Dynamic Path Program Research In Intelligent Transportation System

Optimal path program is a problem people certainly meet with when going out and it is an important application area in ITS. This paper uses partial region of Zhenjiang City as research object for path program, set up road net model of this area on Mapinfo and uses Map Basic for secondary development to realize road net topology and establish experimental platform for the research of path algorithm.Road net model includes all the static information such as length of roads and topological relation and so on. Path program problems based on static information can be solved by traditional path algorithm, Floyd algorithm can work out the most complete solutions among these algorithm. Through the experiment of matrix iteration on weighted network model, this paper proves the minimum weight matrix calculated by Floyd algorithm includes the length between all the nodes in network. Afterwards path program experiment is conducted on the road net model of partial region of Zhenjiang City by Floyd algorithm and the experimental result is compared with navigation software to prove the validity of Floyd algorithm. The computational result is used as the theoretical value for ACO.ACO can get the optimal path between two nodes through imitating ant’s foraging behavior. It is very flexible and combinative. However ACO has the shortcomings of premature convergence easily and falling into the local optimum. So this paper experiments with parameters of ACO, draws the conclusion of the effect on algorithm performance by every parameter and summarize the range of every parameter. Aiming at the disadvantages of ACO, its pheromone concentration range, pheromone update method as well as probability choose method are improved to develop search performance of ACO.According to the feature of Floyd algorithm and ACO, this paper puts forward with FACO Combined with the two algorithm. The improved Floyd ant colony is intelligent and is able to imitate people’s travel in daily life to realize traffic jam simulation. During every dynamic simulation period, optimal path program is carried on by improved ACO. Dynamic path planning experiments show that improved ACO can adjust the travel strategy based on real-time road conditions and make a plan to avoid traffic congestion. It is beneficial for solving the problem of traffic congestion.