ali Chemical Engineering Department King SAUD University.

Introduction to Recurrent neural networks (RNN), Long short-term memory (LSTM) Wenjie Pei In this coffee talk, I would like to present you some basic.

Neural Networks Lab 5. What Is Neural Networks? Neural networks are composed of simple elements( Neurons) operating in parallel. Neural networks are composed.

Tariq Ahamed Wavelet Neural Control Of Cascaded Continuous Stirred Tank Reactors.

Neuron Model and Network Architecture

Eduard Petlenkov, Associate Professor, TUT Department of Computer Control

Data Mining Techniques in Stock Market Prediction

NIMIA October 2001, Crema, Italy - Vincenzo Piuri, University of Milan, Italy NEURAL NETWORKS FOR SENSORS AND MEASUREMENT SYSTEMS Part II Vincenzo.

1 Pattern Recognition: Statistical and Neural Lonnie C. Ludeman Lecture 20 Oct 26, 2005 Nanjing University of Science & Technology.

Radial Basis Function Networks:

Learning From Demonstration. Robot Learning A good control policy u=  (x,t) is often hard to engineer from first principles Reinforcement learning 

An informal description of artificial neural networks John MacCormick.

1 Adaptive Control Neural Networks 13(2000): Neural net based MRAC for a class of nonlinear plants M.S. Ahmed.

Dünaamiliste süsteemide modelleerimine Identification for control in a nonlinear system world Eduard Petlenkov, Automaatikainstituut, 2013.

Approaches to task solution of complex objects identification as example of an induction motor Mubarakzyanov N.R.

CHAPTER 7 MULTI-LEVEL GATE CIRCUITS / NAND AND NOR GATES

Nonlinear & Neural Networks LAB. CHAPTER 8 Combinational Circuit design and Simulation Using Gate 8.1Review of Combinational Circuit Design 8.2Design of.

Introduction to the first laboratory exercise Continuous neural network sliding mode controller DSP2 realization Andreja Rojko

An Artificial Neural Network Approach to Surface Waviness Prediction in Surface Finishing Process by Chi Ngo ECE/ME 539 Class Project.

Eduard Petlenkov, TTÜ automaatikainstituudi dotsent

Modelleerimine ja Juhtimine Tehisnärvivõrgudega Identification and Control with artificial neural networks (2nd part) Eduard Petlenkov, Automaatikainstituut.

Logarithm Basics. The logarithm base a of b is the exponent you put on a to get b: i.e. Logs give you exponents! Definition of Logarithm a > 0 and b >

Playing Tic-Tac-Toe with Neural Networks

CHEE825 Fall 2005J. McLellan1 Nonlinear Empirical Models.

Modelleerimine ja Juhtimine Tehisnärvivõrgudega Identification and Control with artificial neural networks.

Dynamic Neural Network Control (DNNC): A Non-Conventional Neural Network Model Masoud Nikravesh EECS Department, CS Division BISC Program University of.

Real Time Nonlinear Model Predictive Control Strategy for Multivariable Coupled Tank System Kayode Owa Kayode Owa Supervisor - Sanjay Sharma University.

Nonlinear balanced model residualization via neural networks Juergen Hahn.

CS206Evolutionary Robotics Artificial Neural Networks Input Layer Output Layer Input Layer Output Layer ++ ++

Intro. ANN & Fuzzy Systems Lecture 3 Basic Definitions of ANN.

Reactor Design. تحت شعار العيد فرحة : الجمهور : طبعا النهاردة نص يوم علشان العيد خلص امبارح؟ أنا : لأ الجمهور : يعني النهاردة هناخد سكشن؟ أنا : ونص الجمهور.

IGEM 2009 Tutorial Modelling. What? Model A model in science is a symplified physical, mathematical, or logical representation of a system of entities,

Eduard Petlenkov, Associate Professor, TUT Department of Computer Control

Creating Neural Networks & FL Systems Objectives : By the end of this session Student will be able to: 1- Create ANN & FL Systems using toolboxes. 2-

Ryszard Gessing Silesian Technical University Gliwice, Poland

ANN-based program for Tablet PC character recognition

MATLAB Implementation of the Optimal Brain Surgeon (OBS) Algorithm

Modelleerimine ja Juhtimine Tehisnärvivõrgudega

Jeopardy $100 $100 $100 $100 $100 $200 $200 $200 $200 $200 $300 $300

Feature Selection for Pattern Recognition

Date of download: 12/26/2017 Copyright © ASME. All rights reserved.

ISS0023 Intelligent Control Systems Arukad juhtimissüsteemid

ISS0023 Intelligent Control Systems Arukad juhtimissüsteemid

ISS0023 Intelligent Control Systems Arukad juhtimissüsteemid

Artificial Intelligence ppt

Java Implementation of Optimal Brain Surgeon

Perceptrons for Dummies

Neural Networks based control of nonlinear systems

Neural Networks & MPIC.

شبکه عصبی تنظیم: بهروز نصرالهی-فریده امدادی استاد محترم: سرکار خانم کریمی دانشگاه آزاد اسلامی واحد شهرری.

Neural Networks & MPIC.

أنماط الإدارة المدرسية وتفويض السلطة الدكتور أشرف الصايغ

Notes Over 7.4 Finding an Inverse Relation

Exponential and Logarithmic Forms

Masoud Nikravesh EECS Department, CS Division BISC Program

Dynamics of Training Noh, Yung-kyun Mar. 11, 2003

Bell Work: Machine Lab What is the equation to find Input Work? (look at note guide from yesterday) Calculate the input work for your ramp in trial 1.

13. Reactor Engineering: Reactor Design

Learning Combinational Logic

Presentation transcript:

Neural Networks based control of nonlinear systems Lab Neural Networks based control of nonlinear systems

Identification of dynamic systems with Artificial Neural Networks Dynamic/ feedback network: Eduard Petlenkov, Automaatikainstituut, 2013

Identification with Artificial Neural Networks is theinput of the system and the model is the output of the system is the output of the model Eduard Petlenkov, Automaatikainstituut, 2013

Identification with Artificial Neural Networks Eduard Petlenkov, Automaatikainstituut, 2011

Inverse model Eduard Petlenkov, Automaatikainstituut, 2013

Inverse model Consider a system: Eduard Petlenkov, Automaatikainstituut, 2013

Inverse model based control Siin Eduard Petlenkov, Automaatikainstituut, 2013

Inverse model based adaptive control Eduard Petlenkov, Automaatikainstituut, 2013

Example: Jacketed CSTR (Continuous Stirred Tank Reactor) Input-Output equation:

Collecting input-output data

P=[output(3:N)';output(2:N-1)';output(1:N-2)'] T=input(2:N-1)' Training of inverse model td=1 N=size(output,1) P=[output(3:N)';output(2:N-1)';output(1:N-2)'] T=input(2:N-1)' global net_c net_c=newff([0 1; 0 1; 0 1],[5 1],{'tansig','purelin'}) net_c.trainParam.show=1; net_c.trainFcn='traingd'; net_c.trainParam.epochs=3000; net_c=train(net_c,P,T)

Implementation of the controller in MATLAB (m-file controller.m) function control=controller(u) global net_c control=sim(net_c,u);

Control scheme

“dynamic_controller” block

Adaptive control scheme

“adaptive_dynamic_controller” block

function control=controller_adaptive(u) global net_c inp=u(1:3); Adaptive controller = function “controller_adaptive” (m-file controller_adaptive.m) function control=controller_adaptive(u) global net_c inp=u(1:3); error=u(4); time=u(5); control=sim(net_c,inp); if time>10 net_c=adapt(net_c,inp,control+error); end