A novel intelligent predictive maintenance procedure for electrical machines

Size: px
Start display at page:

Download "A novel intelligent predictive maintenance procedure for electrical machines"

Transcription

1 A novel intelligent predictive maintenance procedure for electrical machines D.-M. Yang Department of Automation Engineering, Kao-Yuan University, No.1821 Chung-Shan Road, Loju Hsiang, Kaohsiung County, Taiwan 821 Abstract In this paper, a novel procedure has been developed for the condition monitoring of electrical machines. It is based on wavelet analysis and a singular value decomposition approach. The singular value decomposition method is used to extract salient features from the continuous wavelet coefficients. These features are used as inputs to an artificial neural networ (ANN) trained to identify the conditions of the monitored electrical machine under steady state operation. The ANNs are used to be an intelligent electrical machine condition detector. Results presented have shown that the approach developed can successfully applied and this proposed approach is effective for condition monitoring of electrical machines. KEY WORDS : Artificial neural networ, singular value decomposition, wavelet analysis 1 Introduction Diagnosis of malfunctions of process automation is important in modern manufacturing industries. Today s highly automated complex machinery systems require intelligently automated maintenance systems to effectively achieve the goals of Computer Aided Manufacturing (CAM) and Computer Integrated Manufacturing (CIM) and the maintenance function should involved in the move toward highly automated and integrated manufacturing systems. Three maintenance strategies have practically used in industry. The first is simply react to the machine breadown as and when it happens. This maintenance scheme is called breadown maintenance. This strategy can be used if the machine is inexpensive and the breadown does not cause any other damage. Otherwise, the cost of lost production, safety riss, and additional damage to other machines mae this scheme unacceptable. The second is to perform fixed time interval maintenance. Although this method reduced the chance of unexpected breadown, it has been found to be uneconomical. The stoppage for maintenance involves not only lost production time but also a high ris of introducing imperfections due to human error. The third is predictive (or condition-based) maintenance through condition monitoring, which depends on the condition of the machine. Condition-based maintenance (CBM) consists of continuously evaluating the condition of a monitored machine, without interrupting its operation, and successfully predicting the presence of faults before catastrophic breadown occurs. Such an approach can provide for properly planned maintenance and replacement of failing components at optimum periods. Thus, the application of a condition monitoring-based maintenance policy can help to minimise unnecessary costs and delays caused by the need to carry out unscheduled repairs. Induction machines are critical components in many industrial processes. In spite of their robustness and reliability, some faults are still unavoidable. Therefore, proper monitoring of electrical machines is highly cost-effective in reducing operation cost. The wavelet approach has advantages over traditional Fourier methods for signal analysis, particularly for signals containing discontinuities and shape spie. When Fourier transformation is performed, the

2 signal representation is moved from the time-domain to the frequency-domain. And this change of domain can lead to loss of information and to interpretation difficulties. This disadvantage is overcome in wavelet analysis which represents a signal by using shifted and scaled versions of a so-called mother wavelet and this enables examination of the frequency information of the signal as it evolves with time. This ability to represent simultaneously both time-domain and frequency-domain information is a significant advantage of the wavelet approach. The complex nature of information obtained from wavelet transforms mae subjective interpretation and diagnostic use difficult. In order to effectively interpret the wavelet map, the time-frequency domain is used instead of time-scale domain. Hence, time-frequency distributions of a couple of specifically simulated signals and the associated singular values performed by the singular value decomposition (SVD) technique are presented in this paper. In order to develop a reliable automatic diagnosis procedure to monitor the motor condition, artificial neural networs are employed as a motor automatic condition detector since they do not require an indepth nowledge of the behaviour of the motor system. Neural networs is a suitable analysis tool (Awadallah & Morcos, 2003) since their structure allows them to be trained to learn the characteristics of various motor conditions under different operating models. The training is performed using data obtained from the motor under various operating conditions. A training period is used to establish the weights and biases of the inter-connections of the networ. Once appropriately trained, the inter-connections within the networ itself form the desired input-output mapping which enhances recognition and diagnosis of the various motor conditions. In the following sections, a novel diagnostic procedure contains three procedures which will be introduced briefly in Section 2. Section 3 presents the simulation results of the proposed approach for a couple of specifically simulated signals and a real-time data acquisition system for experimental data collection. Section 4 introduces the implementation of artificial neural networs for automatic detection of the induction electrical machine condition and a conclusion will be presented in Section 6. 2 A novel diagnostic procedure 2.1 Wavelet analysis as the signal processing technique Wavelet analysis is an approach which decomposes a time-domain signal into components in different time windows and different frequency bands and presents the resulting information in the form of a surface in the time frequency plane, sometimes referred to as a scalogram (Rioul & Vetterli, 1991). The scalogram is similar in concept to the spectrogram but differs from in that the frequency resolution of the scalogram is logarithmic rather than linear, as is the case for the spectrogram. Because of the nature of the frequency resolution, the wavelet approach is more effective in analysing both the long-time, lowfrequency and the short-time, high-frequency content of a time signal. This characteristic is very useful for analysing pulse-lie and non-stationary signals.

3 Fig. 1 Morlet wavelet for f o = 1 The continuous wavelet transform of a square-integrable, continuous time signal x(t) is the inner product between x (t) and the analysing wavelet ψ ( ), which gives the wavelet coefficients a, b t 1 + * t b Wx ( a, b) = x( t), ψ a, b ( t) = x( t) ψ ( ) dt a a (1) where a is the dilation parameter governing the wavelet frequency b is the parameter localizing the wavelet function in the time domain and ψ * ( t) is the complex conjugate of the analysing wavelet ψ (t). There are a number of different complex or real valued functions used as analysing wavelets. The analysing wavelet used in this paper is the complex-valued Morlet wavelet (Farge, 1992), given by 2 M j2πf 0t t / 2 ψ ( t) = e e (2) where f is the central frequency of the Morlet wavelet and the value of f is taen 1 in this paper. o Figure 1 shows the real and imaginary parts of the Morlet wavelet together with its confining Gaussian envelope. An alternative formulation of the continuous wavelet transform can be obtained by transforming both the signal x (t) and the analysing wavelet ψ ( ) in the frequency domain: a, b t o ) * j(2πf ) b W ( a, b) = a X ( f ) Ψ ( af ) e df x + (3) * j(2πf ) b where X ( f ) and Ψ ) * t b ( af ) e are the Fourier transform of x (t) and ψ ( ) respectively. For an a easier implementation of wavelet transform, Equation (3) can be expressed in a discrete form as: ) W( a, b ) = a X ( f ) Ψ 2.2 Singular value decomposition for feature extraction m n m * ( a m f ) e j2πf b n The purpose of using a feature extraction procedure is to significantly reduce the quantity of data but, at the same time, retain the original salient information. It is well nown that the SVD has optimal decorrelation properties and provides a means of detecting dominant characteristics on a set of data. The (4)

