Showing papers on "Hybrid neural network published in 2000"

A stator flux oriented vector-controlled induction motor drive with space vector PWM and flux vector synthesis by neural networks

Abstract: A stator flux oriented vector-controlled induction motor drive is described where the space vector PWM (SVM) and stator flux vector estimation are implemented by artificial neural networks (ANN). Artificial neural networks, when implemented by dedicated hardware ASIC chips, provide extreme simplification and fast execution for control and feedback signal processing functions in high performance AC drives. In the proposed project, a feedforward ANN-based SVM, operating at 20 kHz sampling frequency, generates symmetrical PWM pulses in both undermodulation and overmodulation regions covering the range from DC (zero frequency) up to square-wave mode at 60 Hz. In addition, a programmable cascaded low-pass filter (PCLPF), that permits de offset-free stator flux vector synthesis at very low frequency using the voltage model, has been implemented by a hybrid neural network which consists of a recurrent neural network (RNN) and a feedforward neural network (FFANN). The RNN-FFANN based flux estimation is simple, permits faster implementation, and gives superior transient performance when compared with a standard DSP-based PCLPF. A 5-hp open loop volts/Hz controlled drive incorporating the proposed ANN-based SVM and RNN-FFANN based flux estimator was initially evaluated in the frequency range of 1.0 Hz to 58 Hz to validate the performances of SVM and flux estimator. Next, the complete 5 hp drive with stator flux oriented vector control was evaluated extensively using the PWM modulator and flux estimator. The drive performances in both volts/Hz control and vector control were found to be excellent.

Virtual-Intelligence Applications in Petroleum Engineering: Part 2—Evolutionary Computing

Abstract: The goal of this second article is to provide an overview of evolutionary computing, its potential combination with neural networks to produce powerful intelligent applications, and its applications in the oil and gas industry. The most succesful intelligent applications incorporate several virtual-intelligence tools in a hybrid manner. Virtual-intelligence tools complement each other and are able to amplify each other's effectiveness. This article also presents the background of evolutionary computation as related to Darwinian evolution theory. This is followed by a more detailed look at genetic algorithms, the primary evolutionary-computing paradigm currently used. The article concludes by exploring application of a hybrid neural network/genetic algorithm system to a petroleum-engineering-related problem.

Improving learning accuracy of fuzzy decision trees by hybrid neural networks

Abstract: Although the induction of fuzzy decision tree (FDT) has been a very popular learning methodology due to its advantage of comprehensibility, it is often criticized to result in poor learning accuracy. Thus, one fundamental problem is how to improve the learning accuracy while the comprehensibility is kept. This paper focuses on this problem and proposes using a hybrid neural network (HNN) to refine the FDT. This HNN, designed according to the generated FDT and trained by an algorithm derived in this paper, results in a FDT with parameters, called weighted FDT. The weighted FDT is equivalent to a set of fuzzy production rules with local weights (LW) and global weights (GW) introduced in our previous work (1998). Moreover, the weighted FDT, in which the reasoning mechanism incorporates the trained LW and GW, significantly improves the FDTs' learning accuracy while keeping the FDT comprehensibility. The improvements are verified on several selected databases. Furthermore, a brief comparison of our method with two benchmark learning algorithms, namely, fuzzy ID3 and traditional backpropagation, is made. The synergy between FDT induction and HNN training offers new insight into the construction of hybrid intelligent systems with higher learning accuracy.

Hybrid neural-network and rule-based expert system for automatic sleep stage scoring

Abstract: In order to increase the performance of automatic sleep stage scoring, we propose a hybrid neural-network and expert system taking advantages of each system. After signal cleaning and feature extraction from polysomnographic signals using several algorithms we suggested, the rule-based expert system classified the sleep states with symbolic reasoning. The neural network supplemented the shortcomings of rule-based system by dealing with exceptions of rules. The result shows that the combination of computational and symbolic intelligence is promising approach to automatic sleep signal analysis.

