Dynamic Brightness Adjustment of Tunnel Lighting Based on ETC Transaction Data

In recent years, with the continuous extension of the highway network, the power demand of the highway lighting system has also become increasingly serious, and the energy-saving problem brought by tunnel lighting has gradually attracted everyone's attention. As a special section of the highway network, the lighting facilities in the tunnel need to run uninterruptedly for 24 h to solve the “black hole phenomenon” and “zebra effect” in part of the driving process, which has undoubtedly become the most important expense in the highway operating costs. Moreover, excessive brightness distribution runs counter to the national policy of “carbon neutrality” and “energy conservation and emission reduction”. In order to improve the lighting efficiency and thus minimize the operating costs, this paper proposes a dynamic adjustment of tunnel lighting brightness based on ETC transaction data. This method uses the ETC gantry transaction data in front of the tunnel as input to train the bidirectional LSTM (BiLSTM) model, establishes a long-term and short-term tunnel traffic flow prediction model, and realizes the dynamic adjustment of the brightness of the tunnel lighting system under the premise of ensuring traffic safety, thereby reducing energy consumption by 35%-40% and minimizing the energy loss of highway tunnels.