AMR-Sensor-based Critical Information Detection For Intelligent Transport Systems

Along with the rapid growth of social economy, road congestion, traffic acci-dents, environmental pollution and energy shortages have become a traffic-relatedproblems faced by cities in China in recent years. The urban area is under growingtraffic problems plagued without exception, and has seriously hampered its furtherdevelopment. To ease the urban area’s traffic congestion, reduce traffic accidents, saveenergy, improve road capacity has become a top priority the development of the city.Therefore data by using advanced detection means to realize the effective, accurate,real-time traffic information parameter detection, become intelligent transportationsystem (ITS) optimizing traffic organization, improving road traffic capacity, the keyto ease traffic jams. Currently the fixed traffic detection technology is the most widelyused technology, the traffic flow detection products including the sense coil detector,ultrasonic detector, infrared detector, microwave detector and video vehicle detection.And the sense coil because of its low price, become the most widely used detector.However the installation and repair process is very tedious, and would cause somedamage to roads. Weak real-time, low accuracy is the biggest shortcoming of the de-tection scheme.A wireless network vehicle information detection system based AMR (Aniso-tropic Magneto Resistive) sensors was built,which can effectively avoid the lack ofmost of the detector. The system through the real-time traffic information collectionand statistical traffic information acquired vehicle type, speed, traffic flow, drivingdirections, headway, road share this information for the intelligent transportation sys-tem provides a wealth of critical traffic information parameters. The system mainlyconsists of three functional modules, AMR geomagnetic detector node detector dedi-cated route, intersection or sections of collection centers. Geomagnetic intensity vari-ations caused by the detector node detects vehicles processed by a dedicated routingtraffic information parameter sent to collection centers after corresponding traffic dataanalysis and management software detection algorithm.Secondly, a new fuzzy data fusion algorithm to achieve the discrimination ofthe type of vehicle was used in this system. The algorithm reflects the strong robust-ness, but also solve the problem of external interference led to the base value of fail-ure, to ensure that the high sensitivity of the detector, due to its superior base valueupdate mechanism. The algorithm uses Gaussian membership functions described onthe detection waveform extraction of characteristic parameters, and then combined with the corresponding data fusion formula for fuzzy data fusion, determined accord-ing to the principle of maximum degree detection vehicle belongs models. After test-ing, the system models distinguish accuracy rate as high as98%.In conclusion, the design of the entire system, including the composition of thehardware system and the overall architecture of the software, data communicationprotocols and systems management software user interface are introduced in detail.Also by way of example, according to the frequency of occurrence of the miscarriageof justice models, the distinguish model system performance evaluation of the modeldiscrimination process.