Abstract:

          We deployed velocity coils at some suitable locations and employed RBF neural network to predict the velocity between two adjacent coils for the implementation of such intelligent control technology as light following vehicle. We also constructed a tunnel characteristic based tunnel stopping sight distance model, which determined a realistic and safe light length. We eventually presented the idea of smart illumination control.The analysis of practical cases shows that energy saving rate is more than 90% when the traffic throughput is lower than 3 000 pcu a day.

Key words: tunnel illumination, RBF neural network model, tunnel stopping sight distance, intelligent control