• Journal of Institute of Control, Robotics and Systems
• /
• v.16 no.12
• /
• pp.1220-1225
• /
• 2010

This paper proposes a novel active noise control (ANC) approach that uses an IIR filter and neural network techniques to effectively reduce interior noise. We construct a multiple-channel IIR filter module which is a linearly augmented framework with a generic IIR model to generate a primary control signal. A three-layer perceptron neural network is employed for establishing a secondary-path model to represent air channels among noise fields. Since the IIR module and neural network are connected in series, the output of an IIR filter is transferred forward to the neural model to generate a final ANC signal. A gradient descent optimization based learning algorithm is analytically derived for the optimal selection of the ANC parameter vectors. Moreover, re-estimation of partial parameter vectors in the ANC system is proposed for online learning. Lastly, we present the results of a numerical study to test our ANC methodology with realistic interior noise measurement obtained from Korean railway trains.

• Journal of Institute of Control, Robotics and Systems
• /
• v.7 no.3
• /
• pp.209-217
• /
• 2001

In this paper, we propose a linear quadratic regulator(LQR) controller design for the active suspension using evolution strategy(ES) and neural network. We can improve the inherent suspension problem, the trade-off between ride quality and suspension travel by selecting appropriate weight in the LQR-objective function. Since any definite rules for selecting weights do not exist, we replace the designers trial-and-error method with ES that is an optimization algorithm. Using the ES, we can find the proper control gains for selected frequencies, which have major effects on the vibrations of the vehicle. The relationship between the frequencies and proper control gains are generalized by use of the neural networks. When the vehicle is driven, the trained neural network is activated and provides the proper gains for operating frequencies. And we adopted double sky-hook control to protect car component when passing large bump. Effectiveness of our design has been shown compared to the conventional sky-hook controller through simulation studies.

• Journal of Intelligence and Information Systems
• /
• v.22 no.2
• /
• pp.127-142
• /
• 2016

Deep learning model is a kind of neural networks that allows multiple hidden layers. There are various deep learning architectures such as convolutional neural networks, deep belief networks and recurrent neural networks. Those have been applied to fields like computer vision, automatic speech recognition, natural language processing, audio recognition and bioinformatics where they have been shown to produce state-of-the-art results on various tasks. Among those architectures, convolutional neural networks and recurrent neural networks are classified as the supervised learning model. And in recent years, those supervised learning models have gained more popularity than unsupervised learning models such as deep belief networks, because supervised learning models have shown fashionable applications in such fields mentioned above. Deep learning models can be trained with backpropagation algorithm. Backpropagation is an abbreviation for "backward propagation of errors" and a common method of training artificial neural networks used in conjunction with an optimization method such as gradient descent. The method calculates the gradient of an error function with respect to all the weights in the network. The gradient is fed to the optimization method which in turn uses it to update the weights, in an attempt to minimize the error function. Convolutional neural networks use a special architecture which is particularly well-adapted to classify images. Using this architecture makes convolutional networks fast to train. This, in turn, helps us train deep, muti-layer networks, which are very good at classifying images. These days, deep convolutional networks are used in most neural networks for image recognition. Convolutional neural networks use three basic ideas: local receptive fields, shared weights, and pooling. By local receptive fields, we mean that each neuron in the first(or any) hidden layer will be connected to a small region of the input(or previous layer's) neurons. Shared weights mean that we're going to use the same weights and bias for each of the local receptive field. This means that all the neurons in the hidden layer detect exactly the same feature, just at different locations in the input image. In addition to the convolutional layers just described, convolutional neural networks also contain pooling layers. Pooling layers are usually used immediately after convolutional layers. What the pooling layers do is to simplify the information in the output from the convolutional layer. Recent convolutional network architectures have 10 to 20 hidden layers and billions of connections between units. Training deep learning networks has taken weeks several years ago, but thanks to progress in GPU and algorithm enhancement, training time has reduced to several hours. Neural networks with time-varying behavior are known as recurrent neural networks or RNNs. A recurrent neural network is a class of artificial neural network where connections between units form a directed cycle. This creates an internal state of the network which allows it to exhibit dynamic temporal behavior. Unlike feedforward neural networks, RNNs can use their internal memory to process arbitrary sequences of inputs. Early RNN models turned out to be very difficult to train, harder even than deep feedforward networks. The reason is the unstable gradient problem such as vanishing gradient and exploding gradient. The gradient can get smaller and smaller as it is propagated back through layers. This makes learning in early layers extremely slow. The problem actually gets worse in RNNs, since gradients aren't just propagated backward through layers, they're propagated backward through time. If the network runs for a long time, that can make the gradient extremely unstable and hard to learn from. It has been possible to incorporate an idea known as long short-term memory units (LSTMs) into RNNs. LSTMs make it much easier to get good results when training RNNs, and many recent papers make use of LSTMs or related ideas.

• Journal of KIISE:Information Networking
• /
• v.29 no.4
• /
• pp.386-396
• /
• 2002

This paper presents a neural network-based near-optimal routing algorithm. It employs a modified Hopfield Neural Network (MHNN) as a means to solve the shortest path problem. It uses every piece of information that is available at the peripheral neurons in addition to the highly correlated information that is available at the local neuron. Consequently, every neuron converges speedily and optimally to a stable state. The convergence is faster than what is usually found in algorithms that employ conventional Hopfield neural networks. Computer simulations support the indicated claims. The results are relatively independent of network topology for almost all source-destination pairs, which nay be useful for implementing the routing algorithms appropriate to multi -hop packet radio networks with time-varying network topology.

