• 전자공학회논문지B
• /
• 제32B권11호
• /
• pp.1536-1544
• /
• 1995

We propose a design scheme of a binary decision tree and apply it to the tire tread pattern recognition problem. In this scheme, a binary decision tree is constructed by using fuzzy C-means( FCM ) algorithm. All the available features are used while clustering. At each node, the best feature or feature subset among these available features is selected based on proposed similarity measure. The decision tree can be used for the classification of unknown patterns. The proposed design scheme is applied to the tire tread pattern recognition problem. The design procedure including feature extraction is described. Experimental results are given to show the usefulness of this scheme.

• Elastomers and Composites
• /
• 제56권4호
• /
• pp.223-233
• /
• 2021

In this study, epoxidized liquid isoprene rubber (E-LqIR) was used as a processing aid in a silica-filled natural rubber compound to improve the fuel efficiency, abrasion resistance, and oil migration problems of truck and bus radial tire tread. The wear resistance, fuel efficiency, and extraction resistance of the compound were evaluated according to the molecular weight of E-LqIR. Results of the evaluation showed that the E-LqIR compound had a lower chemical crosslink density than that of a treated distillate aromatic extract (TDAE) oil compound because of the sulfur consumption of E-LqIR. However, the filler-rubber interaction improved because of the reaction of E-LqIR with silica and crosslink with the base rubber by sulfur. As the molecular weight of E-LqIR increased, crosslink with sulfur was facilitated, and the filler-rubber interaction improved, resulting in improved abrasion resistance. The fuel efficiency performance of the E-LqIR compound was poorer than that of the TDAE oil compound because of the low chemical crosslink density and hysteresis loss at the free chain end of E-LqIR. However, the fuel efficiency performance improved as the molecular weight of E-LqIR increased.

• Elastomers and Composites
• /
• 제55권4호
• /
• pp.289-299
• /
• 2020

Highly aromatic (HA) oils are common processing aids used in tire tread compounds. However, they often bleed and evaporate from the vulcanizates during tire use. Thus, the mechanical and dynamical properties of the tire decrease. To overcome this problem, we investigated nonfunctionalized liquid butadiene rubber (LBR-305, Kuraray) and center-functionalized liquid butadiene rubber (C-LqBR), polymerized by anionic polymerization. In addition to the liquid butadiene rubbers, p-tert-octylphenol (P-Resin) and C5 hydrocarbon (H-Resin) tackifier resins, which can induce entanglement of rubber compounds, were researched as a processing aid to solve the bleeding problem. Liquid butadiene rubbers have significantly reduced extraction loss by crosslinking with the main rubber chain. They have also increased the abrasion resistance and showed similar or better mechanical and dynamical properties against HA oils. However, resin compounds did not show differences in extraction loss compared to HA oil compounds; instead, they showed increased wet traction.

• 한국대기환경학회지
• /
• 제2권2호
• /
• pp.60-65
• /
• 1986

Atmospheric particulates were collected at a site near the front gate of the Yonsei University using nine stages Andersen air sampler and the distribution of seasonal particle size was investigated. Rubber, Pb and Zn contents of the collected particulates in each stage were determined. Particle size distribution of atmospheric particulate, which was made by concentration distribution curve method, was usually divided into two groups, course (particles larger than 1 - 2 $\mu m in diameter$) and fine (particles smaller than 1 - 2 \mu m in diameter$) groups, regardless of sampling period. More than 80 percent of the total rubber contents in atmospheric particulates were larger than $5 \mu m$ in diameter, meaning that most of rubber particles were originated from tire tread. After benzene extraction for 4 hrs, the extracts were analyzed by Curie-point pyrolysis gas chromatography for rubber content. Pb and Zn contents were determined by atomic absorption spectroscopy. The annual average concentration of rubber particles was $4.2 \mu g/m^3$, which corresponded to 2.2% of the annual average total suspended particulates. Average concentration of styrene brtadiene rubber was about five times that of natural rubber. Annual average concentrations of Pb and Zn were $1.2 \mu g/m^3 and 0.4 \mu g/m^3$ respectively, which corresponded to about 0.7% and 0.2% of the annual average total suspended particulates.

• 한국지능시스템학회논문지
• /
• 제24권6호
• /
• pp.640-645
• /
• 2014

인터넷 상에서 많은 사람들은 사용자 간의 의사소통과 정보 공유, 사회적 관계를 생성하기 위한 방법으로 소셜 네트워크 서비스를 이용한다. 그 중 대표적인 트위터는 하루에 수백만 건의 소셜 데이터가 발생하기 때문에 수집되고 있는 데이터의 양이 엄청나다. 이 방대한 양의 데이터로부터 의미 있는 정보를 추출하는 소셜 마이닝이 집중적으로 연구되고 있다. 트위터는 일반적으로 유용한 정보 혹은 공유하고자 하는 내용을 팔로잉-팔로워 관계를 이용해 쉽게 전달하고 리트윗할 수 있다. 소셜 미디어에서 트윗 데이터에 대한 토픽 모델링은 이슈를 추적하기 위한 좋은 도구이다. 짧은 텍스트 기반인 트윗 데이터의 제한점을 극복하기 위해, 사용자를 노드로 사용자간 댓글과 리트윗 메시지의 여부를 간선으로 하는 그래프 구조를 갖는 댓글 그래프의 개념을 소개한다. 토픽 모델링의 대표적인 방법인 LDA 토픽 모델이 짧은 텍스트 데이터에 대해 비효율적인 것을 보완하기 위한 방법으로, 이 논문에서는 짧은 문서의 수를 줄이고 마이닝 결과의 질을 향상시키기 위한 댓글 그래프를 사용하는 토픽 모델링 방법을 소개한다. 제안한 모델은 토픽 모델링 방법으로 LDA 모델을 사용하였으며, 7일간 수집한 트윗 데이터에 대한 실험 결과를 보인다.