A hybrid feedforward neural network model for the cephalosporin C production process

Abstract: At present, direct on-line measurements of key bioprocess variables as biomass, substrate and product concentrations is a difficult task. Many of the available hardware sensors are either expensive or lack reliability and robustness. To overcome this problem, indirect estimation techniques have been studied during the last decade. Inference algorithms rely either on phenomenological or on empirical models. Recently, hybrid models that combine these two approaches have received great attention. In this work, a hybrid neural network algorithm was applied to a fermentative process. Mass balance equations were coupled to a feedforward neural network (FNN). The FNN was used to estimate cellular growth and product formation rates, which are inserted into the mass balance equations. On-line data of cephalosporin C fed-batch fermentation were used. The measured variables employed by the inference algorithm were the contents of CO2 and O2 in the effluent gas. The fairly good results obtained encourage further studies to use this approach in the development of process control algorithms.

Hybrid neural networks for color identification

Abstract: A method uses a hybrid neural network including a self organizing mapping neural network (SOM NN) and a, back-propagation neural network (BP NN) for color identification. In the method the red, green and blue (RGB) of color samples are input as features of training samples and are automatically classified by way of SOM NN. Afterwards, the outcomes of SOM NN are respectively delivered to various BP NN for further learning; and the map relationship of the input and the output defines the X,Y, Z corresponding the x, y and z values of a coordinate system of the standard color samples of RGB and IT8. By way of the above learning structure, a non-linear model of color identification can be set up. After color samples are self organized and classified by SOM NN network, data can be categorized in clusters as a result of characteristic difference thereof. Then the data are respectively sent to BP NN for learning whereby-the learning system not only can be quickly converged but also lower error discrepancy in operation effectively.

A hybrid structured neural network receiver in digital communication systems

Abstract: In order to reduce the complexity of a radial basis function (RBF) network as a multiuser demodulator and an equalizer, we propose a simplified hybrid neural network architecture. The proposed neural network, which is called RN, has the structure of combining a radial basis function network with multilayer perceptrons (MLPs). The RBF network yields the linear combining output of the hidden layer while the proposed hybrid neural network produces the output using nonlinear combining techniques. From computer simulation results, the RN with the reduced structure from about 50% to about 70% over the RBF network shows better than or almost equal performance to the RBF network as a multiuser demodulator and an equalizer.

Fuzzy self-organizing hybrid neural network for gas analysis system

Abstract: The paper presents the gas analysis system applying the self-organizing fuzzy hybrid neural network. The network is composed of the self-organizing competitive fuzzy layer and the supervised multilayer perceptron (MLP) subnetwork, connected in cascade. The characteristic features of this network structure for gas analysis systems are discussed and the results of experiments compared to standard neural solutions based on MLP or classical hybrid network employing the Kohonen layer.

Hybrid neural network approach in description and prediction of dynamic behavior of chaotic chemical reaction systems

Abstract: A chaotic system with available prior knowledge is identified with both the sequential hybrid neural network and the standard Artificial Neural Network (ANN). The identified models are validated with phase portrait, return map, the largest Lyapunov exponent and correlation dimension instead of using Sum of Square Errors (SSE). Interpolation and Extrapolation capability of the models are compared. This is demonstrated for nonisothermal, irreversible, first-order, series reaction A≇B≇C in a CSTR.

A novel hybrid classifier for recognition of handwritten numerals

Abstract: A hybrid neural network and tree classification system for handwritten numeral recognition is proposed. The recognition system consists of coarse and fine classification based on a variety of stable and reliable global features and local features. For the coarse classifier: a four-layer feedforward neural network with backpropagation learning algorithm is employed to distinguish six subsets {0}, {6}, {8}, {1,7}, {4,9}, {2,3,5} based on the similarity of character's geometrical features. Three character classes {0}, {6} and {8} are directly recognized from ANN. For each of the last three subsets, a decision tree classifier is built for fine classification as follows: firstly, the specific feature-class relationship is heuristically and empirically created between the feature primitives and corresponding semantic class. Then, an iterative growing and pruning algorithm is used to form a tree classifier. Experiments demonstrated that the proposed hybrid recognition system is robust and flexible, which can achieve a high recognition rate.