• Journal of Institute of Control, Robotics and Systems
• /
• v.8 no.2
• /
• pp.120-125
• /
• 2002

An optimization approach is used to partition the field of view. A cost function is defined to represent the constraints on the solution, which is then mapped onto a two-dimensional Hopfield neural network for minimization. Each neuron in the network represents a possible match between a field of view and one or multiple objects. Partition is achieved by initializing each neuron that represents a possible match and then allowing the network to settle down into a stable state. The network uses the initial inputs and the compatibility measures between a field of view and one or multiple objects to find a stable state.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.58 no.12
• /
• pp.2511-2519
• /
• 2009

In this paper, we introduce the methodological system design via feature selection using Principal Component Analysis and Particle Swarm Optimization algorithms. The overall methodological system design comes from three kinds of modules such as preprocessing module, feature extraction module, and recognition module. First, Histogram equalization enhance the quality of image by exploiting contrast effect based on the normalized function generated from histogram distribution values of 2D face image. Secondly, PCA extracts feature vectors to be used for face recognition by using eigenvalues and eigenvectors obtained from covariance matrix. Finally the feature selection for face recognition among the entire feature vectors is considered by means of the Particle Swarm Optimization. The optimized Polynomial-based Radial Basis Function Neural Networks are used to evaluate the face recognition performance. This study shows that the proposed methodological system design is effective to the analysis of preferred face recognition.

• KSCE Journal of Civil and Environmental Engineering Research
• /
• v.29 no.3C
• /
• pp.115-125
• /
• 2009

The preconsolidation pressure has been commonly determined by oedometer test. However, it can also be determined by insitu test, such as piezocone test with theoretical and(or) empirical correlations. Recently, Neural Network (NN) theory was applied and some models were proposed to estimate the preconsolidation pressure or OCR. It was already found that NN model can come over the site dependency and prediction accuracy is greatly improved when compared with present theoretical and empirical models. However, since the optimization process of synaptic weights of NN model is dependent on the initial synaptic weights, NN models which are trained with different initial weights can't avoid the variability on prediction result for new database even though they have same structure and use same transfer function. In this study, Committee Neural Network (CNN) model is proposed to improve the initial weight dependency of multi-layered neural network model on the prediction of preconsolidation pressure of soft clay from piezocone test result. Prediction results of CNN model are compared with those of conventional empirical and theoretical models and multi-layered neural network model, which has the optimized structure. It was found that even though the NN model has the optimized structure for given training data set, it still has the initial weight dependency, while the proposed CNN model can improve the initial weight dependency of the NN model and provide a consistent and precise inference result than existing NN models.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.60 no.2
• /
• pp.398-406
• /
• 2011

In this paper, we introduce a new architecture of PSO-based Polynomial Neural Networks (PNN) and discuss its comprehensive design methodology. The conventional PNN is based on a extended Group Method of Data Handling (GMDH) method, and utilized the polynomial order (viz. linear, quadratic, and modified quadratic) as well as the number of node inputs fixed (selected in advance by designer) at Polynomial Neurons located in each layer through a growth process of the network. Moreover it does not guarantee that the conventional PNN generated through learning results in the optimal network architecture. The PSO-based PNN results in a structurally optimized structure and comes with a higher level of flexibility that the one encountered in the conventional PNN. The PSO-based design procedure being applied at each layer of PNN leads to the selection of preferred PNs with specific local characteristics (such as the number of input variables, input variables, and the order of the polynomial) available within the PNN. In the sequel, two general optimization mechanisms of the PSO-based PNN are explored: the structural optimization is realized via PSO whereas in case of the parametric optimization we proceed with a standard least square method-based learning. To evaluate the performance of the PSO-based PNN, the model is experimented with using Gas furnace process data, and pH neutralization process data. For the characteristic analysis of the given entire data with non-linearity and the construction of efficient model, the given entire system data is partitioned into two type such as Division I(Training dataset and Testing dataset) and Division II(Training dataset, Validation dataset, and Testing dataset). A comparative analysis shows that the proposed PSO-based PNN is model with higher accuracy as well as more superb predictive capability than other intelligent models presented previously.

• Journal of Computing Science and Engineering
• /
• v.9 no.2
• /
• pp.98-107
• /
• 2015

Hierarchical constraints and constraint satisfaction were analyzed in order to solve the problem of conflict detection in collaborative design. The constraints were divided into two sets: one set consisted of known constraints and the other of unknown constraints. The constraints of the two sets were detected with corresponding methods. The set of the known constraints was detected using an interval propagation algorithm, a back propagation (BP) neural network was proposed to detect the set with the unknown constraints. An immune algorithm (IA) was utilized to optimize the weights and the thresholds of the BP neural network, and the steps were designed for the optimization process. The results of the simulation indicated that the BP neural network that was optimized by IA has a better performance in terms of convergent speed and global searching ability than a genetic algorithm. The constraints were described using the eXtensible Markup Language (XML) for computers to be able to automatically recognize and establish the constraint network. The implementation of the conflict detection system was designed based on constraint satisfaction. A wind planetary gear train is taken as an example of collaborative design with a conflict detection system.

• International Journal of Control, Automation, and Systems
• /
• v.4 no.6
• /
• pp.725-735
• /
• 2006

This paper deals with the problem of self-learning cooperative motion control for the pushing task of a humanoid robot in the sagittal plane. A model with 27 linked rigid bodies is developed to simulate the system dynamics. A simple genetic algorithm(SGA) is used to find the cooperative motion, which is to minimize the total energy consumption for the entire humanoid robot body. And the multi-layer neural network based on backpropagation(BP) is also constructed and applied to generalize parameters, which are obtained from the optimization procedure by SGA, in order to control the system.