Font Size: a A A

Study And Implementation On Roadside Wireless Sensor Networks Based Traffic Flow Surveillance System

Posted on:2019-08-29Degree:MasterType:Thesis
Country:ChinaCandidate:Z D LiuFull Text:PDF
GTID:2382330563495451Subject:Computer application technology
Abstract/Summary:PDF Full Text Request
The sensing technology of current traffic flow monitoring is mainly based on radar,laser,video,and induction coils.To my best knowledge,traffic flow monitoring systems with the sensing technology above are difficult to meet various detection demands,such as high accuracy,low cost,convenient portability and easy installation,even though they have several advantages.Due to the differences and correlations of detection requirements that extracted from the long-term surveillance and temporary monitoring on traffic flow detection,a traffic flow detection prototype system is implemented based on roadside sensing networks and wireless geomagnetic sensors.Further,some effective methods about high-precision vehicle detection are studied.Firstly,the framework of roadside wireless sensor networks based traffic flow surveillance system is analyzed,and the architectures of hardware,software,and work flow are described in details.Secondly,the design work of system hardware is carried out by components selection,communication implementation and external circuit design.The system is mainly consisted of JN5139 wireless chip and FXOS8700 CQ sensor.Meanwhile,several application programs are realized and compiled,such as data frame structure,node drive program,wireless communication and signal data collection.Thirdly,a few of algorithms are constructed for vehicle detection,drift compensation,speed and length estimation.Including an adaptive geomagnetic baseline drift compensation method,a dynamic threshold five finite-states machine algorithm,and a moving median speed estimation algorithm.Finally,numerous experiments are conducted to evaluate the performance of the proposed system.According to the test results,the proposed system can realize vehicle detection,speed estimation and vehicle classification.And the experiments reported 95.06% detection accuracy,92.45% speed estimation accuracy,and 93.48% classification accuracy.Roadside wireless sensor networks based traffic flow surveillance system has many characteristics,such as low cost,easy deployment,strong adaptability,as well as cooperative perception.The system can be installed not only for the long-term surveillance but also for the temporary traffic monitoring,especially using in work zone safety.Moreover,the effort of the dissertation will pave the way for high-precision traffic flow detection systems based on wireless geomagnetic sensors.
Keywords/Search Tags:Intelligent transportation system(ITS), traffic flow detection, wireless sensor networks, magnetometer sensors
PDF Full Text Request
Related items