4 SVD theorem states that any m by n matrix W can be decomposed and written as the product of matrices (Therrien, 1992). p T T W ( a, b ) = U V = λ u v (5) m n mm mn where the superscript, T, donates the transpose. U is an orthogonal m by m matrix made up of left singular vectors u, V is an orthogonal n by n matrix made up of right singular vectors v, and is an m by n matrix of non-negative real singular values and decrease monotonically from the upper left to the lower right of. The singular value of W are represented by λ 1 λ 2 λ 3... λp 0, where p = min( m, n) In most machine condition monitoring and diagnosis schemes, only the first one or two of the singular values will be considered since these account for the large majority of variations in the analysed data set. 2.3 Artificial neural networs as an automatic classifier Artificial neural networs (ANNs) can be used to identify and classify complex fault patterns without the need for a detailed nowledge of the behaviour of the system in which they occur or of the mechanics responsible for the generation of their characteristics. ANN design is premised on mimicing the human nervous system by using massively parallel nets composed of many computational elements connected by lins with adjustable weights. Of all the ANN types in current use, the multi-layer perceptron (MLP), trained using the bac-propagation algorithm (or a variant), is the most commonly applied. A full description of the MLP can be found in (Lippman, 1987). Pre-processing of the signals to be used as inputs to the networs can mae neural networ training more efficient due to a significant reduction of the dimensionality of the input data. Before training, it is also useful to scale the inputs and target outputs so that their values fall within a specified range. The normalization procedure used here, described in (Bishop, 1995), is V Vmin V n = ( ) (6) V V max min where V is the maximum magnitude across all of the input patterns and V is the minimum max magnitude across all of the input patterns. When the ANN training is complete, the performance of the ANN can be judged by the success rate (SR) of how many unseen test patterns are correctly classifier. The SR is defined as follows: UPC SR = 100% (7) TP where UPC is the number of unseen patterns correctly classified and TP is the total number of patterns. 3 Implementation In order to evaluate the performance of the proposed procedure, both simulated time series data and real time machine data collected in lab experiments are used to demonstrate the approach. 3.1 Simulated data generation nn = 1 min

5 Fig. 2 (a) A periodic impulse signal Fig.2 (b) Wavelet representation of Fig. 2(a) Fig. 2(c) The first right singular vector ( v1 ) Fig. 2(d) The second right singular vector ( v 2 ) To understand better wavelet representations, a number of simulated signals are analysed using the continuous wavelet transform approach and the singular value decomposition method for the preliminary study. The simulated signals are sampled at 1000 Hz. Figure 2(a) shows the periodic impulse signal x( t) = δ (cos(2πt)), where δ (t) is a Dirac delta function. A wide-band impulse frequency spectrum can be clearly seen from Figure 2(b). Figure 2(c) and (d) show the first and second right singular value decomposition from the wavelet coefficients. The impulse patterns in the periodic impulse signal can be easily caught by the SVD analysis. Figure 3(a) shows a stationary signal, which consists of two sinusoids, one at 25Hz and one at 60 Hz. The signal is given by: x( t) = cos(2π 25t) + cos(2π 60t) (6) In this signal, the frequencies and amplitudes of the two frequency components (25 Hz and 60Hz) do not change with time, which can be clearly identified in the wavelet representation, shown in Figure 3(b). The two sets of the first and second left singular vectors are plotted in Figures 3(c) and (d), which distinctively show the two separate frequency components in the stationary signal by the SVD analysis. From above simulation results obtained, they have demonstrated that SVD approach is capable of identifying salient patterns from stationary or non-stationary signals. From these singular values sets, they

6 can be used as characteristic features for on-line condition monitoring of electrical machines under various operations. Fig. 3(a) A stationary signal Fig. 3(b) Wavelet representation of Fig. 3(a) Fig. 3(c) The first left singular vector ( u1 ) Fig. 3(d) The second left singular vector ( u 2 ) 3.2 Experimental setup and real-time data acquisition Fig 4 : Experimental rig set-up layout

7 The experimental setup used in the study consists of a 2.2 W, 1740 rpm, 4-pole induction machine driving a 5 W DC generator via a flexible coupling, as shown in Figure 4. In order to test the proposed predictive maintenance procedure for condition monitoring of electrical machines, there are four conditions investigated in this experiment. The four conditions studied are: (1) Normal condition under no-load operation (NN) (2) Normal condition under full-load operation (NF) (3) Single phase open under no-load operation (SN) (4) Single phase open under full-load to dead operation (SF) A piezo-electric accelerometer was mounted on the housing of the induction electrical machine to measure vibration. The measured vibration signal was fed to the National Instruments SCC-ACC01 signal conditioning installed into the SC A real-time data acquisition device (Type NI 6062E) was used to record vibration signal. In this experiment, a total of 870 vibration signatures of 2048 points sampled at 4 Hz were recorded. 4 Experimental Results and discussion Figure 5 is depicted one of the vibration signatures for the induction electrical machine from normal operation to one of three phases open under no-load operation. Fig. 5 Vibration signal under no-load operation In order to automatic detection of the induction electrical machine condition, the multi-layer perceptron (MLP), trained using the bac-propagation algorithm is used in this study. Before training the ANN, it is useful to extract salient features. Otherwise garbage in and garbage out. The continuous wavelet transforms are used to perform signal analysis using equation (4) in order to get more diagnostic information. All the ANN input data are normalized between 0.1 and 0.9 using equation (6). The networ is trained using the bac-propagation algorithm. Since the MLP-type of neural networs is of the supervised learning class, target data is required to implement networ training. The target vectors are T defined as the four different patterns; T 1 = T [0.9,0.1,0.1,0.1 ], T 2 = [0.1,0.9,0.1,0.1 ], T T 1 = T [0.1,0.1,0.9,0.1 ] and T 1 = [0.1,0.1,0.1,0.9 ], for the no-load normal condition, full-load normal condition, no-load single phase open and full-load single phase open, respectively. 0.9 indicates the presence of a condition and 0.1 indicates absence of that condition.

