• Journal of Power Electronics
• /
• v.13 no.4
• /
• pp.516-527
• /
• 2013

This paper presents a new overall system for state-of-available-power (SoAP) prediction for a lithium-ion battery pack. The essential part of this method is based on an adaptive network architecture which utilizes both fuzzy model (FIS) and artificial neural network (ANN) into the framework of adaptive neuro-fuzzy inference system (ANFIS). While battery aging proceeds, the system is capable of delivering accurate power prediction not only for room temperature, but also at lower temperatures at which power prediction is most challenging. Due to design property of ANN, the network parameters are adapted on-line to the current battery states (state-of-charge (SoC), state-of-health (SoH), temperature). SoC is required as an input parameter to SoAP module and high accuracy is crucial for a reliable on-line adaptation. Therefore, a reasonable way to determine the battery state variables is proposed applying a combination of several partly different algorithms. Among other SoC boundary estimation methods, robust extended Kalman filter (REKF) for recalibration of amp hour counters was implemented. ANFIS then achieves the SoAP estimation by means of time forward voltage prognosis (TFVP) before a power pulse occurs. The trade-off between computational cost of batch-learning and accuracy during on-line adaptation was optimized resulting in a real-time system with TFVP absolute error less than 1%. The verification was performed on a software-in-the-loop test bench setup using a 53 Ah lithium-ion cell.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.35 no.12B
• /
• pp.1219-1226
• /
• 2010

It is necessary to implementation of system contain intelligent decision making algorithm because discriminant and prediction system for Red Tide is insufficient development and the study of red tide are focused for the investigation of chemical and biological causing. In this paper, we designed inference system using case based reasoning method and implemented knowledge base that case for Red Tide. We used K-Nearest Neighbor algorithm for recommend best similar case and input 375 EA by case for Red Tide case base. As a result, conducted 10-fold cross verification for minimal impact from learning data and acquired confidence, we obtained about 84.2% average accuracy for Red Tide case and the best performance results in case by number of similarity classification k is 5. And, we implemented Red Tide monitoring system using inference result.

• Journal of Industrial Convergence
• /
• v.20 no.4
• /
• pp.23-30
• /
• 2022

Recently, with the development of artificial intelligence (AI) and deep learning, the importance of conversational artificial intelligence chatbots is being highlighted. In addition, chatbot research is being conducted in various fields. To build a chatbot, it is developed using an open source platform or a commercial platform for ease of development. These chatbot platforms mainly use RNN and application algorithms. The RNN algorithm has the advantages of fast learning speed, ease of monitoring and verification, and good inference performance. In this paper, a method for improving the inference performance of RNNs and applied algorithms was studied. The proposed method used the word group expansion learning technique of key words for each sentence when RNN and applied algorithm were applied. As a result of this study, the RNN, GRU, and LSTM three algorithms with a cyclic structure achieved a minimum of 0.37% and a maximum of 1.25% inference performance improvement. The research results obtained through this study can accelerate the adoption of artificial intelligence chatbots in related industries. In addition, it can contribute to utilizing various RNN application algorithms. In future research, it will be necessary to study the effect of various activation functions on the performance improvement of artificial neural network algorithms.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.8C
• /
• pp.1149-1158
• /
• 2004

In this thesis, we created a Fuzzy rule in a Fuzzy logic that are Fuzzy logic which is composed of linguistic rules and Fuzzy inference engine for effective traffic control in ATM networks. The parameters of the Fuzzy rules are adapted to minimize the given performance index in both cases. In other words, the difuzzification value controls the threshold in the buffer to arrival ratio to traffic priority (low or high) using fuzzy set theory for traffic connected after reasoning. Also, show experiment result about rule by MATLAB6.5 and on-line bulid-up to verify validity of created Fuzzy rule. As a result, we can verify that threshold value in buffer is efficiently controlled by the traffic arrival ratio.

• Journal of Korean Library and Information Science Society
• /
• v.44 no.1
• /
• pp.385-408
• /
• 2013

The purpose of this study is to build the tax ontology which can be used to manage imposables by the state or local governments. In this, the tax and related concepts were analyzed and then concept hierarchy i.e., taxonomies were formed. Especially, in the concept hierarchy, after multiple inherits were decomposed as 'primitive concepts' and then Rector's 'methodology of ontology implementation normalization', in which defined concepts were recombined, was used. The methodology employed was that the tax system, which was entangled with the direct taxes, local taxes, and property taxes that has multiple-inherits, was expressed explicitly and logically. After that, automatic classification was carried out through the inference engine, consistency was verified. Finally, some practical cases of ontology created were enumerated.

