Abstract:

Aiming at the two problems in the traffic signal control system of SCATS, poor realtime optimization capability for the phaseoffset and having few options, the generation technique of the adaptive control strategy for the SCATS was investigated in combination with the probe vehicle data. The realtime phaseoffset optimization capability and the accuracy of the input data of the model were improved through enhancing the accuracy of the probe vehicle link travel time evaluation by improvement of the map matching algorithm. Using a fuzzy neural network and a multiple attribute decisionmaking technique, the traffic signal control parameters were generated adaptively based on the realtime detector and probe vehicle data. The suggested model was validated using a SCATS control region of Shanghai city as an example of the simulation traffic network, the data sampled by the detectors of SCATS and the probe vehicles as the input of simulation, and the Paramics V6 as the simulation platform. The results showed that it can improve the SCATS signal control efficiency.

Key words: engineering of communication and transportation system, probe vehicle, traffic control, control strategy, SCATS, fuzzy neural network, multiple attribute decision-making