Introduction
Automating tunnel entrance lighting is a critical balance between driver safety and energy efficiency. When entering a tunnel on a bright day, a driver's eyes are adapted to high outdoor light levels. If the entrance is insufficiently lit, it appears as a "black hole," causing a dangerous period of temporary blindness. Conversely, over-lighting wastes significant electricity. Traditional Automatic Tunnel Lighting (ATL) systems rely on vertical illuminance, but research shows this does not always correlate with the actual luminance—the brightness within the driver's field of view. This paper proposes a system that directly measures outdoor luminance to optimize tunnel lighting and eliminate the "black hole" effect.
System Architecture and Methodology
The proposed system shifts from indirect illuminance measurement to direct sensing of the driver's visual environment. As illustrated in Figure 1, the system utilizes luminance detectors positioned at the access zone to feed data into a measurement and control unit.
Key technical features include:
- Luminance Detectors: Two sensors measure the average luminance of a 20° circular field of view centered on the tunnel entrance from 150 meters away, covering a range of 16 to 4500 cd/m².
- Signal Transmission: To prevent signal degradation over long distances (several kilometers), luminance readings are converted into voltage-frequency signals.
- Stability Controls: A 15-second time constant prevents flickering from momentary shadows. A 20% hysteresis between "turn-on" and "turn-off" thresholds stabilizes the system during rapid weather changes. Additionally, a minimum-on timer of 20 minutes prevents frequent cycling that shortens lamp life.
- Fail-Safe Logic: In the event of power or microprocessor failure, the system defaults to "Full Brightness" (Sunny 1 mode) to ensure driver safety.
Operational Logic and Results
The system categorizes environmental conditions into specific tiers, each with corresponding dimming ratios. This mapping ensures a smooth transition for the human eye while maximizing energy savings.
As shown in Figure 2, the system defines categories such as Sunny (晴天), Cloudy (曇天), Daytime (昼間), Night (夜間), and Late Night (深夜). For instance, in "Sunny 1" conditions, entrance lighting is maintained at 100%, while "Night" levels drop to a 50% base. The use of a base luminance setting ($L$)—typically 3000 to 6000 cd/m²—allows for fine-tuning based on local geography or road conditions after the tunnel is opened.
Empirical Basis
The system's design was informed by a one-year longitudinal study of outdoor luminance and illuminance (both horizontal and vertical) at actual tunnel sites. This empirical evidence confirmed that the correlation between illuminance and luminance was insufficient for precise control, necessitating the direct luminance-based approach to accurately mirror driver eye adaptation.
Conclusion and Limitations
The developed hardware and control logic successfully mirror driver visibility needs and maintain operational stability through hysteresis and time constants. However, future work should address long-term field performance data regarding energy savings and accident reduction. Additionally, the fixed 20° field of view may need adjustment for complex tunnel geometries (e.g., sharp curves), and maintenance protocols for sensor lens fouling (dirt/snow) remain to be detailed to ensure long-term accuracy.
Summary updated