• Journal of Power Electronics
• /
• v.2 no.1
• /
• pp.46-54
• /
• 2002

In this paper a neural network controller for achieving maximum power tracking as well as output voltage regulation, for a wind energy conversion system (WECS) employing a permanent magnet synchronous generator is proposed. The permanent magnet generator (PMG) supplies a dc load via a bridge rectifier and two buck-boost converters. Adjusting the switching frequency of the first buck-boost converter achieves maximum power tracking. Adjusting the switching frequency of the second buck-boost converter allows output voltage regulation. The on-time of the switching devices of the two converters are supplied by the developed neural network (NN). The effect of sudden changes in wind speed and/ or in reference voltage on the performance of the NN controller are explored. Simulation results showed the possibility of achieving maximum power tracking and output voltage regulation simulation with the developed neural network controllers. The results proved also the fast response and robustness of the proposed control system.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2003.07b
• /
• pp.1163-1167
• /
• 2003

As a PD classification method, statistical distribution parameters have been used during several ten years. And this parameters are recently finger print method, NN(Neural Network) and etc. So in this paper we studied finger print method and NN with BP(Back propagation) learning algorithm using the statistical distribution parameter, and compared with two method as classification method. As a result of comparison, classification of NN is more good result than Finger print method in respect to calculation speed, visible effect and simplicity. So, NN has more advantage as a tool for PD classification.

• Journal of the Korea Computer Industry Society
• /
• v.2 no.9
• /
• pp.1191-1198
• /
• 2001

An increasingly important facet of software development is the ability to estimate the associate cost and effort of development early in the development life cycle. Most of the proposed models are based upon a combination of intuition, expert judgement, and regression analysis of empirical data. Overall, the work has failed to produce any single model that can be applied with a reasonable degree of success to a variety of environments. This paper presents a neural network (NN) model that related software development effort to software size measured in function element types. The heuristic approach is applied to decide the number of hidden neurons in NN from the relationship between input-output pairs. The research describes appropriate NN modeling in the context of a case study for 24 software development projects. Also, this paper compared the NN model with a regression analysis model and found the NN model has better accuracy.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• v.8 no.3
• /
• pp.185-191
• /
• 2008

In this paper, experimental studies of a neural network (NN) control technique for non-model based position control of the x-y table robot are presented. Decentralized neural networks are used to control each axis of the x-y table robot separately. For an each neural network compensator, an inverse control technique is used. The neural network control technique called the reference compensation technique (RCT) is conceptually different from the existing neural controllers in that the NN controller compensates for uncertainties in the dynamical system by modifying desired trajectories. The back-propagation learning algorithm is developed in a real time DSP board for on-line learning. Practical real time position control experiments are conducted on the x-y table robot. Experimental results of using neural networks show more excellent position tracking than that of when PD controllers are used only.

• Journal of Korean Society for Quality Management
• /
• v.41 no.2
• /
• pp.249-260
• /
• 2013

Purpose: The main objective of this paper is to propose an RD method by developing a neural network (NN)-based estimation approach in order to provide an alternative aspect of response surface methodology (RSM). Methods: A specific modeling procedure for integrating NN principles into response function estimations is identified in order to estimate functional relationships between input factors and output responses. Finally, a comparative study based on simulation is performed as verification purposes. Results: This simulation study demonstrates that the proposed NN-based RD method provides better optimal solutions than RSM. Conclusion: The proposed NN-based RD approach can be a potential alternative method to utilize many RD problems in competitive manufacturing nowadays.

• Proceedings of the KIPE Conference
• /
• 2001.10a
• /
• pp.656-659
• /
• 2001

In this paper a neural network controller for achieving maximum power tracking as well as output voltage regulation, for a wind energy conversion system(WECS) employing a permanent magnet synchronous generator, is proposed. The permanent magnet generator (PMG) supplies a dc load via a bridge rectifier and two buck-boost converters. Adjusting the switching frequency of the first buck-boost converter achieves maximum power tracking. Adjusting the switching frequency of the second buck-boost converter allows output voltage regulation. The on-times of the switching devices of the two converters are supplied by the developed neural network(NN). The effect of sudden changes in wind speed ,and/or in reference voltage on the performance of the NN controller are explored. Simulation results showed the possibility of achieving maximum power tracking and output voltage regulation simultaneously with the developed neural network controller. The results proved also the fast response and robustness of the proposed control system.