Design of neural network-based microchip for color segmentation

Abstract: This paper presents an overview of current ongoing research and design efforts conducted by Intelligent Optical Systems, Inc. in the area of hardware-based color segmentation. We discuss the specifics of the design of a microchip that combines a hardwired hybrid neural network with on-chip imaging. Several preliminary tests show high approximation ability of our scheme. The single-chip implement has many advantages. The final product will consists of an RGB pixel array with infinite color depth and a neural network capable of high speed image segmentation.© (2000) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

A semi-parametric hybrid neural model for nonlinear blind signal separation.

Abstract: Nonlinear blind signal separation is an important but rather difficult problem. Any general nonlinear independent component analysis algorithm for such a problem should specify which solution it tries to find. Several recent neural networks for separating the post nonlinear blind mixtures are limited to the diagonal nonlinearity, where there is no cross-channel nonlinearity. In this paper, a new semi-parametric hybrid neural network is proposed to separate the post nonlinearly mixed blind signals where cross-channel disturbance is included. This hybrid network consists of two cascading modules, which are a neural nonlinear module for approximating the post nonlinearity and a linear module for separating the predicted linear blind mixtures. The nonlinear module is a semi-parametric expansion made up of two sub-networks, one of which is a linear model and the other of which is a three-layer perceptron. These two sub-networks together produce a "weak" nonlinear operator and can approach relatively strong nonlinearity by tuning parameters. A batch learning algorithm based on the entropy maximization and the gradient descent method is deduced. This model is successfully applied to a blind signal separation problem with two sources. Our simulation results indicate that this hybrid model can effectively approach the cross-channel post nonlinearity and achieve a good visual quality as well as a high signal-to-noise ratio in some cases.

Fault location in transmission line using self-organizing neural network

Abstract: The paper presents the application of self-organizing neural network for the location of the fault in transmission line and estimation of the parameter of the faulty element. The location of fault is done on the basis of the measurement of some node voltages of the line and appropriate preprocessing to enhance the differences between different faults. The hybrid neural network is used to solve the problem. The self-organizing layer of this network is used as the classifier. The output postprocessing MLP structure realizes the association of the place of fault and its parameter with the measured set of node voltages. The results of computer experiments are given in the paper and discussed.

A fuzzy-neural adaptive control for MIMO nonlinear system with application

Abstract: The fuzzy identification proposed by Takaki and Sugeno (1985) is extended to a MIMO adaptive controller based on a hybrid neural network structure. The network is roughly divided into the premise and consequence corresponding to the T-S model. Each parameter of the consequence function can be adjusted by the extended Bp algorithm so that automatic rule modification can be realized. The membership function of each fuzzy subset can be modified by a genetic algorithm. In this way, more pre-knowledge for the plant need not be required. Finally, the MIMO fuzzy-neural control is used to simulate a real example.

Adaptive Radio Resource Control via Cascaded Neural Networks for Sequenced Propagation Estimation and Multi-user Detection in Third-generation Wireless Networks

Abstract: A hybrid neural network approach is presented to predict radio propagation characteristics and multiuser interference and to evaluate their combined impact on throughput, latency and information loss in third-generation (3G) wireless networks. The three performance parameters influence the quality of service (QoS) for multimedia services for 3G networks. These networks are based on hierarchical cell structures and operate in mobile urban and indoor environments with service demands emanating from diverse traffic sources. Candidate radio interfaces for these networks employ a form of wideband CDMA.

