Elliptical Basis Function Research Articles

Optimization strategy for curing ultra-thick composite laminates based on multi-objective genetic algorithm

Thermal gradient and temperature overshoot during the curing process of ultra-thick carbon fibre composite have a major impact on its properties. We present a strategy to optimize the curing process based on the surrogate model and genetic algorithm. Firstly, a FE model based on heat transfer was developed for curing a 30 mm thick laminate and validated by experimental data. Then a multi-objective optimization strategy was developed by combining the optimal Latin hypercube sampling method, elliptical basis function (EBF) neural network model and non-dominated sorting genetic algorithm-II (NSGA-II). It is found that the optimized cure cycle can effectively reduce the undesirable maximum DoC (Degree of cure) gradient and the temperature gradient leading to improved mechanical properties for the 30 mm thick laminate.

Organic solar cells defects classification by using a new feature extraction algorithm and an EBNN with an innovative pruning algorithm

Physical defects reduce the organic solar cells (OSC) functioning. Throughout the OSC fabrication process, the defects can occur, for instance, by scratches or uneven morphologies. In general, bulk defects, interface defects, and interconnect defects can promote shunt and series resistance of the cell. It is crucial to properly detect and classify such defects and their amount in the structure. Correlating such defects with the performance of the cell is important both during the R&D stages to optimize processes, and for mass production stages where defects detection is an integral part of the production line. For the recognition of texture variations in the scanning electron microscope images caused by these defects is crucial the definition of a set of features for texture representation. Because the low-order Zernike moments can represent the whole shape of the image and the high-order Zernike moments can describe the detail. Then, in our case, the feature of the image can be represented by a small number of Zernike moments. In fact, the feature set extracted and described by the Zernike moments are not sensitive to the noises and are hardly redundant. So it possible concentrate the signal energy over a set of few vectors. Finally, for classification, an elliptical basis function neural networks was used. The results show effectiveness of the proposed methodology. In fact, we obtained correct classification of 89.3% over testing data set.

Optimization of fuzzy control energy management strategy for fuel cell vehicle power system using a multi‐islandgenetic algorithm

AbstractEnergy storage system can be used to increase the fuel economy of fuel cell system (FCS). In this study, a new method was introduced for optimizing the energy management strategy (EMS) for fuel cell vehicle (FCV) to reduce fuel consumption. The membership function (MF) in fuzzy control is subjective; thus, 12 design variables in the input‐output MFs were selected using sensitivity analysis, and elliptical basis function neural network method was used to establish a high‐precision approximate model of FCV. Multi‐island genetic algorithm was used to optimize the MFs. The effectiveness of the optimized fuzzy control EMS and the proposed optimization method were demonstrated in simulations of two EMSs under four driving cycles. The simulation results confirmed that the optimized fuzzy control EMS provided smoother and more stable output power from FCS reducing hydrogen consumption by 8.4%, 1.1%, 5.1%, and 7.6%, respectively, compared to the original fuzzy control EMS; and hydrogen saved by the optimized EMS provided extra range of 9.15, 1.10, 5.37, and 8.25 km per 100 km in the four driving cycles, respectively. The optimized EMS can reduce hydrogen consumption to increase fuel economy and extend the life span of the fuel cell.

Optimal isotropic, reusable truss lattice material with near-zero Poisson’s ratio

Cork is a natural amorphous material with near-zero Poisson’s ratio that is ubiquitously used for sealing glass bottles. It is an anisotropic, transversally isotropic, composite that can hardly be scaled down. Here, we propose a new class of isotropic and reusable cork-like metamaterial that is designed from a hybrid truss-lattice material to show an isotropic Poisson’s ratio close to zero. Optimization is conducted using a multi-objective genetic algorithm, assisted by an elliptical basis function neural network, and coupled with finite element simulations. The optimal micro-structured metamaterial, fabricated by two-photon lithography with a lattice constant of 300 μm, has an almost isotropic Poisson’s ratio smaller than 0.08 in all directions. It can recover 96.6% of its original shape after a compressional test exceeding 20% strain.

Speech emotion recognition using hybrid spectral-prosodic features of speech signal/glottal waveform, metaheuristic-based dimensionality reduction, and Gaussian elliptical basis function network classifier

