• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.13 no.4
• /
• pp.352-359
• /
• 1988

A learning control system which should have memory elements can be designed by utilizing the concept of parameter adaptation for unknown control object system parameters and regard it as a hybrid adaptive control system. A parameter adaptive learning control system applicable to a continuous time system has been already reported. Since there have been rapid developments in digital technology, it is possible to extend a continuous time parameter adaptive learning control system concept to a discrete time case. This problem is treated in this paper. Its justfication is proved and a simulation shows that this algorithms is effective.

• Proceedings of the KIEE Conference
• /
• 1998.11b
• /
• pp.648-650
• /
• 1998

This paper proposes a constructive algorithm for generating the Takagi-Sugeno type fuzzy model through the sequential learning from training data set. The proposed algorithm has a two-stage learning scheme that performs both structure and parameter learning simultaneously. The structure learning constructs fuzzy model using two growth criteria to assign new fuzzy rules for given observation data. The parameter learning adjusts the parameters of existing fuzzy rules using the LMS rule. To evaluate the performance of the proposed fuzzy modeling approach, well-known benchmark is used in simulation and compares it with other modeling approaches.

• Journal of Auto-vehicle Safety Association
• /
• v.13 no.4
• /
• pp.129-143
• /
• 2021

This paper proposes actuator fault detection and adaptive fault-tolerant control algorithms using performance index and human-like learning for longitudinal autonomous vehicles. Conventional longitudinal controller for autonomous driving consists of supervisory, upper level and lower level controllers. In this paper, feedback control law and PID control algorithm have been used for upper level and lower level controllers, respectively. For actuator fault-tolerant control, adaptive rule has been designed using the gradient descent method with estimated coefficients. In order to adjust the control parameter used for determination of adaptation gain, human-like learning algorithm has been designed based on perceptron learning method using control errors and control parameter. It is designed that the learning algorithm determines current control parameter by saving it in memory and updating based on the cost function-based gradient descent method. Based on the updated control parameter, the longitudinal acceleration has been computed adaptively using feedback law for actuator fault-tolerant control. The finite window-based performance index has been designed for detection and evaluation of actuator performance degradation using control error.

• 제어로봇시스템학회:학술대회논문집
• /
• 1996.10a
• /
• pp.95-98
• /
• 1996

Universal Learning Network(U.L.N.), which can model and control the large scale complicated systems naturally, consists of nonlinearly operated nodes and multi-branches that may have arbitrary time delays including zero or minus ones. Therefore, U.L.N. can be applied to many kinds of systems which are difficult to be expressed by ordinary first order difference equations with one sampling time delay. It has been already reported that learning algorithm of parameter variables in U.L.N. by forward and backward propagation is useful for modeling, managing and controlling of the large scale complicated systems such as industrial plants, economic, social and life phenomena. But, in the previous learning algorithm of U.L.N., time delays between the nodes were fixed, in other words, criterion function of U.L.N. was improved by adjusting only parameter variables. In this paper, a new learning algorithm is proposed, where not only parameter variables but also time delays between the nodes can be adjusted. Because time delays are integral numbers, adjustment of time delays can be carried out by a kind of random search procedure which executes intensified and diversified search in a single framework.

• Journal of Korea Multimedia Society
• /
• v.21 no.4
• /
• pp.441-449
• /
• 2018

Deep learning has been used for various processing centered on image recognition. One core algorithms of the deep learning, convolutional neural network is an deep neural network that specialized in image recognition. In this paper, we use a convolutional neural network to classify forest insects and propose an optimization method. Experiments were carried out by combining two weight initialization and six parameter update methods. As a result, the Xavier-SGD method showed the highest performance with an accuracy of 82.53% in the 12 different combinations of experiments. Through this, the latest learning algorithms, which complement the disadvantages of the previous parameter update method, we conclude that it can not lead to higher performance than existing methods in all application environments.

• Journal of the Korea Society of Computer and Information
• /
• v.22 no.12
• /
• pp.27-34
• /
• 2017

Design of an intelligent controller for efficient control in smart building is one of the effective technologies to reduce energy consumption by reducing response time with keeping comfortable level for inhabitants. In this paper, we focus on how to find major parameters in order to enhance the ability of HVAC(heating, ventilation, air conditioning) PID controller. For the purpose of that, we use machine learning technologies for tuning HVAC devices. We show the simulation results to illustrate the behavioral relation of whole system and each control parameter while learning process.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2004.10a
• /
• pp.313-316
• /
• 2004

Model selection is the process that sets up the regularization parameter in the support vector machine or regularization network by using the external methods such as general cross validation or L-curve criterion. This paper suggests that the regularization parameter can be obtained simultaneously within the learning process of neural networks without resort to separate selection methods. In this paper, extended kernel method is introduced. The relationship between regularization parameter and the bias term in the extended kernel is established. Experimental results show the effectiveness of the new model selection method.

• International Journal of Contents
• /
• v.4 no.2
• /
• pp.19-23
• /
• 2008

This paper presents a novel method to construct a personalized e-Learning system based on dynamic estimations of item parameters and learners' abilities, where the learning content objects are of the same intrinsic quality or homogeneously distributed and the estimations are carried out using IRT(Item Response Theory). The system dynamically connects the test and the corresponding learning procedures. Test results are directly applied to estimate examinee's ability and are used to modify the item parameters and the difficulties of learning content objects during the learning procedure is being operated. We define the learning unit 'Node' as an amount of learning objects operated so that new parameters can be re-estimated. There are various content objects in a Node and the parameters estimated at the end of current Node are directly applied to the next Node. We offer the most appropriate learning Node for a person's ability throughout the estimation processes of IRT. As a result, this scheme improves learning efficiency in web-base e-Learning environments offering the most appropriate learning objects and items to the individual students according to their estimated abilities. This scheme can be applied to any e-Learning subject having homogeneous learning objects and unidimensional test items. In order to construct the system, we present an operation scenario using the proposed system architecture with the essential databases and agents.

• Proceedings of the KIEE Conference
• /
• 2005.05a
• /
• pp.15-17
• /
• 2005

Recently genetic algorithm techniques have widely used in adaptive and control schemes for production systems. However, generally it costs a lot of time for learning in the case applied in control system. Furthermore, the physical meaning of genetic algorithm constructed as a result is not obvious. And this method has been used as a learning algorithm to estimate the parameter of a genetic algorithm used for identification of the process dynamics of nonlinear IIR filter and it was shown that this method offered superior capability over the genetic algorithm. A genetic algorithm is used to solve the parameter identification problem for linear and nonlinear digital filters. This paper goal estimate nonlinear IIR filter parameter using the genetic algorithm.

• Proceedings of the KIEE Conference
• /
• 2005.10b
• /
• pp.502-504
• /
• 2005

Recently genetic algorithm techniques have widely used in adaptive and control schemes for production systems. However, generally it costs a lot of time for learning in the case applied in control system. Furthermore, the physical meaning of genetic algorithm constructed as a result is not obvious. And this method has been used as a learning algorithm to estimate the parameter of a genetic algorithm used for identification of the process dynamics of FIR filter and it was shown that this method offered superior capability over the genetic algorithm. A genetic algorithm is used to solve the parameter identification problem for linear and nonlinear digital filters. This paper goal estimate FIR filter parameter using the genetic algorithm.