A hybrid neural network model for predicting silicon content

Abstract: The silicon content of the hot metal in the blast furnace ironmaking process normally reflects the thermal state of the furnace and affects the fuel rate. In this paper a hybrid neural network model is proposed to predict the silicon contentn steps ahead. A time-delay neural network, which has self-loops to represent dynamics, is adopted here. The learning procedure of this network has been divided into two states. A BP algorithm with forgetting factor is first introduced to find the appropriate structure of the network. The temporal difference (TD) method with forgetting factor is then used forn-step-ahead prediction. The results show that the method can perform satisfactoryn-step-ahead prediction and is suited for implementation.

Fast learning using multilayer perceptron networks with adaptive centroid layer

Abstract: A hybrid neural network architecture is investigated for classification purposes The proposed hybrid is based on the multilayer perceptron (MLP) network In addition to the usual hidden layers the first hidden layer is selected to be an adaptive centroid layer Each unit in this new layer incorporates a centroid vector that is located somewhere in the space spanned by the input variables The output of these units is the Euclidean distance between the centroid vector and the inputs The centroid layer has some resemblance to the hidden layer of the radial basis function (RBF) networks Therefore the proposed design can be regarded as a sort of hybrid of the MLP and RBF networks The presented benchmark experiments demonstrate that the proposed hybrid can provide significant advantages over standard MLPs in terms of fast and efficient learning, and compact network structure

Vector control of induction motor using a hybrid neural network/fuzzy logic model

Abstract: This paper presents the practical implementation scheme of a three-phase induction motor vector drive using a hybrid neural network/fuzzy logic control model. In vector control, the three phase currents are transformed into a single current vector rotating at synchronous speed. The transformation enabled a single control output to control current values in all the three phases thus greatly simplifying the control model. This paper incorporates a hybrid neural network/fuzzy logic model into the vector control model to make it more intelligent. The neural network/fuzzy logic model is used as a universal function approximator that transforms the user input command and the speed sensor feedback to /spl Delta/l/sub q/ (the change in q current quantities) and the machine synchronous speed.

Similarity match (SM) technique for the development of client barcode

Abstract: A hybrid neural network is proposed for speaker verification (SV). The basic idea in this system is the usage of vector quantization preprocessing as the feature extractor. The experiments were carried out using a neural network model (NNM) with frame labeling performed from a client codebook known as NNM-C. The work also examines how the neural network model with enhance features from the client barcode compares to NNM client codebook with linear time normalization (LTN). Improved performance for NNM (client barcode) with more inputs and proper alignment of the speech signals supports the hypothesis that a more detailed representation of the speech patterns proved helpful for the system. The flexibility of this system allows an equal error rate (EER) of 0.62% (speaker specific EER) on a single isolated digit and 1.9% (SI EER) on a sequence of 12 isolated digits.

Application of an Adaptive Hybrid Neural Network to Medical Diagnosis

Abstract: We have previously devised a hybrid neural network, based on the synergism of the Fuzzy ARTMAP and Probabilistic Neural Networks, for on-line pattern classification and probability estimation tasks. In this paper, we investigate the applicability of the hybrid network to medical diagnosis problems. In particular, the network was employed to predict and classify Myocardial Infarction patients into two categories (positive and negative cases) using a database of real records collected from a hospital. A number of experiments was conducted to evaluate the effects of several network parameters on its performance. The results are discussed and compared with those from the Fuzzy ARTMAP network.

Hybrid neural network and variable structure control for robot

Abstract: This paper proposes a neural-network and continuous sliding mode hybrid control method for robot manipulators, which has a unknown model. First, a feedforward neural network is used to learn the characteristics of the robot system (or specially its inverse dynamics) for accurate trajectory following and smooth torque control. Then, a saturated-function-based continuous sliding mode controller is used to guarantee the convergence of the tracking errors, reduces or even eliminates the chattering. Simulations of a two-link robot are given to illustrated the good transient performance and smooth control torque.