Design And Implementation Of Intelligent Building Monitoring Data Processing System Based On Spark

With the development of Intelligent building technology,more and more sensors device are integrated into the intelligent buildings.Intelligent buildings,security systems,building automation,fire protection systems,such as a variety of sensor data integration,abnormal monitoring,fault alarm put forward higher requirements.The existing building monitoring data processing system problem gradually highlights.Mainly reflected in the anomaly detection rule can not support the system administrator to expand and flexibly configure the service according to the business expansion;The real-time detection of abnormal data is difficult to ensure in the case of sharp increase of intelligent building sensing equipment data;fault alarm data can not be timely and effectively notified to the designated equipment operation and maintenance personnel,resulting in failure can not be timely and effective treatment.In view of the above problems,this paper designs and implements the intelligent building monitoring data processing system based on spark,through the research and practice of open source stream data process framework SparkStreaming,kafka,Flume and function programming language scala.The system mainly includes functions of building information maintenance,equipment information maintenance,rule information maintenance,alarm information maintenance,user information maintenance,sensor data real-time acquisition,transmission and abnormal monitoring.The test and application of the system show that the system is feasible and effective.Compared with other similar monitoring systems,the system designed and implemented in this paper has the following characteristics:1)Anomaly detection rules are divided into simple rules and complex rules.Complex rules are composed of simple directed directed acyclic graphs(DAG).Support for system administrators on the rules of self-expansion and simple and flexible through the web interface configuration operations.2)The stream data process framework is applied to the intelligent building monitoring data processing system,intelligent building transmission of data collection,transmission and abnormal detection of real-time greatly enhanced by the traditional data processing time to upgrade to the level of minutes or even seconds level.At the same time,the anomaly monitoring data processing system is based on the distributed framework to facilitate the expansion.3)The fault alarm notification and real-time anomaly monitoring and processing system using distributed message middleware kafka decoupling to ensure the rapid detection of abnormal processing and fault alarm timely notification.Support the priority of the fault alarm information processing,to protect the degree of damage and emergency priority notification of the failure to deal with timely and effective fault handling.