Fault Diagnosis for Closed Loop Nonlinear System Using Generalized Frequency Response Functions and Least Square Support Vector Machine

Aimed at the non-stationary characteristics of rolling bearing vibration signal,a fault diagnosis method was proposed based on local-wave method and KPCA(kernel principal component analysis)-LSSVM(least squares support vector machine).Firstly,local wave decomposition was used to decompose rolling bearing vibration signal into several intrinsic mode function(IMF),whose feature energy,singular values and AR model parameters were computed as initial feature vectors.Secondly,ini tial feature vectors were mapped into a higher-dimensional space with KPCA,and the kemel principal components were extracted as new feature vectors,which used as the input of LSSVM for fault classification.The experimental results show the KPCA-LSSVM method improves LSSVM's classification performance by KPCA obtaining additional discriminative information,and has better generalization than direct LSSVM method,and can identify rolling bearing fault patterns more accurately.

This paper presents a novel Gabor-based kernel Principal Component Analysis (PCA) method by integrating the Gabor wavelet representation of face images and the kernel PCA method for face recognition. Gabor wavelets first derive desirable facial features characterized by spatial frequency, spatial locality, and orientation selectivity to cope with the variations due to illumination and facial expression changes. The kernel PCA method is then extended to include fractional power polynomial models for enhanced face recognition performance. A fractional power polynomial, however, does not necessarily define a kernel function, as it might not define a positive semidefinite Gram matrix. Note that the sigmoid kernels, one of the three classes of widely used kernel functions (polynomial kernels, Gaussian kernels, and sigmoid kernels), do not actually define a positive semidefinite Gram matrix either. Nevertheless, the sigmoid kernels have been successfully used in practice, such as in building support vector machines. In order to derive real kernel PCA features, we apply only those kernel PCA eigenvectors that are associated with positive eigenvalues. The feasibility of the Gabor-based kernel PCA method with fractional power polynomial models has been successfully tested on both frontal and pose-angled face recognition, using two data sets from the FERET database and the CMU PIE database, respectively. The FERET data set contains 600 frontal face images of 200 subjects, while the PIE data set consists of 680 images across five poses (left and right profiles, left and right half profiles, and frontal view) with two different facial expressions (neutral and smiling) of 68 subjects. The effectiveness of the Gabor-based kernel PCA method with fractional power polynomial models is shown in terms of both absolute performance indices and comparative performance against the PCA method, the kernel PCA method with polynomial kernels, the kernel PCA method with fractional power polynomial models, the Gabor wavelet-based PCA method, and the Gabor wavelet-based kernel PCA method with polynomial kernels.

Least squares support vector machine (LS-SVM) is a modified version of traditional support vector machine (SVM). LS-SVM considers equality constraints, therefore it solves a set of linear equations instead of quadratic programming problem in SVM. However, the sparseness of LS-SVM is lost due to itpsilas isin-sensitive cost function. Sparseness can be obtained by applying a pruning method, which eliminates some vectors with smallest support values and retrains the remaining samples. But iterative retraining is a time-consuming process. Motivated by the fact that boundary samples are more significant for constructing a LS-SVM classifier, this paper proposes a method of using pulse coupled neural networks (PCNNs) to search boundary samples of original data sets. The original data sets are mapped into some PCNN neurons, and a firing algorithm is designed to determine which samples lie at boundary region. It gives a novel approach to impose sparsity for LS-SVM. Experiments show that the proposed method can effectively detect boundary samples and speed up LS-SVM classifiers.

Gearbox affect the normal operation of the wind turbines, to study the fault diagnosis, support vector method was used. Parameters selection is very important and decides the fault diagnosis precision. In order to overcome the blindness of man-made choice of the parameters in least squares support vector machine (LSSVM) and improve the accuracy and efficiency of fault diagnosis, a method based on LSSVM trained by genetic algorithm was proposed. This method searches the optimized parameters in LSSVM by taking advantage of the genetic algorithms powerful global searching ability. The research is provided using this method on the fault diagnosis of wind turbine gearbox and compared with the diagnostic method of LSSVM. The experimental results show that the method achieves a higher diagnostic accuracy.

This study presents a multi-class least square support vector machines (LS-SVM)-based classifier for transformer fault diagnosis. First, the original binary classifier is extended for multi-class classification that is common in fault diagnosis by using combination schemes, that is, the minimal output coding, error correcting output codes, one-against-one and one-against-all schemes. Second, the algorithm of particle swarm optimisation is implemented to select the optimal feature parameters for the multi-class LS-SVM classifiers. Then the effectiveness of the proposed approach is verified on the basis of the experiments on benchmark classification data and real-world transformer data. For comparison purpose, three widely used transformer diagnosis methods such as the IEC criteria, back propagation neural network and standard support vector machines are utilised. The results show the proposed approach has a better performance both in training and testing accuracies.

A vibrating screen efficiency prediction modeling method based on autoregressive (AR) model and least square support vector machine (LS-SVM) was proposed. The vibration signals of a self-synchronized vibrating screen were collected to establish the AR model. Nonlinear principal components of the signals were extracted by the kernel principal component analysis (KPCA), followed by the regression model reconstruction using LS-SVM to accomplish reduced complexity of the prediction model from AR coefficients and improved generalization capacity and learning speed. The results show that the model predictions are consistent with the experimental data, which indicates that the modeling method is applicable and feasible in adjusting the design and process parameters of vibrating screens. Furthermore, the work condition monitoring method in the experiment is feasible for faults diagnosis of mechanical equipment.