• Advances in concrete construction
• /
• v.17 no.2
• /
• pp.53-66
• /
• 2024

Domain Name Systems (DNS) provide critical performance in directing Internet traffic. It is a significant duty of DNS service providers to protect DNS servers from bandwidth attacks. Data mining techniques may identify different trends in detecting anomalies, but these approaches are insufficient to provide adequate methods for querying traffic data in significant network environments. The patterns can enable the providers of DNS services to find anomalies. Accordingly, this research has used a new approach to find the anomalies using the Neural Network (NN) because intrusion detection techniques or conventional rule-based anomaly are insufficient to detect general DNS anomalies using multi-enterprise network traffic data obtained from network traffic data (from different organizations). NN was developed, and its results were measured to determine the best performance in anomaly detection using DNS query data. Going through the R² results, it was found that NN could satisfactorily perform the DNS anomaly detection process. Based on the results, the security weaknesses and problems related to unpredictable matters could be practically distinguished, and many could be avoided in advance. Based on the R² results, the NN could perform remarkably well in general DNS anomaly detection processing in this study.

• The Transactions of the Korean Institute of Electrical Engineers
• /
• v.40 no.2
• /
• pp.224-233
• /
• 1991

The multilayer expanson of single layer NN (Neural Network) was needed to solve the linear seperability problem as shown by the classic example using the XOR function. The EBP (Error Back Propagation ) learning rule is often used in multilayer Neural Networks, but it is not without its faults: 1)D.Rimmelhart expanded the Delta Rule but there is a problem in obtaining Ca from the linear combination of the Weight matrix N between the hidden layer and the output layer and H, wich is the result of another linear combination between the input pattern and the Weight matrix M between the input layer and the hidden layer. 2) Even if using the difference between Ca and Da to adjust the values of the Weight matrix N between the hidden layer and the output layer may be valid is correct, but using the same value to adjust the Weight matrixd M between the input layer and the hidden layer is wrong. Recognitron III was proposed to solve these faults. According to simulation results, since Recognitron III does not learn the three layer NN itself, but divides it into several single layer NNs and learns these with learning patterns, the learning time is 32.5 to 72.2 time faster than EBP NN one. The number of patterns learned in a EBP NN with n input and output cells and n+1 hidden cells are 2**n, but n in Recognitron III of the same size. [5] In the case of pattern generalization, however, EBP NN is less than Recognitron III.

• KSBB Journal
• /
• v.16 no.5
• /
• pp.452-458
• /
• 2001

Concentrations of ammonia in biological processes were controlled by PID controllers and also neural network based controllers (NN controllers). A flow injection analysis system has been to on-line monitor the concentrations of ammonia in a bioreactor. The effect of the analysis error and the residence time of samples on the control performance were studied. The optimal neural network structure was investigated by using computer simulation and found to be a 3(input layer)-2(hidden layer)-1(output layer). The NN controller is often time consuming, but it has advantage over the PID controller in sensitivity. The 3-2-1 NN controller has been applied to control the ammonia concentrations in a simulated bioprocess and also a real cultivation process of yeast. The good control performance showed that the 3-2-1 NN controller based on the FIA system can be used to control the concentration of substrates in biological processes very well.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.168.1-168
• /
• 2001

This article describes a novel model-based compensating neural network (NN) model developed to be used in our active binocular head controller, which addresses both the kinematics and dynamics aspects in trying to precisely track a moving object of interest to keep it in view. The compensating NN model is constructed using two classes of self-tuning neural models: namely Neural Gas (NG) algorithm and SoftMax function networks. The resultant servo controller is shown to be able to handle the tracking problem with a minimum knowledge of the dynamic aspects of the system.