• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1997.10a
• /
• pp.181-184
• /
• 1997

This paper propose the systematic procedure of the fuzzy model based controller for the continuous nonlinear system. Fuzzy controller have been successfully applied to many uncertain and complex industrial plants. The design of the fuzzy controller mainly depends on the knowledge from the expert who are familiar with the plant by trial and error. Therefore we need more systematic approach to the design of the fuzzy controller. In this paper, we design fuzzy model based controller applied to the nonlinear system. Unlike the design procedures reported in[8] and[9], we use the nonlinear process directly in designing the controller. This controller has been successfully applied to an inverted pendulum.

• The Transactions of the Korean Institute of Electrical Engineers A
• /
• v.48 no.6
• /
• pp.800-806
• /
• 1999

This paper proposes IIKBAG(Intelligent Interlocking Knowledge Base Generator) which can build automatically the interlocking knowledge base utilized as the real-time interlocking strategy of the electronic interlocking system in order to enhance it's reliability and expansion. The IIKBAG consists of the inference engine and the knowledge base. The former has an auto-learning function which searches all the train routes for the given station model based on heuristic search technique while dynamically searching the model, and then generates automatically the interlocking patterns obtained from the interlocking relations of signal facilities on the routes. The latter is designed as the structure which the real-time expert system embedded on IS(Interlocking System) can use directly in order to enhances the reliability and accuracy. The IIKBAG is implemented in C computer language for the purpose of the build and interface of the station structure database. And, a typical station model is simulated to prove the validity of the proposed IIKBAG.

• Journal of Information Technology Services
• /
• v.11 no.sup
• /
• pp.139-152
• /
• 2012

In the medical web forum, people share medical experience and information as patients and patents' families. Some people search medical information written in non-expert language and some people offer words of comport to who are suffering from diseases. Medical web forums play a role of the informative support and the emotional support. We propose the automatic classification model of articles in the medical web forum into the information support and emotional support. We extract text features of articles in web forum using text mining techniques from the perspective of linguistics and then perform supervised learning to classify texts into the information support and the emotional support types. We adopt the Support Vector Machine (SVM), Naive-Bayesian, decision tree for automatic classification. We apply the proposed model to the HealthBoards forum, which is also one of the largest and most dynamic medical web forum.

• Proceedings of the KSR Conference
• /
• 2008.06a
• /
• pp.1499-1504
• /
• 2008

This study shows risk-based evaluation results of casualty accidents for passengers, railway staffs and MOP(Member of public) on the national railway in South Korea. To evaluate risk of these accidents, the hazardous events and the hazardous factors were identified by the review of the accident history and engineering interpretation of the accident behavior. A probability evaluation model for each hazardous event which was based on the accident appearance scenario was developed by using the Fault Tree Analysis (FTA) technique. The probability for each hazardous event was evaluated from the historical data and structured expert judgment. In addition, the severity assessment model utilized by the Event Tree Analysis (ETA) technique was composed of the accident progress scenarios. And the severity for the hazardous events was estimated using fatalities and weighted injuries. The risk assessment model developed can be effectively utilized in defining the risk reduction measures in connection with the option analysis.

• Proceedings of the KIEE Conference
• /
• 2000.07d
• /
• pp.2333-2335
• /
• 2000

The main motivation of this research is to develop an intelligent control strategy for Activated Sludge Process(ASP). ASP is a complex and nonlinear dynamic system because of the characteristic of a wastewater, the change of an influent flow rate, weather conditions, and etc. The mathematical model of ASP also includes uncertainties which are ignored or not considered by process engineer or controller designer. The ASP is generally controlled by a PID controller that consists of fixed proportional, integral, and derivative gain values. The PID gains can be adjusted by the expert in the ASP. The ASP model based on Matlab$^{(R)}$5.3/Simulink$^{(R)}$3.0 is developed in this paper. Various control methods are applied to the ASP model and the control results are disscussed. Three control methods are designed and tested: conventional PID controller, fuzzy logic control approach to modify setpoints, and fuzzy-PID control method.

• Proceedings of the KIEE Conference
• /
• 1998.07c
• /
• pp.1073-1075
• /
• 1998

This paper presents the method of parameter estimation of diffusion model for monitoring Demand-Side Management Program. Bass diffusion model was applied in this paper, which has different values according to parameters ; coefficients of innovation, imitation and potential adopters. Though it is very important to estimate three parameter, there are no empirical results in practice. Thus, this paper presents the method of parameter estimation in case of few data with constraints to reduce the possibility of bad estimation. The constraints are empirical results or expert's decision. Case studies show the diffusion curves of high-efficient lighting and also forecasting of the peak value for power demand considering diffusion of high-efficient lighting, the feedback and least-square parameter estimation method used in this paper enable us to evaluate the status and forecasting of the effect of DSM program.

• Journal of the Korean Dietetic Association
• /
• v.15 no.4
• /
• pp.356-363
• /
• 2009

The study aimed to determine the optimal mixing ratio of two different amounts of Letinus edodes powder and egg for preparation of pasta. The complete analysis was conducted using the Design Expert 7 program (State-Easy, Minneapolis, MN). Response surface methodology revealed 10 experimental points, including two replicates for L. edodes powder and egg. L. edodes pasta formulation was optimized using rheology. Lightness and redness displayed a lineal model pattern, whereas yellowness was represented by a quadratic model. The sensory evaluation parameters of texture, color, flavor, appearance and overall quality showed significant differences for color (p<0.01), flavor (p<0.05), texture (p<0.05) and overall quality (p<0.05). All sensory parameters showed significant values in a quadratic model. The optimum formulations processed by numerical and graphical optimization were determined as 15 g of L. edodes powder and 37.5 g of egg.

• ETRI Journal
• /
• v.46 no.3
• /
• pp.513-525
• /
• 2024

Deep neural networks trained on labeled medical data face major challenges owing to the economic costs of data acquisition through expensive medical imaging devices, expert labor for data annotation, and large datasets to achieve optimal model performance. The heterogeneity of diseases, such as Alzheimer's disease, further complicates deep learning because the test cases may substantially differ from the training data, possibly increasing the rate of false positives. We propose a reconstruction-based self-supervised anomaly detection model to overcome these challenges. It has a dual-subnetwork encoder that enhances feature encoding augmented by skip connections to the decoder for improving the gradient flow. The novel encoder captures local and global features to improve image reconstruction. In addition, we introduce an entropy-based image conversion method. Extensive evaluations show that the proposed model outperforms benchmark models in anomaly detection and classification using an encoder. The supervised and unsupervised models show improved performances when trained with data preprocessed using the proposed image conversion method.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.317-320
• /
• 2004

In this work we have managed to find parameters for defining athlete's aerobic and anaerobic thresholds. Thresholds which are of vital importance for top athletes. It is shown how differential evolution and different similarity measures has been used to tune computational model for threshold definitions. From our results it is obvious that the use of right parameter values for this kind expert system is of vital importance.

• Proceedings of the IEEK Conference
• /
• 2003.07c
• /
• pp.2827-2830
• /
• 2003

Tele manipulator is distingushed from industrial robot by iterating same specified work. Manipulator operator is included in control loop for controlling the telemanipulator because he decide directly during the work and order controllabily reducing the modelling error of telemanipulator which depend on the PID controller. But position-force control method of bidirectional control impose unsafety of vibiration and Analytic Hierachy Method can stabilize for reducing nonlinear modelling error by expert operator because of transformation empirical control rule to linear model.