8 The networs topology has 50 inputs nodes (corresponding to the first columns of the matrices U mm ), four output nodes (corresponding to four machine conditions examined) and 15 hidden nodes. The number of hidden nodes is determined using geometric pyramid rule which suggests the number of hidden layer nodes as being the square root of the product of number of input and output nodes (Hansen, 1998). The sigmoid function is chosen as the transfer function to ensure node outputs between 0 and 1. The collected files listed in the table 1 are divided into two groups. One is used to train the ANNs, the other is used to test the ANN performances. The learning coefficient and the momentum coefficient are chosen as 0.95 and 0.01, respectively. The trained networs gave 100% correct prediction for the training patterns, which indicates that the networs have been successfully trained. Table 1. File descriptions Case NN NF SN SF File index The performances of the trained networs were checed by using unseen patterns. The trained neural networs can classify the four different conditions with a high SR using equation (7), as listed in Table 2. Table 2. Performance of the networs Case NN NF SN SF SR 100% 100% 100% 80.71% 5 Conclusions The primary objective of this wor is focus on developing an integrated system using continuous wavelet transforms as a signal preprocessor and a SVD technique as salient feature extraction. The ANNs are used to be an intelligent motor condition detector. Results presented have shown that the approach developed can successfully applied and this proposed approach is effective for condition monitoring of electrical machines. Acnowledgements The author wish to than the National Science Council for their financial support under project numbers E and E References Awadallah, M.A. & Morcos M.M. (2003). Application of AI Tools in Fault Diagnosis of Electrical Machines and Drives An Overview. IEEE Transactions on Energy Conversion, 18(2), Bishop, C.M. (1995). Neural Networs for Pattern Recognition. Oxford: Oxford University Press. Farge, M. (1992). Wavelet Transforms and Their Applications to Turbulence. Annual Review of Fluid Mechanics, 24, Hansen, J.V. (1998). Comparative Performance of Bacpropagation Networs Designed by Genetic Algorithms and Heuristics. International Journal of Intelligent Systems In Accounting, Finance & Management. 7(2), Lippman, R.P. (1987). An Introduction to Computing with Neural Nets. IEEE ASSP Magazine. 4(2), Rioul, O and Vetterli M. (1991). Wavelets and Signal Processing. IEEE Signal Processing Magazine, 8,

9 Therrien, C.W. (1992). Discrete Random Signals and Statistical Signal Processing. Prentice Hall. New Jersy :Englewood Cliffs.

Induction Motor Bearing Fault Detection with Non-stationary Signal Analysis

Induction Motor Bearing Fault Detection with Non-stationary Signal Analysis Proceedings of International Conference on Mechatronics Kumamoto Japan, 8-1 May 7 ThA1-C-1 Induction Motor Bearing Fault Detection with Non-stationary Signal Analysis D.-M. Yang Department of Mechanical

More information

Application of cepstrum and neural network to bearing fault detection

Application of cepstrum and neural network to bearing fault detection Journal of Mechanical Science and Technology 23 (2009) 2730~2737 Journal of Mechanical Science and Technology www.springerlin.com/content/738-494x DOI 0.007/s2206-009-0802-9 Application of cepstrum and

More information

Application of Artificial Neural Networks in Evaluation and Identification of Electrical Loss in Transformers According to the Energy Consumption

Application of Artificial Neural Networks in Evaluation and Identification of Electrical Loss in Transformers According to the Energy Consumption Application of Artificial Neural Networks in Evaluation and Identification of Electrical Loss in Transformers According to the Energy Consumption ANDRÉ NUNES DE SOUZA, JOSÉ ALFREDO C. ULSON, IVAN NUNES

More information

Journal of Engineering Science and Technology Review 6 (2) (2013) Research Article. Received 25 June 2012; Accepted 15 January 2013

Journal of Engineering Science and Technology Review 6 (2) (2013) Research Article. Received 25 June 2012; Accepted 15 January 2013 Jestr Journal of Engineering Science and Technology Review 6 () (3) 5-54 Research Article JOURNAL OF Engineering Science and Technology Review www.jestr.org Fault Diagnosis and Classification in Urban

More information

A Machine Intelligence Approach for Classification of Power Quality Disturbances

A Machine Intelligence Approach for Classification of Power Quality Disturbances A Machine Intelligence Approach for Classification of Power Quality Disturbances B K Panigrahi 1, V. Ravi Kumar Pandi 1, Aith Abraham and Swagatam Das 1 Department of Electrical Engineering, IIT, Delhi,

More information

Intelligent Fault Classification of Rolling Bearing at Variable Speed Based on Reconstructed Phase Space

Intelligent Fault Classification of Rolling Bearing at Variable Speed Based on Reconstructed Phase Space Journal of Robotics, Networking and Artificial Life, Vol., No. (June 24), 97-2 Intelligent Fault Classification of Rolling Bearing at Variable Speed Based on Reconstructed Phase Space Weigang Wen School

More information

An artificial neural networks (ANNs) model is a functional abstraction of the

An artificial neural networks (ANNs) model is a functional abstraction of the CHAPER 3 3. Introduction An artificial neural networs (ANNs) model is a functional abstraction of the biological neural structures of the central nervous system. hey are composed of many simple and highly

More information

Improving Electromotor Process in Water Pump by Using Power Spectral Density, Time Signal and Fault Probability Distribution Function

Improving Electromotor Process in Water Pump by Using Power Spectral Density, Time Signal and Fault Probability Distribution Function Improving Electromotor Process in Water Pump by Using Power Spectral Density, Time Signal and Fault Probability Distribution Function Hojjat Ahmadi, Zeinab Khaksar Department of Agricultural Machinery

More information

Part 8: Neural Networks

Part 8: Neural Networks METU Informatics Institute Min720 Pattern Classification ith Bio-Medical Applications Part 8: Neural Netors - INTRODUCTION: BIOLOGICAL VS. ARTIFICIAL Biological Neural Netors A Neuron: - A nerve cell as

More information

ARTIFICIAL NEURAL NETWORK PART I HANIEH BORHANAZAD

ARTIFICIAL NEURAL NETWORK PART I HANIEH BORHANAZAD ARTIFICIAL NEURAL NETWORK PART I HANIEH BORHANAZAD WHAT IS A NEURAL NETWORK? The simplest definition of a neural network, more properly referred to as an 'artificial' neural network (ANN), is provided

More information

Measurement and interpretation of instantaneous autospectrum of cyclostationary noise

Measurement and interpretation of instantaneous autospectrum of cyclostationary noise The 33 rd International Congress and Exposition on Noise Control Engineering Measurement and interpretation of instantaneous autospectrum of cyclostationary noise K. Vokurka Physics Department, Technical

More information

Application of Fourier Descriptors & Artificial Neural Network to Bearing Vibration Signals for Fault Detection & Classification

Application of Fourier Descriptors & Artificial Neural Network to Bearing Vibration Signals for Fault Detection & Classification Universal Journal of Aeronautical & Aerospace Sciences 2 (2014), 37-54 www.papersciences.com Application of Fourier Descriptors & Artificial Neural Network to Bearing Vibration Signals for Fault Detection

More information

CSE 352 (AI) LECTURE NOTES Professor Anita Wasilewska. NEURAL NETWORKS Learning

CSE 352 (AI) LECTURE NOTES Professor Anita Wasilewska. NEURAL NETWORKS Learning CSE 352 (AI) LECTURE NOTES Professor Anita Wasilewska NEURAL NETWORKS Learning Neural Networks Classifier Short Presentation INPUT: classification data, i.e. it contains an classification (class) attribute.