• Korean Institute of Interior Design Journal
• /
• v.19 no.6
• /
• pp.38-46
• /
• 2010

The objective of this study is to find an educational method that is able to increase creativity using both left and right directed thinking with complementary cooperation. The premise of experimental study is that analogical inference is a great help to produce a creative design, and the design tasks of residential space are given to 20 students, voluntary participants in four experimental tests. The first test is conducted with fundamental conditions such as site or location, users and their design requirements. Other three tests make a clear distinction with three cases using verbal analogy from many keywords, using visual analogy from many images and using verbal visual analogy from keywords and visual images. Consequently, when students use both verbal and visual analogy in solving design tasks their creative ability qualitatively as well as quantitatively is higher than in using only verbal analogy or visual analogy. Further study will be progressed with the design tasks of residential space in order to have an effective verification by comparing students' design results classified into two groups. One is a control group that consists of sophomore students in a college and another is a comparison group that consists of sophomore students in an university.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.41 no.10
• /
• pp.69-77
• /
• 2004

In this thesis, we created a Fuzzy rule in a Fuzzy logic that are fuzzy logic which is composed of linguistic rules and Fuzzy inference engine for effective traffic control in ATM networks. The parameters of the Fuzzy rules are adapted to minimize the given performance index in both cases. In other words, the difuzzification value controls the service rate in the server to total traffic arrival ratio and buffer occupancy ratio using fuzzy set theory for traffic connected after reasoning. Also, show experiment result about rule by MATLAB6.5 and on-line bulid-up to verify validity of created Fuzzy rule. As a result, we can verify that service ratio in server is efficiently controlled by the total traffic arrival ratio and buffer occupancy ratio.

• Korea Trade Review
• /
• v.44 no.3
• /
• pp.253-269
• /
• 2019

This study examines the utilization effect of FTA on the sales and profit of the secondary auto-part suppliers in two aspects: 1) firm-level survey on the utilization of FTA and 2) statistical inference on the FTA effects on their sales and profit. We use the GLS statistical technique, panel and survey data of 130 secondary auto-part suppliers in the Daegu and Kyungpook area in 2007-2016. The evidence shows that the secondary auto-part suppliers struggle to prepare documents for the FTA certificates of origin due to their small firm size, although they utilize the FTA at the relatively high rate of 92 percent. Statistical evidence shows that an increase in the export of the first vender significantly affects the sale of the secondary auto-part vendors, but not their profit. The low profit and high managing cost of utilizing the FTA deteriorate the utilization effect of the FTA of the secondary auto-part vendors, which is a key factor in a global supply chain and for the competitiveness of the automobile industry. Government policies are required for the secondary auto-part suppliers to utilize the FTA more effectively and share the benefits of tariff reduction with first auto-part vendors under FTA transaction.

• Science of Emotion and Sensibility
• /
• v.17 no.1
• /
• pp.79-92
• /
• 2014

This study is aimed at minimizing the errors in using a sensitivity evaluation model that could occur when sensitivity analysis method is actively used for design evaluation. To extract words of a contrantion-type model by product, interior space, and streetscape design, primary word refinement was conducted with the words extracted from preceding studies. The analysis revealed that 19 words were used in all three fields. A reliability analysis revealed that different words had a bad impact on the reliability in each field. The applicability was reviewed through reliability analysis of EPA model as contraction-type and PAD model as inference-type. The results are as follows. Although the reliability of the contrantion-type model was higher than that of inference-type model in all three fields, the differences in Cronbach's Alpha were small. Also, When the reliability was analyzed after deleting the words that had a bad impact on reliability, the differences in the reliability's coefficients were clearly significant. Therefore, it is necessary to select words suitable for sensitivity evaluation target and objectivity of the evaluation can be boosted by using a proper model. Analysis of the sensitivity evaluation model suitable for future environmental evaluation should be analyzed with various statistical methods, beyond verification of reliability.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.12
• /
• pp.59-68
• /
• 2022

This paper presents a method for 1:1 verification by comparing the similarity between the given real product image and the drawing image. The proposed method combines two existing CNN-based deep learning models to construct a Siamese Network. After extracting the feature vector of the image through the FC (Fully Connected) Layer of each network and comparing the similarity, if the real product image and the drawing image (front view, left and right side view, top view, etc) are the same product, the similarity is set to 1 for learning and, if it is a different product, the similarity is set to 0. The test (inference) model is a deep learning model that queries the real product image and the drawing image in pairs to determine whether the pair is the same product or not. In the proposed model, through a comparison of the similarity between the real product image and the drawing image, if the similarity is greater than or equal to a threshold value (Threshold: 0.5), it is determined that the product is the same, and if it is less than or equal to, it is determined that the product is a different product. The proposed model showed an accuracy of about 71.8% for a query to a product (positive: positive) with the same drawing as the real product, and an accuracy of about 83.1% for a query to a different product (positive: negative). In the future, we plan to conduct a study to improve the matching accuracy between the real product image and the drawing image by combining the parameter optimization study with the proposed model and adding processes such as data purification.