차세대 스마트 해양 교통 안전을 위한 고신뢰·고효율 LED 등명기 연구

다국어 초록 (Multilingual Abstract)

Aids to Navigation (AtoN) are critical infrastructure for safe vessel operations, and marine lanterns constitute the largest portion of these facilities. Although light sources have advanced rapidly―from Fresnel-lens optics to solar-powered unmanned automation and, more recently, high-efficiency long-lifetime LEDs—research on fundamentally new optical structures beyond lens-based systems remains limited. This study proposes and validates a lensless double-reflection optical architecture to address the optical loss, structural vulnerability, and high manufacturing cost of conventional Fresnel/aspheric-lens marine lanterns. In addition, compact all-in-one marine lanterns integrating a solar module, battery, and lantern body were developed with a highly reliable housing designed to withstand impact, salinity, moisture, and dust, enabling long-term unmanned operation. Four prototypes were fabricated and certified through accredited testing: all-in-one 3 NM, 5 NM, and 9 NM lanterns, and 11 NM lantern In case of 11NM lantern, it is suitable for adaptive control, utilizing AI vision-based technology to facilitate real-time intensity adjustment under dynamic conditions. The all-in-one 3 NM lantern met comprehensive environmental and EMC requirements, including salt spray, IP67 waterproof/dustproof performance, thermal endurance, and low-temperature operation. It achieved 75 cd central luminous intensity with a vertical distribution of -2.2° to 2.3°, while exhibiting high energy efficiency (1.56 W power consumption, 2.96 mA no-load current). The all-in-one 5 NM lantern enhanced optical output via reflector coating, delivering 199 cd with 1.63 W consumption and 2.47 mA no-load current while maintaining the same form factor as the all-in-one 3 NM lantern. Environmental testing revealed that the all-in-one 9 NM lantern maintains structural integrity under salt spray and IP67 conditions. In addition, the system's compliance with KS X 3140:2014 verifies its operational stability regarding electromagnetic compatibility. Optical characterization of the integrated lantern confirmed a fixed luminous intensity of 1,657 cd with a stable vertical distribution ranging from -1.7° to 2.1°. Electrical evaluation indicated a power consumption of 10.36 W and a no-load current of 2.47 mA. The lantern featured a high-voltage solar charging module containing ETFE-coated solar panels to support long-term autonomous operation. Furthermore, the 11 NM lantern passed IP66 and salt spray tests, achieving 8,447 cd approximately 2.7 times higher than the 3,080 cd standard. It exhibited excellent uniformity (≥50% average intensity over 360°) and a vertical spread of -3.0° to 1.9°, while significantly reducing power consumption to 47.12 W, a 34% lower than the 72 W standard. For intelligent operation under low visibility, an AI vision system based on DehazeNet running on NVIDIA Jetson Nano maintained 12 ∼ 18 FPS at 720p in a 200 m fog simulation tunnel, enabling real-time haze estimation and automatically increasing lantern intensity by approximately 2.0 under dense fog conditions, effectively compensating for the conspicuity loss of constant-output lanterns. Overall, the developed lensless double-reflection marine lantern all-in-one 3 NM, 5 NM, and 9 NM lanterns, and 11 NM lantern Showed the advanced performance compared to the conventional marine lanterns including in environmental durability, optical characteristics, electrical/thermal stability, EMC, and energy efficiency, With the feasibility of AI the feasibility of AI-driven adaptive intensity control. The results indicate a practical pathway toward high-reliability smart AtoN systems capable of actively responding to sea fog and rapidly changing maritime weather, with strong potential for future expansion into integrated smart AtoN platforms.

목차 (Table of Contents)

• 1 서론 1
• 1.1 해상용 등명기 1
• 1.2 국내·외 해상용 등명기 운영현황 6
• 1.3 국내·외 해상용 등명기 연구동향 10
• 1.4 연구 목적 14
• 2 이론 20
• 2.1 국내 해상용 등명기 표준 규격 20
• 2.2 이중반사형 구조를 갖는 등명기 28
• 2.3 광학시뮬레이션 30
• 2.4 태양전지 32
• 2.5 반사판 코팅 35
• 2.6 LED 방열 39
• 2.7 딥러닝 기반 해무 측정 기법 41
• 3 연구방법 45
• 3.1 해상용 LED 등명기 제작 45
• 3.1.1 해상용 LED 등명기 설계 45
• 3.1.2 해상용 LED 등명기 광학 시뮬레이션 46
• 3.1.3 해상용 LED 등명기 외함 제작 47
• 3.2 국내 해상용 LED 등명기 측정 방법 48
• 3.2.1 광학계 측정 51
• 3.2.2 섬광기 측정 52
• 3.3.3 등명기 환경 특성 측정 53
• 4 연구결과 54
• 4.1 포물선 초점 위치의 LED를 이용한 해상용 일체형 3 NM 등명기 연구 54
• 4.1.1 해상용 일체형 3 NM 등명기 광학 시뮬레이션 55
• 4.1.2 해상용 일체형 3 NM 등명기 설계 56
• 4.1.3 해상용 일체형 3 NM 등명기 성능 결과 60
• 4.2 해양플랜트에 적용 가능한 해상용 일체형 5 NM 등명기 연구 67
• 4.2.1 해상용 일체형 5 NM 광학 시뮬레이션 68
• 4.2.2 해상용 일체형 5 NM 등명기 설계 70
• 4.2.3 해상용 일체형 5 NM 등명기 성능 결과 71
• 4.3 기상관측 부이에 적용 가능한 해상용 일체형 9 NM 등명기 연구 74
• 4.3.2 해상용 일체형 9 NM 등명기 광학 시뮬레이션 75
• 4.3.1 해상용 일체형 9 NM 등명기 설계 77
• 4.3.3 해상용 일체형 9 NM 등명기 성능 결과 82
• 4.4 위험 기상 대응 및 자동 광도 제어 기능이 적용된 AI 비전 기반 해상용
• 분리형 11 NM 등명기 연구 89
• 4.4.1 해상용 분리형 11 NM 등명기 광학 시뮬레이선 90
• 4.4.2 해상용 분리형 11 NM 등명기 설계 92
• 4.4.3 위험 기상 대응을 위한 AI 비전 영상처리 기법 연구 94
• 4.4.4 해상용 분리형 11 NM 등명기 성능 결과 102
• 5 결론 111
• 참고문헌 115