More information

APPLICATION OF WAVELET TRANSFORM TO DETECT FAULTS IN ROTATING MACHINERY

APPLICATION OF WAVELET TRANSFORM TO DETECT FAULTS IN ROTATING MACHINERY APPLICATION OF WAVELET TRANSFORM TO DETECT FAULTS IN ROTATING MACHINERY Darley Fiácrio de Arruda Santiago UNICAMP / Universidade Estadual de Campinas Faculdade de Engenharia Mecânica CEFET-PI / Centro

More information

Rolling Element Bearing Faults Detection, Wavelet De-noising Analysis

Rolling Element Bearing Faults Detection, Wavelet De-noising Analysis Universal Journal of Mechanical Engineering 3(2): 47-51, 2015 DOI: 10.13189/ujme.2015.030203 http://www.hrpub.org Rolling Element Bearing Faults Detection, Wavelet De-noising Analysis Maamar Ali Saud AL-Tobi

More information

Introduction to Biomedical Engineering

Introduction to Biomedical Engineering Introduction to Biomedical Engineering Biosignal processing Kung-Bin Sung 6/11/2007 1 Outline Chapter 10: Biosignal processing Characteristics of biosignals Frequency domain representation and analysis

More information

CHAPTER 4 FAULT DIAGNOSIS OF BEARINGS DUE TO SHAFT RUB

CHAPTER 4 FAULT DIAGNOSIS OF BEARINGS DUE TO SHAFT RUB 53 CHAPTER 4 FAULT DIAGNOSIS OF BEARINGS DUE TO SHAFT RUB 4.1 PHENOMENON OF SHAFT RUB Unwanted contact between the rotating and stationary parts of a rotating machine is more commonly referred to as rub.

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

Novelty Detection based on Extensions of GMMs for Industrial Gas Turbines

Novelty Detection based on Extensions of GMMs for Industrial Gas Turbines Novelty Detection based on Extensions of GMMs for Industrial Gas Turbines Yu Zhang, Chris Bingham, Michael Gallimore School of Engineering University of Lincoln Lincoln, U.. {yzhang; cbingham; mgallimore}@lincoln.ac.uk

More information

Multilayer Perceptron

Multilayer Perceptron Outline Hong Chang Institute of Computing Technology, Chinese Academy of Sciences Machine Learning Methods (Fall 2012) Outline Outline I 1 Introduction 2 Single Perceptron 3 Boolean Function Learning 4

More information

Neural Networks and the Back-propagation Algorithm

Neural Networks and the Back-propagation Algorithm Neural Networks and the Back-propagation Algorithm Francisco S. Melo In these notes, we provide a brief overview of the main concepts concerning neural networks and the back-propagation algorithm. We closely

More information

Bearing fault diagnosis based on EMD-KPCA and ELM

Bearing fault diagnosis based on EMD-KPCA and ELM Bearing fault diagnosis based on EMD-KPCA and ELM Zihan Chen, Hang Yuan 2 School of Reliability and Systems Engineering, Beihang University, Beijing 9, China Science and Technology on Reliability & Environmental

More information

Multiple Wavelet Coefficients Fusion in Deep Residual Networks for Fault Diagnosis

Multiple Wavelet Coefficients Fusion in Deep Residual Networks for Fault Diagnosis Multiple Wavelet Coefficients Fusion in Deep Residual Networks for Fault Diagnosis Minghang Zhao, Myeongsu Kang, Baoping Tang, Michael Pecht 1 Backgrounds Accurate fault diagnosis is important to ensure

More information

Eliminating the Influence of Harmonic Components in Operational Modal Analysis

Eliminating the Influence of Harmonic Components in Operational Modal Analysis Eliminating the Influence of Harmonic Components in Operational Modal Analysis Niels-Jørgen Jacobsen Brüel & Kjær Sound & Vibration Measurement A/S Skodsborgvej 307, DK-2850 Nærum, Denmark Palle Andersen

More information

Application of Fully Recurrent (FRNN) and Radial Basis Function (RBFNN) Neural Networks for Simulating Solar Radiation

Application of Fully Recurrent (FRNN) and Radial Basis Function (RBFNN) Neural Networks for Simulating Solar Radiation Bulletin of Environment, Pharmacology and Life Sciences Bull. Env. Pharmacol. Life Sci., Vol 3 () January 04: 3-39 04 Academy for Environment and Life Sciences, India Online ISSN 77-808 Journal s URL:http://www.bepls.com

More information

Impeller Fault Detection for a Centrifugal Pump Using Principal Component Analysis of Time Domain Vibration Features

Impeller Fault Detection for a Centrifugal Pump Using Principal Component Analysis of Time Domain Vibration Features Impeller Fault Detection for a Centrifugal Pump Using Principal Component Analysis of Time Domain Vibration Features Berli Kamiel 1,2, Gareth Forbes 2, Rodney Entwistle 2, Ilyas Mazhar 2 and Ian Howard

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

A FUZZY NEURAL NETWORK MODEL FOR FORECASTING STOCK PRICE

A FUZZY NEURAL NETWORK MODEL FOR FORECASTING STOCK PRICE A FUZZY NEURAL NETWORK MODEL FOR FORECASTING STOCK PRICE Li Sheng Institute of intelligent information engineering Zheiang University Hangzhou, 3007, P. R. China ABSTRACT In this paper, a neural network-driven

More information

Radial Basis Function Networks. Ravi Kaushik Project 1 CSC Neural Networks and Pattern Recognition

Radial Basis Function Networks. Ravi Kaushik Project 1 CSC Neural Networks and Pattern Recognition Radial Basis Function Networks Ravi Kaushik Project 1 CSC 84010 Neural Networks and Pattern Recognition History Radial Basis Function (RBF) emerged in late 1980 s as a variant of artificial neural network.

More information

ON NUMERICAL ANALYSIS AND EXPERIMENT VERIFICATION OF CHARACTERISTIC FREQUENCY OF ANGULAR CONTACT BALL-BEARING IN HIGH SPEED SPINDLE SYSTEM

ON NUMERICAL ANALYSIS AND EXPERIMENT VERIFICATION OF CHARACTERISTIC FREQUENCY OF ANGULAR CONTACT BALL-BEARING IN HIGH SPEED SPINDLE SYSTEM ON NUMERICAL ANALYSIS AND EXPERIMENT VERIFICATION OF CHARACTERISTIC FREQUENCY OF ANGULAR CONTACT BALL-BEARING IN HIGH SPEED SPINDLE SYSTEM Tian-Yau Wu and Chun-Che Sun Department of Mechanical Engineering,

More information

Introduction to Artificial Neural Networks

Introduction to Artificial Neural Networks Facultés Universitaires Notre-Dame de la Paix 27 March 2007 Outline 1 Introduction 2 Fundamentals Biological neuron Artificial neuron Artificial Neural Network Outline 3 Single-layer ANN Perceptron Adaline

