Research And Design Of Intelligent Traffic Information System For Single Intersection

With the development of urbanization,the number of motor vehicles has been increasing rapidly,which makes the urban road traffic more and more crowded,and the congestion mainly occurs at the road intersection.Therefore,the traffic signal management of a single intersection plays an important role in the effective operation of road traffic.Combining the actual road traffic conditions,the paper applies the machine vision-based traffic parameter acquisition technology and the fuzzy neural network control algorithm to the traffic management system.The intelligent traffic information system can be used to map traffic lights in all directions according to real-time intersection traffic flow information.Intelligent management to guide vehicles to pass.Firstly,the paper analyzes the demand of intelligent traffic information system,designs the overall system design according to the actual functional requirements,introduces the key technologies in the system,and focuses on the traffic parameter acquisition technology and fuzzy neural network algorithm in the intelligent traffic information system.In order to obtain the traffic parameter information of the road junction,OpenCV machine vision technology will be used to detect the three traffic parameters of the queue length,traffic volume and vehicle speed of the intersection:The vehicle queue length detection is mainly to detect the vehicle motion detection and presence detection of the lane image to determine the position of the vehicle queue tail.The lane parking line is used as the queue head position of the vehicle,and then the camera calibration technology is used to realize the conversion of the vertical pixel distance to the actual three-dimensional space.Vehicle flow and speed detection is done by setting virtual coil between parking line and crosswalk in the image,using Gao Si mixed model in virtual coil to establish background,background difference method to detect moving vehicles.Finally,according to the status of virtual coil to obtain traffic flow information and vehicle speed detection.Based on the acquisition of accurate traffic parameters,the fuzzy control algorithm is applied to the traffic signal control system of road intersection.Through the research and analysis,it is found that the algorithm is subjective,so the neural network algorithm with self-learning function will be incorporated on the basis of fiizzy control system to form an intelligent traffic information system with fuzzy neural network as the core.The collected traffic parameters are used as input parameters of the fuzzy neural network algorithm.The control rules are automatically generated according to the actual collected sample data,and the traffic control strategy of the transit phase is obtained by fuzzy inference of the input traffic parameters.In order to verify the effectiveness of the algorithm,the performance control of timing control,fuzzy control and fuzzy neural network are simulated under different traffic flow conditions,with the average vehicle delay as the technical index.Through the scheme design and algorithm research,the hardware platform and software design of the system are completed at last.The embedded industrial control computer is chosen as the main controller of the system,and the video stream of the intersection is collected through the intersection webcam,and then the three traffic parameters of the waiting phase are obtained by processing and obtaining the three traffic parameters.The traffic parameters are processed by fuzzy neural network algorithm,and the green light time of the waiting phase is obtained.Through the LoRa wireless communication module,the control strategy frame data including the traffic phase and the green light time length is transmitted to the traffic light control module;the ARM processor parses the transmitted frame data,and finally realizes the lighting out of the traffic light module and the management of the timing.The intelligent traffic information system researched in this paper carries on the function test from the machine vision multi-traffic parameter detection,the traffic signal light control strategy,the traffic light module timing check and so on.The test results and experimental data show that the system can realize the intelligent management of traffic lights at single intersection,optimize the signal timing of traffic junctions,and achieve the expected research goal.