Vehicular sensor networks (VSN) event region detection method based on Dempster-Shafer (D-S) evidence theory

Abstract

The invention provides a VSN event region detection method based on a D-S evidence theory. The method can be used for effectively detecting an event occurrence region under the condition that a priori knowledge is absent, highly mobile network topology is aimed at and complex changing road traffic scenes are handled with. The method comprises initialization and maintenance of VSN scenarios, a road division sub-cell event monitoring probability module, a road division sub-cell event occurrence probability module, an event occurrence confidence module, an evidence combination conflict computation module, an event region judgment module and an event detection trigger module.

Description

technical field [0001] The present invention relates to the field of collaborative data processing and event monitoring of vehicle-mounted self-organizing wireless sensor networks, and more specifically, relates to a new method that uses D-S evidence theory to describe the inconsistency of data fusion in vehicle-mounted sensor networks, thereby detecting the inconsistencies in the road condition environment. method of the event area. Background technique [0002] With the popularization of automobiles and the development of sensing technology, wireless communication and other technologies, by installing sensor node devices in vehicles driving on the road and interconnecting them through wireless communication, they are self-organized into wireless vehicle sensor networks; vehicle sensor networks can realize vehicle Collaborative perception, processing and transmission of information such as various road traffic conditions in urban areas is an important means and way to reali...

AI Technical Summary

Problems solved by technology

[0005] The implementation method of event area detection technology should be related to specific applications. In the vehicle-mounted sensor network environment, due to the high mobility of vehicles and the complexity of urban road traffic conditions, the network topology changes rapidly, and event monitoring cannot be simply set. The sensor data threshold is realized; therefore, the general wireless ad hoc network and wireless sensor network event area detection methods cannot be well applied to this application scenario

Some researchers have proposed a method of using artificial intelligence for vehicle self-organizing networks, through local node collaboration, using machine learning, support vector machines, Bayesian neural networks, or using hidden Markov models, etc., to extract event features and Classify and judge the probability of event generation to realize event monitoring; these event monitoring methods can effectively monitor road vehicle-related events in the vehicle ad hoc network environment, but these methods need to give prior knowledge and other information in the specific environment of road vehicles in advance, these Since road traffic and vehicle driving are greatly affected by the natural environment, road terrain characteristics and human factors, the rationality of the value of prior knowledge directly affects the performance of the event monitoring system; in addition, these studies are mainly devoted to the judgment of events, rather than Events generate area extent and location information, in fact, the detection of event areas poses technical challenges due to the high mobility of vehicles in the vehicular network environment

Images