More information

ADAPTIVE INVERSE CONTROL BASED ON NONLINEAR ADAPTIVE FILTERING. Information Systems Lab., EE Dep., Stanford University

ADAPTIVE INVERSE CONTROL BASED ON NONLINEAR ADAPTIVE FILTERING. Information Systems Lab., EE Dep., Stanford University ADAPTIVE INVERSE CONTROL BASED ON NONLINEAR ADAPTIVE FILTERING Bernard Widrow 1, Gregory Plett, Edson Ferreira 3 and Marcelo Lamego 4 Information Systems Lab., EE Dep., Stanford University Abstract: Many

More information

Artificial Neural Network

Artificial Neural Network Artificial Neural Network Contents 2 What is ANN? Biological Neuron Structure of Neuron Types of Neuron Models of Neuron Analogy with human NN Perceptron OCR Multilayer Neural Network Back propagation

More information

Artificial Intelligence (AI) Common AI Methods. Training. Signals to Perceptrons. Artificial Neural Networks (ANN) Artificial Intelligence

Artificial Intelligence (AI) Common AI Methods. Training. Signals to Perceptrons. Artificial Neural Networks (ANN) Artificial Intelligence Artificial Intelligence (AI) Artificial Intelligence AI is an attempt to reproduce intelligent reasoning using machines * * H. M. Cartwright, Applications of Artificial Intelligence in Chemistry, 1993,

More information

Identification and Classification of High Impedance Faults using Wavelet Multiresolution Analysis

Identification and Classification of High Impedance Faults using Wavelet Multiresolution Analysis 92 NATIONAL POWER SYSTEMS CONFERENCE, NPSC 2002 Identification Classification of High Impedance Faults using Wavelet Multiresolution Analysis D. Cha N. K. Kishore A. K. Sinha Abstract: This paper presents

More information

Hand Written Digit Recognition using Kalman Filter

Hand Written Digit Recognition using Kalman Filter International Journal of Electronics and Communication Engineering. ISSN 0974-2166 Volume 5, Number 4 (2012), pp. 425-434 International Research Publication House http://www.irphouse.com Hand Written Digit

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

DAMAGE ASSESSMENT OF REINFORCED CONCRETE USING ULTRASONIC WAVE PROPAGATION AND PATTERN RECOGNITION

DAMAGE ASSESSMENT OF REINFORCED CONCRETE USING ULTRASONIC WAVE PROPAGATION AND PATTERN RECOGNITION II ECCOMAS THEMATIC CONFERENCE ON SMART STRUCTURES AND MATERIALS C.A. Mota Soares et al. (Eds.) Lisbon, Portugal, July 18-21, 2005 DAMAGE ASSESSMENT OF REINFORCED CONCRETE USING ULTRASONIC WAVE PROPAGATION

More information

Wear State Recognition of Drills Based on K-means Cluster and Radial Basis Function Neural Network

Wear State Recognition of Drills Based on K-means Cluster and Radial Basis Function Neural Network International Journal of Automation and Computing 7(3), August 2010, 271-276 DOI: 10.1007/s11633-010-0502-z Wear State Recognition of Drills Based on K-means Cluster and Radial Basis Function Neural Network

More information

COMP-4360 Machine Learning Neural Networks

COMP-4360 Machine Learning Neural Networks COMP-4360 Machine Learning Neural Networks Jacky Baltes Autonomous Agents Lab University of Manitoba Winnipeg, Canada R3T 2N2 Email: jacky@cs.umanitoba.ca WWW: http://www.cs.umanitoba.ca/~jacky http://aalab.cs.umanitoba.ca

More information

Unit 8: Introduction to neural networks. Perceptrons

Unit 8: Introduction to neural networks. Perceptrons Unit 8: Introduction to neural networks. Perceptrons D. Balbontín Noval F. J. Martín Mateos J. L. Ruiz Reina A. Riscos Núñez Departamento de Ciencias de la Computación e Inteligencia Artificial Universidad

More information

Data Mining Part 5. Prediction

Data Mining Part 5. Prediction Data Mining Part 5. Prediction 5.5. Spring 2010 Instructor: Dr. Masoud Yaghini Outline How the Brain Works Artificial Neural Networks Simple Computing Elements Feed-Forward Networks Perceptrons (Single-layer,

More information

Mr. Harshit K. Dave 1, Dr. Keyur P. Desai 2, Dr. Harit K. Raval 3

Mr. Harshit K. Dave 1, Dr. Keyur P. Desai 2, Dr. Harit K. Raval 3 Investigations on Prediction of MRR and Surface Roughness on Electro Discharge Machine Using Regression Analysis and Artificial Neural Network Programming Mr. Harshit K. Dave 1, Dr. Keyur P. Desai 2, Dr.

More information

Intelligent Modular Neural Network for Dynamic System Parameter Estimation

Intelligent Modular Neural Network for Dynamic System Parameter Estimation Intelligent Modular Neural Network for Dynamic System Parameter Estimation Andrzej Materka Technical University of Lodz, Institute of Electronics Stefanowskiego 18, 9-537 Lodz, Poland Abstract: A technique

More information

Robust Controller Design for Speed Control of an Indirect Field Oriented Induction Machine Drive

Robust Controller Design for Speed Control of an Indirect Field Oriented Induction Machine Drive Leonardo Electronic Journal of Practices and Technologies ISSN 1583-1078 Issue 6, January-June 2005 p. 1-16 Robust Controller Design for Speed Control of an Indirect Field Oriented Induction Machine Drive

More information

CMSC 421: Neural Computation. Applications of Neural Networks

CMSC 421: Neural Computation. Applications of Neural Networks CMSC 42: Neural Computation definition synonyms neural networks artificial neural networks neural modeling connectionist models parallel distributed processing AI perspective Applications of Neural Networks

More information

HIGH PERFORMANCE ADAPTIVE INTELLIGENT DIRECT TORQUE CONTROL SCHEMES FOR INDUCTION MOTOR DRIVES

HIGH PERFORMANCE ADAPTIVE INTELLIGENT DIRECT TORQUE CONTROL SCHEMES FOR INDUCTION MOTOR DRIVES HIGH PERFORMANCE ADAPTIVE INTELLIGENT DIRECT TORQUE CONTROL SCHEMES FOR INDUCTION MOTOR DRIVES M. Vasudevan and R. Arumugam Department of Electrical and Electronics Engineering, Anna University, Chennai,

More information

Deep Neural Networks (1) Hidden layers; Back-propagation

Deep Neural Networks (1) Hidden layers; Back-propagation Deep Neural Networs (1) Hidden layers; Bac-propagation Steve Renals Machine Learning Practical MLP Lecture 3 4 October 2017 / 9 October 2017 MLP Lecture 3 Deep Neural Networs (1) 1 Recap: Softmax single

More information

Neural Network Based Methodology for Cavitation Detection in Pressure Dropping Devices of PFBR