In this paper, a hybrid system consisting of three stages of feature extraction, dimensionality reduction, and feature classification is proposed for speech emotion recognition (SER). At feature extraction stage, an informationally-rich spectral-prosodic hybrid feature vector comprised of perceptual-spectral features; that is, mel-frequency cepstral coefficient (MFCC), perceptual linear prediction coefficient (PLPC), and perceptual minimum variance distortionless response (PMVDR) coefficient along with the prosodic feature of pitch (i.e. F0) are extracted for each frame. This feature vector is extracted from both speech signal and its glottal-waveform. The first and the second-order derivatives are then added to the above-mentioned vector to form a high-dimensional hybrid feature vector characterized by a large number of dimensions. At the next stage, i.e. dimensionality reduction, the dimensionality of this feature vector is reduced using a new proposed quantum-behaved particle swarm optimization (QPSO)-based approach. In this paper, a new QPSO algorithm (so-called, pQPSO) is presented that makes use of a truncated Laplace distribution (TLD) to generate new particles and thus to produce solutions (i.e. particles) that are all within a valid range of a problem (contrary to the standard QPSO). The contraction-expansion (CE) factor of the proposed pQPSO is also selected adaptively. Using the proposed QPSO algorithm, an optimal discriminative dimensionality reduction matrix (i.e. projection matrix) is estimated with emotion classification accuracy as a class-discriminative criterion. At the subsequent stage, vectors with reduced feature dimensionality are fed into a Gaussian elliptical basis function (GEBF)-type neural network classifier to detect their speech emotion. To accelerate the training phase of the GEBF classifier, a fast-scaled conjugate gradient (SCG) algorithm is correspondingly employed that does not need to adjust the learning rate. Finally, the proposed method is evaluated on three standard emotional speech databases of Berlin Database of Emotional Speech (EMODB), Surrey Audio-Visual Expressed Emotion (SAVEE), and Interactive Emotional Dyadic Motion Capture (IEMOCAP). The experimental results showed that the proposed method was more accurate than state-of-the-art ones in terms of detecting speech emotions.

Aero-thermal behavior and performance optimization of rectangular finned elliptical heat exchangers with different tube arrangements

Tube arrangement largely determines the flow structure as well as aero-thermal performance in finned tube heat exchangers (FTHXs). Most of the previous studies did not cover both overall performance and local thermal flow characteristics due to the physical complexity. This work is related to the comparison of various aero-thermal behaviors of rectangular finned elliptical tube heat exchangers (RFETHXs) with different tube arrangements in a turbulent flow region, from basic characteristics to optimal structures. Using 3-D numerical simulation based on the validated standard k-ε model, the local thermal flow characteristic as well as overall dimensional and dimensionless performances of three arrangements is investigated. Then, using variance analysis, the contribution of each structural parameter of the three arrangements on overall performance is compared after the screening of the factor of longitudinal tube pitch. Using the same CFD data as variance analysis, an elliptical basis function (EBF) model of each type of arrangements is constructed to understand the distribution of j/f0.5 as well as PEC in the parameter range of this work. Moreover, optimal structural parameters of each arrangement are provided. It is observed that, each tube arrangement presents its own characteristic of flow separation and heat transfer coefficient distribution on fins. For the in-line and staggered row arrangements, the contributions of all the main factors and the two-way interactions on various overall performances are close, while the situation of the staggered column is different. The influence of two-way interactions is marginal in the staggered column structure, while it cannot be neglected in the two other arrangements in most cases. For j/f0.5, the optimal structural parameter combination of the in-line and staggered row arrangements is fin thickness 0.6 mm, fin pitch 2 mm and transverse tube pitch 54 mm, whereas that of the staggered column arrangement is 0.6 mm, 2 mm and 70 mm in turn.

Robust optimisation of vibro-acoustic system based on an elliptical basis function neural network

Reducing the sound pressure level (SPL) reliably and effectively is paramount in the presence of uncertainty. Herein, the robust optimisation of a vibro-acoustic system is performed by combining a probabilistic analysis approach and an approximate methodology. A surrogate model containing uncertain variables is established based on the elliptical basis function (EBF) neural network. Using Monte Carlo (MC) simulations, the probability distribution of the output responses are obtained, and the result shows that the sound pressure is influenced by uncertain factors. The simulated annealing (SA) algorithm is adopted to perform the robust optimisation. It shows that the optimisation results not only reduced the SPL inside the passenger car, but also enabled the SPL to be less sensitive to the fluctuations of the uncertain variables.

An oblique elliptical basis function network approach for supervised learning applications

We propose a neural network architecture based on the oblique elliptical basis function for supervised learning problems. In classification, a category can be a disconnected or non-convex region involving several overlapping or disjoint sub-regions of the feature space. Other existing supervised learning methods may have the restriction that only allows decision regions to be convex. Our proposed method overcomes this restriction by employing a rotational self-constructing clustering algorithm to decompose the feature space into a collection of sub-regions which can then be combined to make up individual categories. An unseen instance is classified to a certain category if its similarity to the category exceeds a threshold. The whole framework fits in a five-layer network consisting of input, component-similarity, cluster-similarity, aggregation, and output layers. A similar idea also applies to solving regression problems. A parameter learning algorithm based on least squares estimation is used to derive the weights of the underlying network. Our approach can offer some advantages in practicality. Through the incorporation of rotation, data can be clustered more appropriately than by standard elliptical basis functions. Also, our approach is applicable to single-label classification, multi-label classification, as well as regression problems. A number of experiments are conducted to show the effectiveness of the proposed approach.

