Research On Discrimination System Of False Fire Reporting Based On Internet Of Things And CEP

Although Fire monitoring system based on Internet of Things can receive the fireinformation of networking units and running status of fire fighting facilities inreal-time, but there still exists problems of false fire reporting caused by inefficientmanagement or insufficient maintenance of fire fighting facilities.The false firereporting has reduced the fire warning capability of Fire System and this system facedproblems of data mining,data analysis and discrimination of fire reporting in real-time.To solve problems of false fire reporting, research on internet-based real time firemonitoring and on-line data analysis is proposed. The concept of Complex EventProcessing is introduced into the remote fire fighting security monitoring systems.Complex Event Processing technology is a memory-based data streams analysis tools,which can process massive data as data streams and can be integrated into the firereporting system in the form of components, with low coupling and high flexibility.The thesis employed complex events processing engine (Esper) integrating withMyEclipse to develop a prototype system.As a result, the entire process including datastream processing from IOT, events and rules modeling and on-line discrimination of falsefire reporting is presentedin the system.The thesis’s main researchcontents are as follows:Firstly,the background of this dissertation is described, the Internet of Things,Fire system based on Internet of Things and Complex Event Processing are reviewedand analyzed, focusing on large data streaming data calculation mode, complex eventprocessing language EPL, complex event processing engine Esper, architecture ofEsper engine,its event listeners and event window mechanism processing mode.Secondly,infrastructure network and the overall architecture of traditionalfirenetworking monitoring system are studied.On this basis, the system architecture ofDiscrimination of False Fire Reporting Based on Internet of Things and ComplexEvent Processing are proposed, that using a layered architecture model, includingdata collection and network communication layer, data preprocessing layer, eventsmonitoring based on complex event processing layer and the information release layer. Finally,the false fire alarm type and the processing of false fire complex eventbuilding is described in detail, with an emphasis on the definition of fire false atomicevents, abstract events and complex event rules.Then Based on the description of falsefire complex event,transformed fromsuspicious event algebra definition into EPLquery by developing a Esper-based console desk process application.Then by meansof computational experiments,realized the receiving of the operating state datas of firefacilities in real-time and convert such datas into streams,then the thesis simulatedthese data stream into events processing engine andcarried out analysis on this datastreams and finally presented discrimination of false fire reporting in real time.