An LS-SVM classifier based methodology for avoiding unwanted responses in processes under uncertainties

Utilization of SVM, LSSVM and GP for Predicting the Medical Waste Generation

The prediction of wastes generated in the hospital will help their management for several activities like storage, transport and disposing. This chapter adopts Support Vector Machine (SVM), Least Square Support Vector Machine (LSSVM) and Genetic Programming (GP) in order to estimate the rate of medical waste generation. In the event of predicting the rate, type of hospital, capacity and bed occupancy has been used as inputs of SVM, LSSVM and GP. SVM is based on statistical learning theory, which provides an elegant tool for nonlinear system modeling. LSSVM is the re-formulation to the general SVM. GP, a best part of evolutionary algorithm and also the specification of Genetic Algorithm (GA). These SVM, LSSVM and GP have been used as the regression techniques. The results show the performance of the developed SVM, LSSVM and GP models were elegant and outstanding.

Due to the growth prospect of analog circuit fault diagnosis, this paper tends to introduce a novel arithmetic model based on least squares support vector machine (LSSVM) and the semi-supervised learning (SSL) scheme which is adept at cost-saving. The proposed method contains two steps. Firstly, the fact that large deviation may emerge as a result of the empirical risk inspires the idea of an improved transductive least square support vector machine (T-LSSVM) which aims at obtaining the best hyperplane that equipped with the maximum margin to the support vectors no matter whether the samples are labeled or unlabeled. Secondly, to overcome the drawback of typical T-LSSVM, i.e., sensitivity to local minima, a laplcian-transductive least squares support vector machine (Lap-T-LSSVM) is proposed which can perform the fault diagnosis via a laplcian. The experiment adopts band-pass filter circuit as diagnosis object. Simulation results verify that the proposed method is superior to previous SVM in accuracy.

The objective of this paper was to determine soluble solids content (SSC) of intact Gannan navel orange by a portable near-infrared (NIR) spectrometer with the optical fiber in the wavelength range of 551~950nm. The effective wavelength regions (EWs) were chosen from the spectra pro-processed by second derivative by interval partial least square (iPLS) and backward interval partial least square (Bipls). Then the partial least square (PLS) and least square support vector machine (LS-SVM) models were developed with EWs. 60 unknown samples were used to evaluate the performance of them. The LS-SVM model was better than others with EWs chosen by Bilps. The correlation coefficient (R) and root mean square error of prediction (RMSEP) for LS-SVM (Bipls) were 0.86 and 0.55°Brix. The results showed that the portable NIR combination with LS-SVM was a feasible method to determine SSC of intact Gannan navel orange nondestructively.

The prediction of wastes generated in the hospital will help their management for several activities like storage, transport and disposing. This chapter adopts Support Vector Machine (SVM), Least Square Support Vector Machine (LSSVM) and Genetic Programming (GP) in order to estimate the rate of medical waste generation. In the event of predicting the rate, type of hospital, capacity and bed occupancy has been used as inputs of SVM, LSSVM and GP. SVM is based on statistical learning theory, which provides an elegant tool for nonlinear system modeling. LSSVM is the re-formulation to the general SVM. GP, a best part of evolutionary algorithm and also the specification of Genetic Algorithm (GA). These SVM, LSSVM and GP have been used as the regression techniques. The results show the performance of the developed SVM, LSSVM and GP models were elegant and outstanding.

The minimal complexity support vector machine is a fusion of the support vector machine (SVM) and the minimal complexity machine (MCM), and results in maximizing the minimum margin and minimizing the maximum margin. It works to improve the generalization ability of the L1 SVM (standard SVM) and LP (Linear Programming) SVM. In this paper, we discuss whether it also works for the LS (Least Squares) SVM. The minimal complexity LS SVM (MLS SVM) is trained by minimizing the sum of squared margin errors and minimizing the maximum margin. This results in solving a set of linear equations and a quadratic program, alternatingly. According to the computer experiments for two-class and multiclass problems, the MLS SVM does not outperform the LS SVM for the test data although it does for the cross-validation data.

This article adopts three soft computing techniques including support vector machine (SVM), least square support vector machine (LSSVM) and relevance vector machine (RVM) for prediction of status of epimetemorphic rock slope. The input variables of SVM, LSSVM and RVM are bulk density, height, inclination, cohesion and internal friction angle. There are 53 datasets which have been used to develop the SVM, LSSVM and RVM models. The developed SVM, LSSVM and RVM give equations for prediction of status of epimetemorphic rock slope. The performance of SVM, LSSVM and RVM is 100%. A comparative study has been presented between the developed SVM, LSSVM and RVM. The results confirm that the developed SVM, LSSVM and RVM are effective tools for prediction of status of epimetemorphic rock slope.

In the practical implementation of control of electromyography (sEMG) driven devices, algorithms should recognize the human’s motion from sEMG with fast speed and high accuracy. This study proposes two feature engineering (FE) techniques, namely, feature-vector resampling and time-lag techniques, to improve the accuracy and speed of least square support vector machine (LSSVM) for wrist palmar angle estimation from sEMG feature. The root mean square error and correlation coefficients of LSSVM with FE are 9.50 2.32 degree and 0.971 0.018 respectively. The average training time and average execution time of LSSVM with FE in processing 12600 sEMG points are 0.016 s and 0.053 s respectively. To evaluate the proposed algorithm, its estimation results are compared with those of three other methods, namely, LSSVM, radial basis function (RBF) neural network, and RBF with FE. Experimental results verify that introduction of time-lag into feature vector can greatly improve the estimation accuracy of both RBF and LSSVM; meanwhile the application of feature-vector resampling technique can significantly increase the training and execution speed of RBF neural network and LSSVM. Among different algorithms applied in this study, LSSVM with FE techniques performed best in terms of training and execution speed, as well as estimation accuracy.