Multidisciplinary Design Optimization of Crankshaft Structure Based on Cooptimization and Multi-Island Genetic Algorithm

The feasibility design method with multidisciplinary and multiobjective optimization is applied in the research of lightweight design and NVH performances of crankshaft in high-power marine reciprocating compressor. Opt-LHD is explored to obtain the experimental scheme and perform data sampling. The elliptical basis function neural network (EBFNN) model considering modal frequency, static strength, torsional vibration angular displacement, and lightweight design of crankshaft is built. Deterministic optimization and reliability optimization for lightweight design of crankshaft are operated separately. Multi-island genetic algorithm (MIGA) combined with multidisciplinary cooptimization method is used to carry out the multiobjective optimization of crankshaft structure. Pareto optimal set is obtained. Optimization results demonstrate that the reliability optimization which considers the uncertainties of production process can ensure product stability compared with deterministic optimization. The coupling and decoupling of structure mechanical properties, NVH, and lightweight design are considered during the multiobjective optimization of crankshaft structure. Designers can choose the optimization results according to their demands, which means the production development cycle and the costs can be significantly reduced.

Surrogating for High Dimensional Computationally Expensive Multi-Modal Functions with Elliptical Basis Function Models

In the engineering design process, approximation Technique could guarantee the fitting precision, speed up the design process and reduce design costs. To a certain extent, surrogate models could replace time-consuming and highly accurate computational fluid dynamics analysis gradually. In this paper, we take Optimal Latin Hypercube Sampling experimental design strategies to determine the sample space and error analysis test sample, adopt the principle of infilling criteria based on the maximum error to improve the accuracy of the surrogate model, test the unimodal and multimodal expensive functions of 10 dimension, 20 dimensions and 30 dimensions, study the performance and scope of EBF-NN surrogate model based on infilling criteria by comparing the RBF-NN surrogate model.

A novel approach for change detection of remotely sensed images using semi-supervised multiple classifier system

In this article, a novel approach using ensemble of semi-supervised classifiers is proposed for change detection in remotely sensed images. Unlike the other traditional methodologies for detection of changes in land-cover, the present work uses a multiple classifier system in semi-supervised (leaning) framework instead of using a single weak classifier. Iterative learning of base classifiers is continued using the selected unlabeled patterns along with a few labeled patterns. Ensemble agreement is utilized for choosing the unlabeled patterns for the next training step. Finally, each of the unlabeled patterns is assigned to a specific class by fusing the outcome of base classifiers using some combination rule. For the present investigation, multilayer perceptron (MLP), elliptical basis function neural network (EBFNN) and fuzzy k-nearest neighbor (k-nn) techniques are used as base classifiers. Experiments are carried out on multi-temporal and multi-spectral images and the results are compared with the change detection techniques using MLP, EBFNN, fuzzy k-nn, unsupervised modified self-organizing feature map and semi-supervised MLP. Results show that the proposed work has an edge over the other state-of-the-art techniques for change detection.

Remaining Useful Life Prediction Using Elliptical Basis Function Network and Markov Chain

This paper presents a novel method for remaining useful life prediction using the Elliptical Basis Function (EBF) network and a Markov chain. The EBF structure is trained by a modified Expectation-Maximization (EM) algorithm in order to take into account the missing covariate set. No explicit extrapolation is needed for internal covariates while a Markov chain is constructed to represent the evolution of external covariates in the study. The estimated external and the unknown internal covariates constitute an incomplete covariate set which are then used and analyzed by the EBF network to provide survival information of the asset. It is shown in the case study that the method slightly underestimates the remaining useful life of an asset which is a desirable result for early maintenance decision and resource planning.

A hybrid classification algorithm based on coevolutionary EBFNN and domain covering method

A new hybrid scheme of the elliptical basis function neural network (EBFNN) model combined with the cooperative coevolutionary algorithm (Co-CEA) and domain covering method is presented for multiclass classification tasks. This combination of the Co-CEA EBFNN (CC-EBFNN) and the domain covering method is proposed to enhance the predictive capability of the estimated model. The whole training process is divided into two stages: the evolutionary process, and the heuristic structure refining process. First, the initial hidden nodes of the EBFNN model are selected randomly in the training samples, which are further partitioned into modules of hidden nodes with respect to their class labels. Subpopulations are initialized on modules, and the Co-CEA evolves all subpopulations to find the optimal EBFNN structural parameters. Then the heuristic structure refining process is performed on the individual in the elite pool with the special designed constructing and pruning operators. Finally, the CC-EBFNN model is tested on six real-world classification problems from the UCI machine learning repository, and experimental results illustrate that the EBFNN model can be estimated in fewer evolutionary trials, and is able to produce higher prediction accuracies with much simpler network structures when compared with conventional learning algorithms.

