Design Of Parking Space Detection System Based On Geomagnetic Vector Information

With the advancement of urbanization,people are getting richer.More and more people are buying cars to get around.As the number of motor vehicles increases year by year,the number of parking spaces cannot meet the parking demand.The problem of "parking difficulty" has become more serious.To solve this problem,people turn their attention to sensor technology and Internet of Things communication technology.Using magnetic sensor to collect geomagnetic information to detect parking spaces is becoming more and more popular.Especially,it has a great advantage in the detection of outdoor parking space and roadside parking space.With the rapid development of NB-IoT technology,it is widely used in the field of Internet of Things.Aiming at the problems of low detection accuracy and poor real-time data transmission in the existing geomagnetic parking space detection system,this paper will design a parking space detection algorithm to improve the detection accuracy and use NBIoT to improve the real-time communication.The main work done in this article can be summarized into the following four aspects:(1)The design of parking space detector node is completed.Based on low-power single-chip microcomputer,three-axis magnetic sensor and NB-IoT module,the hardware circuit and embedded software are designed.(2)A parking space detection algorithm based on geomagnetic vector information is designed.By studying the influence of vehicles entering and leaving parking spaces on geomagnetic vector,a multi-state machine algorithm is designed to detect parking spaces.In addition,the method of adjacent parking space judgment based on DS evidence theory is proposed,which can reduce the misjudgment caused by vehicles in adjacent parking spaces and improve the overall accuracy of detection.(3)A parking space management system on Web is designed.The system provides users with a friendly human-computer interaction interface,which is convenient to manage parking spaces and observe parking data.(4)The system is tested.In order to test its function implementation and performance,the system is built in the real scenario.The objects of function implementation test include detector node and parking management system.The performance test includes stability test,real-time test and detection accuracy test.After three months of testing,the system can run stably,and the final detection accuracy rate is over 97%.