Virtual detector in adaptive traffic signal control system based on support vector machine

Abstract

Any traffic responsive system depends on its ability to sense traffic flow. In traffic engineering, detector data can not satisfy the demand of adaptive traffic signal control system because of its high cost, low coverage area, difficult to update and maintenance. Therefore, in order to obtain more detail of traffic condition information such as traffic volume, mean velocity, link travel time, occupancy, etc. for adaptive traffic control system, as the concept of the software-is-namely instrument, we put forward a technique of Virtual Detector. It is developed, applied virtual instrument technology, traffic flow theory, multi-sensor data fusion technology, database and fault diagnostic system.