Research On Vehicle Recognition And Vehicle Speed Based On Microseismic Big Data Detection Algorithm

At present,traffic data acquisition is a prerequisite for the construction of intelligent transportation system.And the vehicle identification and detection part is a very important part in the process of traffic data collection.The method based on vehicle identification and detection can effectively and conveniently obtain real-time traffic data,and then analyze and study the real-time traffic data.More useful parameters(such as vehicle flow and speed,etc.)are extracted from traffic data to provide data basis for intelligent transportation analysis.According to the real-time data,the intelligent transportation system uses the vibration signal data for analysis and processing.At the same time,develop appropriate management plans and measures for the traffic management control center.Finally,traffic operations will become more information and automated.At first,traditional identification conventional vehicle detection method outlined,but also introduces the principle of a sensor detected vehicle identification microseismic.Secondly,the paper analyzes the characteristics and transmission modes of vibration signals produced by automobile vibration,and takes the big data of micro earthquake as the premise,a multi-vehicle recognition algorithm and a vehicle speed determination algorithm are established.The multi-vehicle recognition algorithm mainly uses the mean filtering method to filter the data,and then continuously updates the reference value and threshold value,and finally uses the state machine to identify the vehicle.At the same time,RNN algorithm is used to solve the boundary value of the number of vehicles output from the vibration signal.The vehicle speed determination algorithm mainly uses the fixed distance between two adjacent sensors and the peak time corresponding to the vibration signal generated by the vehicle passing each sensor to solve the vehicle’s traveling speed.Finally,this paper selects more representative 6 sets of data signals for multi-vehicle identification analysis and verification,and 6 time periods for multi-vehicle identification verification.At the same time,traveling speed of the vehicle is verified.The multi-vehicle recognition algorithm uses the recognition precision to measure the correctness of the algorithm.The feasibility of vehicle speed decision algorithm is to measure the value of the algorithm by means of mean absolute error.By experiments in a multi-car recognition accuracy than 96.07%,and the average absolute error of vehicle speed is within ±1.73km/h.The feasibility of the algorithm is proved.