Neural Network Based Methodology for Cavitation Detection in Pressure Dropping Devices of PFBR Indian Society for Non-Destructive Testing Hyderabad Chapter Proc. National Seminar on Non-Destructive Evaluation Dec. 7-9, 2006, Hyderabad Neural Network Based Methodology for Cavitation Detection in

More information

COMPARING PERFORMANCE OF NEURAL NETWORKS RECOGNIZING MACHINE GENERATED CHARACTERS

COMPARING PERFORMANCE OF NEURAL NETWORKS RECOGNIZING MACHINE GENERATED CHARACTERS Proceedings of the First Southern Symposium on Computing The University of Southern Mississippi, December 4-5, 1998 COMPARING PERFORMANCE OF NEURAL NETWORKS RECOGNIZING MACHINE GENERATED CHARACTERS SEAN

More information

Computational and Experimental Approach for Fault Detection of Gears

Computational and Experimental Approach for Fault Detection of Gears Columbia International Publishing Journal of Vibration Analysis, Measurement, and Control (2014) Vol. 2 No. 1 pp. 16-29 doi:10.7726/jvamc.2014.1002 Research Article Computational and Experimental Approach

More information

The Fault extent recognition method of rolling bearing based on orthogonal matching pursuit and Lempel-Ziv complexity

The Fault extent recognition method of rolling bearing based on orthogonal matching pursuit and Lempel-Ziv complexity The Fault extent recognition method of rolling bearing based on orthogonal matching pursuit and Lempel-Ziv complexity Pengfei Dang 1,2 *, Yufei Guo 2,3, Hongjun Ren 2,4 1 School of Mechanical Engineering,

More information

Identification Techniques for Operational Modal Analysis An Overview and Practical Experiences

Identification Techniques for Operational Modal Analysis An Overview and Practical Experiences Identification Techniques for Operational Modal Analysis An Overview and Practical Experiences Henrik Herlufsen, Svend Gade, Nis Møller Brüel & Kjær Sound and Vibration Measurements A/S, Skodsborgvej 307,

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

A. Pelliccioni (*), R. Cotroneo (*), F. Pungì (*) (*)ISPESL-DIPIA, Via Fontana Candida 1, 00040, Monteporzio Catone (RM), Italy.

A. Pelliccioni (*), R. Cotroneo (*), F. Pungì (*) (*)ISPESL-DIPIA, Via Fontana Candida 1, 00040, Monteporzio Catone (RM), Italy. Application of Neural Net Models to classify and to forecast the observed precipitation type at the ground using the Artificial Intelligence Competition data set. A. Pelliccioni (*), R. Cotroneo (*), F.

More information

PATTERN RECOGNITION FOR PARTIAL DISCHARGE DIAGNOSIS OF POWER TRANSFORMER

PATTERN RECOGNITION FOR PARTIAL DISCHARGE DIAGNOSIS OF POWER TRANSFORMER PATTERN RECOGNITION FOR PARTIAL DISCHARGE DIAGNOSIS OF POWER TRANSFORMER PO-HUNG CHEN 1, HUNG-CHENG CHEN 2, AN LIU 3, LI-MING CHEN 1 1 Department of Electrical Engineering, St. John s University, Taipei,

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

A Hybrid Model of Wavelet and Neural Network for Short Term Load Forecasting

A Hybrid Model of Wavelet and Neural Network for Short Term Load Forecasting International Journal of Electronic and Electrical Engineering. ISSN 0974-2174, Volume 7, Number 4 (2014), pp. 387-394 International Research Publication House http://www.irphouse.com A Hybrid Model of

More information

NEURAL NETWORKS APPLICATION FOR MECHANICAL PARAMETERS IDENTIFICATION OF ASYNCHRONOUS MOTOR

NEURAL NETWORKS APPLICATION FOR MECHANICAL PARAMETERS IDENTIFICATION OF ASYNCHRONOUS MOTOR NEURAL NETWORKS APPLICATION FOR MECHANICAL PARAMETERS IDENTIFICATION OF ASYNCHRONOUS MOTOR D. Balara, J. Timko, J. Žilková, M. Lešo Abstract: A method for identification of mechanical parameters of an

More information

STRUCTURED NEURAL NETWORK FOR NONLINEAR DYNAMIC SYSTEMS MODELING

STRUCTURED NEURAL NETWORK FOR NONLINEAR DYNAMIC SYSTEMS MODELING STRUCTURED NEURAL NETWORK FOR NONLINEAR DYNAIC SYSTES ODELING J. CODINA, R. VILLÀ and J.. FUERTES UPC-Facultat d Informàtica de Barcelona, Department of Automatic Control and Computer Engineeering, Pau

More information

Applying the Wavelet Transform to Derive Sea Surface Elevation from Acceleration Signals

Applying the Wavelet Transform to Derive Sea Surface Elevation from Acceleration Signals Applying the Wavelet Transform to Derive Sea Surface Elevation from Acceleration Signals *Laurence Zsu-Hsin Chuang **Li-Chung Wu **Ching-Ruei Lin **Chia Chuen Kao *Institute of Ocean Technology and Marine

More information

Notes on Latent Semantic Analysis

Notes on Latent Semantic Analysis Notes on Latent Semantic Analysis Costas Boulis 1 Introduction One of the most fundamental problems of information retrieval (IR) is to find all documents (and nothing but those) that are semantically

More information

Deep Neural Networks (1) Hidden layers; Back-propagation

Deep Neural Networks (1) Hidden layers; Back-propagation Deep Neural Networs (1) Hidden layers; Bac-propagation Steve Renals Machine Learning Practical MLP Lecture 3 2 October 2018 http://www.inf.ed.ac.u/teaching/courses/mlp/ MLP Lecture 3 / 2 October 2018 Deep

More information

Vibration Signals Analysis and Condition Monitoring of Centrifugal Pump

Vibration Signals Analysis and Condition Monitoring of Centrifugal Pump Technical Journal of Engineering and Applied Sciences Available online at www.tjeas.com 2013 TJEAS Journal-2013-3-12/1081-1085 ISSN 2051-0853 2013 TJEAS Vibration Signals Analysis and Condition Monitoring

More information

Generator Thermal Sensitivity Analysis with Support Vector Regression

Generator Thermal Sensitivity Analysis with Support Vector Regression 2 American Control Conference Marriott Waterfront, Baltimore, MD, USA June 3-July 2, 2 WeB5.3 Generator Thermal Sensitivity Analysis with Support Vector Regression Youliang Yang and Qing Zhao* Abstract

More information

Assessment of the Frequency Domain Decomposition Method: Comparison of Operational and Classical Modal Analysis Results

Assessment of the Frequency Domain Decomposition Method: Comparison of Operational and Classical Modal Analysis Results Assessment of the Frequency Domain Decomposition Method: Comparison of Operational and Classical Modal Analysis Results Ales KUYUMCUOGLU Arceli A. S., Research & Development Center, Istanbul, Turey Prof.

