Traffic Information Monitoring Based On WSN And Data Fusion

In intelligent transportation system, traffic information detection is the important component parts. It sends accurate and real-time traffic data which is detected by various sensors to traffic control center, then traffic control center deal with these data and send corresponding instructions by traffic lights or other traffic control methods, so that realize the intelligent control of purpose.Wireless sensor network used in traffic information detection has broad development prospects because of four good points: high detection precision, easy installation and maintenance, less environmental impact and data fusion. This paper has researched traffic flow detection, vehicle classification, and vehicle speed detection by WSN and data fusion, and presents an area traffic volume collection model based on WSN. It verified that this model is reasonable and feasible.First, a new algorithm is developed with binary proximity magnetic sensors and back propagation neural networks. In this scheme, we use the low cost and high sensitive magnetic sensors that detect the magnetic disturbance when vehicle pass over it and estimate vehicle length with the geometrical characteristics of binary proximity networks, and finally classify vehicles via neural networks. The inputs to the neural networks are the occupancy empty ratio sequence and vehicle length, and the output is seven classified vehicle type. It verified that this scheme enhances the vehicle classification with high accuracy and solid robustness.Secondly, an adaptive Kalman filter, Kalman-PE, is presented to vehicle speed estimation. To higher accuracy, we estimate the measurement noise with an improved patch estimation algorithm, which can track the changes of the measurement noise and estimate it in real time. Using the Kalman-PE, a new method, making the Kalman Filter under unknown measurement noise condition valid, is proposed. Finally, a simulation test is conducted for contrasting the capability of Weighted Average Filter, Kalman Filter and Kalman-PE Filter, which shows Kalman-PE Filter effectiveness.Thirdly, on the basis of existing methods, this paper presents a method of estimating the traffic volume of inter-section and boundary section of traffic zone. This method uses the WSN to detect the traffic volume, and use Kalman Filtering to process the data. By the experiments, this method has improved the precision of traffic volume, and also can obtain the state of change of the traffic zone.At last, we summarize the work of this paper, point out the next step's research orientation.