Research On Credibility Evaluation Method Of Multi-source Heterogeneous Information Fusion Based On Roadway Model

With the development of the Internet of Things technology,many wireless sensor networks have been used in environmental safety monitoring,so that when a safety accident is about to occur,the safety monitoring center can make an early warning based on the information collected by the sensor;but when a safety accident actually occurs,many times the safety monitoring center does not make early warning or does not warn in time,thus leading to casualties.One of the reasons why the monitoring system does not operate normally is that the sensor data input to the monitoring system is incorrect,which causes the monitoring system to fail to recognize the real situation in the environment at the time,and thus does not issue an early warning;therefore,before the monitoring system processes the data,It is necessary to evaluate the credibility of the sensor data first,and eliminate those untrusted data to reduce the probability of false alarms and missed alarms in the monitoring system;so this thesis has launched a research on the credibility evaluation method of sensor data.This thesis first analyzes and researches the data quality evaluation technology involved in big data processing,including the basic concepts,the dimensions and assessment methods involved in different fields.Then the relation and difference between data quality and data credibility are analyzed.After that,the data credibility evaluation framework and commonly used evaluation methods are studied;the key steps in data credibility evaluation are analyzed,and the advantages and disadvantages of commonly used evaluation methods are summarized.Then,based on the previous summary,this thesis specifically researched the sensor data credibility evaluation method in the data credibility evaluation.First,the definition of the credibility of sensor data is summarized,and then various reasons for the unreliability of sensor data are analyzed.Based on these reasons,the shortcomings of existing sensor credibility evaluation methods are discovered.Subsequently,in view of the shortcomings of the existing single-sensor credibility evaluation methods,a credibility evaluation method based on data trends is proposed;it solves the problems of existing methods or problems that cannot be evaluated online or are difficult to obtain due to prior probability or are excessive.Depend on a certain point,leading to unsatisfactory evaluation effect or small application range.Then,for multi-sensor data credibility evaluation,a multi-sensor data credibility evaluation model is proposed,and the calculation method of each parameter in the model is introduced,including trend correlation,mean aggregation,and direct data availability.Reliability,credibility of heterogeneous fusion.The model uses the above four parameters to fuse the historical data of single sensor data,fuse the data of multiple sensors at the same time,and use the spatiotemporal correlation of these data and the connection between various attributes to comprehensively evaluate the sensor data;to solve The existing methods have low accuracy in credible evaluation of data when the sensor is occluded.Finally,the reliability evaluation model based on multi-sensor data is used to evaluate multiple groups of data,and the evaluation effect can achieve the desired goal.