More information

Gear Health Monitoring and Prognosis

Gear Health Monitoring and Prognosis Gear Health Monitoring and Prognosis Matej Gas perin, Pavle Bos koski, -Dani Juiric ic Department of Systems and Control Joz ef Stefan Institute Ljubljana, Slovenia matej.gasperin@ijs.si Abstract Many

More information

CHAPTER 6 FAULT DIAGNOSIS OF UNBALANCED CNC MACHINE SPINDLE USING VIBRATION SIGNATURES-A CASE STUDY

CHAPTER 6 FAULT DIAGNOSIS OF UNBALANCED CNC MACHINE SPINDLE USING VIBRATION SIGNATURES-A CASE STUDY 81 CHAPTER 6 FAULT DIAGNOSIS OF UNBALANCED CNC MACHINE SPINDLE USING VIBRATION SIGNATURES-A CASE STUDY 6.1 INTRODUCTION For obtaining products of good quality in the manufacturing industry, it is absolutely

More information

Slide04 Haykin Chapter 4: Multi-Layer Perceptrons

Slide04 Haykin Chapter 4: Multi-Layer Perceptrons Introduction Slide4 Hayin Chapter 4: Multi-Layer Perceptrons CPSC 636-6 Instructor: Yoonsuc Choe Spring 28 Networs typically consisting of input, hidden, and output layers. Commonly referred to as Multilayer

More information

Introduction to Natural Computation. Lecture 9. Multilayer Perceptrons and Backpropagation. Peter Lewis

Introduction to Natural Computation. Lecture 9. Multilayer Perceptrons and Backpropagation. Peter Lewis Introduction to Natural Computation Lecture 9 Multilayer Perceptrons and Backpropagation Peter Lewis 1 / 25 Overview of the Lecture Why multilayer perceptrons? Some applications of multilayer perceptrons.

More information

Research Article NEURAL NETWORK TECHNIQUE IN DATA MINING FOR PREDICTION OF EARTH QUAKE K.Karthikeyan, Sayantani Basu

Research Article   NEURAL NETWORK TECHNIQUE IN DATA MINING FOR PREDICTION OF EARTH QUAKE K.Karthikeyan, Sayantani Basu ISSN: 0975-766X CODEN: IJPTFI Available Online through Research Article www.ijptonline.com NEURAL NETWORK TECHNIQUE IN DATA MINING FOR PREDICTION OF EARTH QUAKE K.Karthikeyan, Sayantani Basu 1 Associate

More information

ABSTRACT I. INTRODUCTION

ABSTRACT I. INTRODUCTION 2017 IJSRST Volume 3 Issue 2 Print ISSN: 2395-6011 Online ISSN: 2395-602X National Conference on Advances in Engineering and Applied Science (NCAEAS) 16 th February 2017 In association with International

More information

Bearing fault diagnosis based on TEO and SVM

Bearing fault diagnosis based on TEO and SVM Bearing fault diagnosis based on TEO and SVM Qingzhu Liu, Yujie Cheng 2 School of Reliability and Systems Engineering, Beihang University, Beijing 9, China Science and Technology on Reliability and Environmental

More information

ARTIFICIAL NEURAL NETWORKS گروه مطالعاتي 17 بهار 92

ARTIFICIAL NEURAL NETWORKS گروه مطالعاتي 17 بهار 92 ARTIFICIAL NEURAL NETWORKS گروه مطالعاتي 17 بهار 92 BIOLOGICAL INSPIRATIONS Some numbers The human brain contains about 10 billion nerve cells (neurons) Each neuron is connected to the others through 10000

More information

Numerical magnetic field analysis and signal processing for fault diagnostics of electrical machines

Numerical magnetic field analysis and signal processing for fault diagnostics of electrical machines The Emerald Research Register for this journal is available at http://wwwemeraldinsightcom/researchregister The current issue and full text archive of this journal is available at http://wwwemeraldinsightcom/0332-1649htm

More information

A Fractal-ANN approach for quality control

A Fractal-ANN approach for quality control A Fractal-ANN approach for quality control Kesheng Wang Department of Production and Quality Engineering, University of Science and Technology, N-7491 Trondheim, Norway Abstract The main problem with modern

More information

A methodology for fault detection in rolling element bearings using singular spectrum analysis

A methodology for fault detection in rolling element bearings using singular spectrum analysis A methodology for fault detection in rolling element bearings using singular spectrum analysis Hussein Al Bugharbee,1, and Irina Trendafilova 2 1 Department of Mechanical engineering, the University of

More information

Condition monitoring of motor-operated valves in nuclear power plants

Condition monitoring of motor-operated valves in nuclear power plants Author manuscript, published in "8th International Conference on Condition Monitoring and Machinery Failure Prevention Technologies, United Kingdom (2011)" Condition monitoring of motor-operated valves

More information

Artificial Neural Networks Francesco DI MAIO, Ph.D., Politecnico di Milano Department of Energy - Nuclear Division IEEE - Italian Reliability Chapter

Artificial Neural Networks Francesco DI MAIO, Ph.D., Politecnico di Milano Department of Energy - Nuclear Division IEEE - Italian Reliability Chapter Artificial Neural Networks Francesco DI MAIO, Ph.D., Politecnico di Milano Department of Energy - Nuclear Division IEEE - Italian Reliability Chapter (Chair) STF - China Fellow francesco.dimaio@polimi.it

More information

Using Operating Deflection Shapes to Detect Misalignment in Rotating Equipment

Using Operating Deflection Shapes to Detect Misalignment in Rotating Equipment Using Operating Deflection Shapes to Detect Misalignment in Rotating Equipment Surendra N. Ganeriwala (Suri) & Zhuang Li Mark H. Richardson Spectra Quest, Inc Vibrant Technology, Inc 8205 Hermitage Road

More information

A New Model Reference Adaptive Formulation to Estimate Stator Resistance in Field Oriented Induction Motor Drive

A New Model Reference Adaptive Formulation to Estimate Stator Resistance in Field Oriented Induction Motor Drive A New Model Reference Adaptive Formulation to Estimate Stator Resistance in Field Oriented Induction Motor Drive Saptarshi Basak 1, Chandan Chakraborty 1, Senior Member IEEE and Yoichi Hori 2, Fellow IEEE

More information

An ANN based Rotor Flux Estimator for Vector Controlled Induction Motor Drive

An ANN based Rotor Flux Estimator for Vector Controlled Induction Motor Drive International Journal of Electrical Engineering. ISSN 974-58 Volume 5, Number 4 (), pp. 47-46 International Research Publication House http://www.irphouse.com An based Rotor Flux Estimator for Vector Controlled