Applying Elliptical Basis Function Neural Networks to VAD for Wireless Communication Systems

Voice activity detection (VAD) determines whether a slice of waveform is voice or silence. The proposed VAD algorithm applying Elliptical Basis Function (EBF) neural networks uses k-means clustering and least mean square for the update algorithm. The error rates achieved by the EBF network have superior performance to those of G.729 Annex B and RBF.

TEXT-INDEPENDENT SPEAKER VERIFICATION USING MINIMAL RESOURCE ALLOCATION NETWORKS

This paper presents a text-independent speaker verification system based on an online Radial Basis Function (RBF) network referred to as Minimal Resource Allocation Network (MRAN). MRAN is a sequential learning RBF, in which hidden neurons are added or removed as training progresses. LP-derived cepstral coefficients are used as feature vectors during training and verification phases. The performance of MRAN is compared with other well-known RBF and Elliptical Basis Function (EBF) based speaker verification methods in terms of error rates and computational complexity on a series of speaker verification experiments. The experiments use data from 258 speakers from the phonetically balancedcontinuous speech corpus TIMIT. The results show that MRAN produces comparable error rates to other methods with much less computational complexity.

An elliptical basis function network for classification of remote sensing images

An elliptical basis function (EBF) network is employed in this study for the classification of remotely sensed images. Though similar in structure, the EBF network differs from the well-known radial basis function (RBF) network by incorporating full covariance matrices and employing the expectation-maximization (EM) algorithm to estimate the basis functions. Since remotely sensed data often take on mixture-density distributions in the feature space, the network not only possesses the advantage of the RBF mechanism, but also utilizes the EM algorithm to compute the maximum likelihood estimates of the mean vectors and covariance matrices of a Gaussian mixture distribution in the training phase. Experimental results show that the EM-based EBF network is more effective in training and simpler in structure than an RBF network constructed for the same task.

Adaptive forecasting of irregular demand processes

The paper presents some comparative results from different approaches for solving the problem of forecasting irregularly spaced time series, including those representing irregular demand processes in inventory management or production planning problems. The best-known methods for that purpose are based on simple exponential smoothing of the time and magnitude of occurrence of demand. It is shown how supervised neural networks can be used as suitable filtering and forecasting models for many of those processes. In particular, we discuss the identification and estimation issues, as well as the limitations associated to the application of Gaussian radial or elliptical basis function networks.

Feature mapping and recuperation by using elliptical basis function networks for robust speaker verification

The performance of speaker verification systems is often compromised under real-world environments. For example, variations in handset characteristics could cause severe performance degradation. This paper presents a novel method to overcome this problem by using a non-linear handset mapper. Under this method, a mapper is constructed by training an elliptical basis function network using distorted speech features as inputs and the corresponding clean features as the desired outputs. During feature recuperation, clean features are recovered by feeding the distorted features to the feature mapper. The recovered features are then presented to a speaker model as if they were derived from clean speech. Experimental evaluations based on 258 speakers of the TIMIT and NTIMIT corpuses suggest that the feature mappers improve the verification performance remarkably.

Estimation of elliptical basis function parameters by the EM algorithm with application to speaker verification

This paper proposes to incorporate full covariance matrices into the radial basis function (RBF) networks and to use the expectation-maximization (EM) algorithm to estimate the basis function parameters. The resulting networks, referred to as elliptical basis function (EBF) networks, are evaluated through a series of text-independent speaker verification experiments involving 258 speakers from a phonetically balanced, continuous speech corpus (TIMIT).We propose a verification procedure using RBF and EBF networks as speaker models and show that the networks are readily applicable to verifying speakers using LP-derived cepstral coefficients as features. Experimental results show that small EBF networks with basis function parameters estimated by the EM algorithm outperform the large RBF networks trained in the conventional approach. The results also show that the equal error rate achieved by the EBF networks is about two-third of that achieved by the vetor quantization (VQ)-based speaker models.

Determining the Optimal Number of Clusters by an Extended RPCL Algorithm

Determining an appropriate number of clusters is a difficult yet important problem that the rival penalized competitive learning (RPCL) algorithm was designed to solve, but its performance is not satifactory with overlapping clusters or cases where input vectors contain dependent components. We address this problem by incorporating full covariance matrices into the original RPCL algorithm. The resulting extended RPCL algorithm progressively eliminates units whose clusters contain only a small amount of training data. The algorithm is used to determine the number of clusters in a Gaussian distribution. It is also used to optimize the architecture of elliptical basis function networks for speaker verification and vowel classification. We found that covariance matrices obtained by the extended RPCL algorithm have a better representation of clusters than those obtained by the original RPCL algorithm, resulting in a lower verification error rate in speaker verification and a higher recognition accuracy in vowel classification.