In vehicle navigation systems,the aim of path planning techniques is to obtain the appropriate route for drivers.This problem and associated methods are very important in intelligent transportation system.Finding the best route in complex urban streets is the typical nonlinear optimization problem.In these years,swarm intelligence algorithms are widely utilized in such issues and have shown some advantages.In this paper,some blocks in Tangshan city are modeled as nodes with related longitudes and latitudes. An improved fruit fly algorithm is designed for solving path planning task.Fitness function is adjusted for discrete optimization and some operators in genetic algorithm are introduced.Simulation experiments show that the algorithm designed in this paper is able to obtain a better route from the starting node to the destination.The cost of driving is reduced.