More information

Electric Load Forecasting Using Wavelet Transform and Extreme Learning Machine

Electric Load Forecasting Using Wavelet Transform and Extreme Learning Machine Electric Load Forecasting Using Wavelet Transform and Extreme Learning Machine Song Li 1, Peng Wang 1 and Lalit Goel 1 1 School of Electrical and Electronic Engineering Nanyang Technological University

More information

OPTIMIZATION OF MORLET WAVELET FOR MECHANICAL FAULT DIAGNOSIS

OPTIMIZATION OF MORLET WAVELET FOR MECHANICAL FAULT DIAGNOSIS Twelfth International Congress on Sound and Vibration OPTIMIZATION OF MORLET WAVELET FOR MECHANICAL FAULT DIAGNOSIS Jiří Vass* and Cristina Cristalli** * Dept. of Circuit Theory, Faculty of Electrical

More information

Instituto Tecnológico y de Estudios Superiores de Occidente Departamento de Electrónica, Sistemas e Informática. Introductory Notes on Neural Networks

Instituto Tecnológico y de Estudios Superiores de Occidente Departamento de Electrónica, Sistemas e Informática. Introductory Notes on Neural Networks Introductory Notes on Neural Networs Dr. José Ernesto Rayas Sánche April Introductory Notes on Neural Networs Dr. José Ernesto Rayas Sánche BIOLOGICAL NEURAL NETWORKS The brain can be seen as a highly

More information

Offline Parameter Identification of an Induction Machine Supplied by Impressed Stator Voltages

Offline Parameter Identification of an Induction Machine Supplied by Impressed Stator Voltages POSTER 2016, PRAGUE MAY 24 1 Offline Parameter Identification of an Induction Machine Supplied by Impressed Stator Voltages Tomáš KOŠŤÁL Dept. of Electric Drives and Traction, Czech Technical University,

More information

EMD-BASED STOCHASTIC SUBSPACE IDENTIFICATION OF CIVIL ENGINEERING STRUCTURES UNDER OPERATIONAL CONDITIONS

EMD-BASED STOCHASTIC SUBSPACE IDENTIFICATION OF CIVIL ENGINEERING STRUCTURES UNDER OPERATIONAL CONDITIONS EMD-BASED STOCHASTIC SUBSPACE IDENTIFICATION OF CIVIL ENGINEERING STRUCTURES UNDER OPERATIONAL CONDITIONS Wei-Xin Ren, Department of Civil Engineering, Fuzhou University, P. R. China Dan-Jiang Yu Department

More information

ARTIFICIAL NEURAL NETWORKS APPROACH IN MICROWAVE FILTER TUNING

ARTIFICIAL NEURAL NETWORKS APPROACH IN MICROWAVE FILTER TUNING Progress In Electromagnetics Research M, Vol. 13, 173 188, 2010 ARTIFICIAL NEURAL NETWORKS APPROACH IN MICROWAVE FILTER TUNING J. J. Michalski TeleMobile Electronics Ltd. Pomeranian Science and Technology

More information

4. Multilayer Perceptrons

4. Multilayer Perceptrons 4. Multilayer Perceptrons This is a supervised error-correction learning algorithm. 1 4.1 Introduction A multilayer feedforward network consists of an input layer, one or more hidden layers, and an output

More information

Introduction to Neural Networks

Introduction to Neural Networks Introduction to Neural Networks What are (Artificial) Neural Networks? Models of the brain and nervous system Highly parallel Process information much more like the brain than a serial computer Learning

More information

Artificial Neural Network and Fuzzy Logic

Artificial Neural Network and Fuzzy Logic Artificial Neural Network and Fuzzy Logic 1 Syllabus 2 Syllabus 3 Books 1. Artificial Neural Networks by B. Yagnanarayan, PHI - (Cover Topologies part of unit 1 and All part of Unit 2) 2. Neural Networks

More information

CS:4420 Artificial Intelligence

CS:4420 Artificial Intelligence CS:4420 Artificial Intelligence Spring 2018 Neural Networks Cesare Tinelli The University of Iowa Copyright 2004 18, Cesare Tinelli and Stuart Russell a a These notes were originally developed by Stuart

More information

Jean Morlet and the Continuous Wavelet Transform (CWT)

Jean Morlet and the Continuous Wavelet Transform (CWT) Jean Morlet and the Continuous Wavelet Transform (CWT) Brian Russell 1 and Jiajun Han 1 CREWES Adjunct Professor CGG GeoSoftware Calgary Alberta. www.crewes.org Introduction In 198 Jean Morlet a geophysicist

More information

Neural network modelling of reinforced concrete beam shear capacity

Neural network modelling of reinforced concrete beam shear capacity icccbe 2010 Nottingham University Press Proceedings of the International Conference on Computing in Civil and Building Engineering W Tizani (Editor) Neural network modelling of reinforced concrete beam

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

Speaker Representation and Verification Part II. by Vasileios Vasilakakis

Speaker Representation and Verification Part II. by Vasileios Vasilakakis Speaker Representation and Verification Part II by Vasileios Vasilakakis Outline -Approaches of Neural Networks in Speaker/Speech Recognition -Feed-Forward Neural Networks -Training with Back-propagation

More information

Intelligent Handwritten Digit Recognition using Artificial Neural Network

Intelligent Handwritten Digit Recognition using Artificial Neural Network RESEARCH ARTICLE OPEN ACCESS Intelligent Handwritten Digit Recognition using Artificial Neural Networ Saeed AL-Mansoori Applications Development and Analysis Center (ADAC), Mohammed Bin Rashid Space Center

More information

Selection of the Appropriate Lag Structure of Foreign Exchange Rates Forecasting Based on Autocorrelation Coefficient

Selection of the Appropriate Lag Structure of Foreign Exchange Rates Forecasting Based on Autocorrelation Coefficient Selection of the Appropriate Lag Structure of Foreign Exchange Rates Forecasting Based on Autocorrelation Coefficient Wei Huang 1,2, Shouyang Wang 2, Hui Zhang 3,4, and Renbin Xiao 1 1 School of Management,

More information

A Wavelet Neural Network Forecasting Model Based On ARIMA

A Wavelet Neural Network Forecasting Model Based On ARIMA A Wavelet Neural Network Forecasting Model Based On ARIMA Wang Bin*, Hao Wen-ning, Chen Gang, He Deng-chao, Feng Bo PLA University of Science &Technology Nanjing 210007, China e-mail:lgdwangbin@163.com

More information

Fault Detection and Diagnosis of an Electrohydrostatic Actuator Using a Novel Interacting Multiple Model Approach

Fault Detection and Diagnosis of an Electrohydrostatic Actuator Using a Novel Interacting Multiple Model Approach 2011 American Control Conference on O'Farrell Street, San Francisco, CA, USA June 29 - July 01, 2011 Fault Detection and Diagnosis of an Electrohydrostatic Actuator Using a Novel Interacting Multiple Model

More information