Intelligent Transportation Demonstration System Based On Electromagnetic Navigation And Visual Detection

With the rapid development of economic and society, car ownership has increased considerably. Intelligent Transportation System (ITS) is a real-time efficient traffic management system that combines advanced electronic information technology, which has been paid great attention to by domestic and foreign research institutions. Being a basic platform for the demonstration and simulation of key techniques, prototype system based on scaled model has been an important method in ITS research. This thesis adopts scaled city sandbox and car models to design and implement a dynamic demonstration system of ITS, combined with high-speed camera, ZigBee wireless communication technology, electromagnetic navigation technology and wireless sensor network positioning technology.First of all, on the basis of summarizing current research, an overall architecture of demonstration system has been designed with short-range wireless communications technology, and the hardware design of concept car models has been given out. For the convenience of setting system parameters and monitoring system status, this thesis has designed a graphical display interface program on.NET platform. Wireless data communication protocols have been designed to achieve communication between PC and other devices. ZigBee protocol has been selected to establish the communication network of the ITS demonstration system by considering that ZigBee has characteristics of self-organization, short range and so on. Experiment has been designed to test the network.According to the actual environment, electromagnetic navigation is employed to navigate the concept car models. PD control algorithm is adopted to correct the bias of car models. In order to achieve dynamic path setting function, RFID tag is introduced to build digital identity of intersections that can guide the car models to turn at the intersections and select the right path. A C# PC software has been developed to monitor information of vehicle status and traffic conditions. Car models perceive speed information with infrared sensors and distinguish the state of intersection traffic signal light by detecting infrared frequency modulation signal. High-speed camera is introduced to establish an Intersection traffic flow detection system that is used to recognize car icon with the image feature extraction algorithm. This thesis adopts pulse numbers and point correction to monitor the car position and carries out validation and analysis.On analyzing the basic concept and characteristics of chirp spread spectrum technology (CSS), this thesis has done a lot of testing on the CSS node localization performance. Combined with wireless sensor network localization and tracking algorithm, this thesis achieves the car trajectory tracking by using the least square localization algorithm and the extended Kalman filtering algorithm (EKF). And then I introduce the trajectory information control method to improve the tracking accuracy, achieving the similar architecture as GPS indoor positioning. The experimental results indicate that, the algorithm proposed in this article can obtain fairly accurate trajectory and has a strong robustness for the environment. The tracking accuracy can reach 0.179 meters at the velocity of 0.4m/s with a length of 2 meters square trajectory.Finally, according to the design of this thesis, some summaries are made and the direction of subsequent improvement is indicated.