http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Effects of humidity and concentration on the light scattering dust sensor
김선태,황운호,전형재,김건호,이휘 한국냄새환경학회 2022 실내환경 및 냄새 학회지 Vol.21 No.4
Low-cost particulate matter (PM) sensors based on the light scattering principle measure the concentration of particles by the change of scattering intensity after light is irradiated onto the particles. It has been reported that when the relative humidity is high, water vapor may cause the expansion of airborne particles and affect the accuracy of the light scattering method for PM measurement, but it has also been shown that the effect of humidity is not significant or even negligible. Therefore, to determine the effect of humidity on the Plantower PMS7003 light scattering sensor, in this study, a BAM1020 (Beta Attenuation Monitoring) was installed alongside to continuously monitor the ambient atmospheric PM concentration for approximately four weeks. The sensors collected data at 10-minute intervals, resulting in a 1-hour average for comparative analysis. To accurately measure humidity, the performance of the Arduino + DHT22 humidity sensor was also pre-evaluated using a series of saturated salt solutions. The humidity was grouped into five intervals and analyzed by visual analysis. The results confirmed that there was no significant correlation between PM2.5 differences and humidity, which were randomly and uniformly distributed around the mean. However, since in the very low and high concentration ranges based on the beta-ray monitor measurements, the difference between the light scattering sensor PM10 measurement and the reference value is much larger than the difference between the PM2.5 and the reference value., there is an additional need to investigate the appropriate correction method for dust season or PM10. The results show that the outcomes of the light scattering sensor are more influenced by particle size and concentration than by humidity.
스팸 대응 시스템에서 특징 추출 방법 비교에 관한 연구
김종완,김희재,강신재,황운호 한국지능시스템학회 2006 한국지능시스템학회 학술발표 논문집 Vol.16 No.2
본 논문에서는 스팸 대응 시스템의 특징 추출 방법들을 비교한다. 실험 결과는 퍼지추론 방법이 정보획득량, 카이제곱 통계량, 상호정보 방법에 비하여 정확률과 재현율의 결합 척도인 F-척도면에서 월등한 성능을 보여주지는 않는다. 하지만 제안된 퍼지추론 방법은 사용된 특징들의 수에 비례하여 성능이 증가하므로 좋은 특징 추출 방법으로 간주된다. 따라서 본 연구는 무수한 스팸 메일로 고통 받는 전자우편 사용자들을 위한 스팸 메일 필터링 시스템 개발에 도움을 줄 수 있다.