Study On Self-adaptive Traffic Control Based On Inductive Vehicle Detection Technology

Intelligent control technique for urban roadway traffic is an important topic in control domain and traffic engineering domain. With the fast development of artificial intelligence (AI), automatic control, computer and communication techniques, many new theories and research achievements have been published in recent year, and their applications in engineering have shown tremendous powers and huge potentials.Due to the heavily nonlinear, stochastic, time-variant and uncertain characteristics, mathematical model of urban traffic system is very difficult, some traditional methods can not get ideal effect. However, fuzzy control can have a good performance by simulating manned decision process without mathematical modeling.In this paper we present a multi-phase traffic light control strategy based on the length of the waiting queue according to fuzzy control theory. In this strategy, the length of queue is regarded as the control object and the length of queue on the contiguous phase is used to determine the timing plan. The information of the length is provided by vehicle detecting card, and the idea of fixed period technology is presented, we also give the design of hardware and software of this card.The main contributions of this paper are:1. By summarizing the characteristics of modern urban traffic control systems adopted widely in cities home and abroad, a modularized intelligent traffic control system which is suitable for the development of Chinese cities is presented. It can improve traffic efficiency according to the waiting queue of the intersection.2. The idea of fixed period technology for the vehicle detecting card is presented. Compared with traditional detecting productions adopted frequency divided method, system's stability under poor working conditions is improved, and the 0.01% detection accuracy is achieved. The design of hardware and software of this card is given.3. We develop a macroscopic dynamic deterministic traffic flow model for control.Furthermore, a multi-phase traffic light control strategy is proposed, which determines the timing plan based on the length of the waiting queue of every phase. Simulations show its good effect on shortening delay time.