Key Technology Research And Typical Application Of Vector Extraction Based On Perception Data

With the rapid development of the Internet of Things technology and its related applications(such as smart cities,sensing China,intelligent transportation,emergency management,intelligent security,etc.),how to effectively store,query,and analyze the massive Internet of things perception data has become a problem to be solved.Based on the idea of vector-based data extraction and storage of the Internet of Things,vector data is extracted and stored.Not only will the data store update frequency and amount of data be reduced,but it will also reduce the amount of data for query and analysis operations.With the help of the PageRank algorithm,the road state vector is introduced to implement dynamic path planning.The design is based on the spatial diversity theory of crowd evacuation to achieve efficient evacuation of large-scale activities.Based on the vector extraction of trajectory data,this paper analyzes studies and improves the dynamic path planning and efficient evacuation of large-scale activities:(1)Based on the storage of sensing data,this paper proposes a vector storage method.For the state vector extraction method of spatial-temporal data with a path,this paper selects the Bessel curve fitting algorithm,designs the state vector representation,matching and updating algorithm,extracts the vector of the sensing data and obtains the state vector sequence of the monitoring object.Based on the vector data stored method of sensing data,an efficient data query is proposed to achieve efficient data analysis.Through the vector extraction and storage of sensing data,not only the amount of stored data and the frequency of data update are reduced,but also the efficiency of data query and the accuracy of interpolation query data are improved.(2)In this paper,by using the PageRank algorithm idea,using the new PageRank algorithm of the link state vector,the congestion coefficient value CR of the intersection is calculated.The CR value of the intersection reflects the congestion state of the road section.Then we consider the distance of the road section,design the heuristic function of the A* algorithm based on the distance and congestion state,and use the A* algorithm to achieve real-time dynamic route planning and reduce road section congestion.Compared with the navigation path planning of Baidu and Autonavi map,the path planning method designed in this paper tends to select the current road section with small congestion coefficient,achieve the expected navigation effect and reduce traffic pressure.(3)In view of the current research status of large-scale active crowd evacuation schemes at home and abroad,this paper establishes an evacuation plan model based on the spatial diversity theory of crowd evacuation design.According to the established network model,plan the crowd evacuation plan,use the single point path planning studied in this paper to carry out the evacuation route planning,and formulate a reasonable,accurate and efficient evacuation plan to minimize the evacuation time and minimize traffic congestion.Taking the large-scale activities in the National Stadium as an example,experiments are conducted,and the results of the experiments showed that the evacuation model proposed in this paper greatly improves the evacuation efficiency.