Research And Implementation Of Data Filtering And Code Generation Methods In Domain-specific Operation Monitoring System

Along with the rapid development of computer technology,computer operation monitoring system is also being used in more and more enterprises.The operation monitoring system usually consists of two parts: monitoring software deployed at the central end and a number of data acquisition devices deployed in the outstations.In recent years,the operation monitoring system in aerospace and other domains have revealed many problems.Among which the data acquisition device in the system is often deployed in places with limited network infrastructure,often facing small network bandwidth,network intermittency and other situations,resulting in the collection of a large amount of monitoring data can not be timely transferred to the monitoring center for processing.In addition,with more and more functions of data collection devices in the system and more nodes to be monitored,the software development becomes more complex,requiring a lot of time and manpower for coding and testing.Moreover,the operation monitoring system often faces the addition of new devices and the removal of old devices,and the code modules change frequently.Therefore,the current operation monitoring software in aerospace and other domains generally have problems such as high code repetition rate and low development efficiency.By analyzing the issues of the current operation monitoring system,this thesis proposes a key data filtering algorithm in monitoring data stream to reduce the data transmission volume in response to the problem of high network fluctuation and bandwidth limitation often faced by data transmission of collection devices.At the same time,a template-based automatic code generation scheme is proposed to address the problems of high code repetition rate and low development efficiency commonly found in the software development process of operation monitoring system.The main research content can be divided into the following two points.(1)By researching several different skyline query algorithms and combining the characteristics of monitoring data,the DSSW algorithm is adapted to the monitoring data stream scenario with "greater than" as the key criterion.And improvements are made from two aspects.On the one hand,dynamic skyline is used instead of traditional skyline to further reduce query skyline.On the other hand,the algorithm filtering strategy is adjusted to reduce the key data filtering time by increasing the filtering of consecutive identical data arriving in sliding windows.on the basis of the above work,a -dominant dynamic skyline query over sliding window(DDSSW)more suitable for monitoring data is proposed.Finally,it is demonstrated experimentally that the DDSSW algorithm significantly improves the filtering ability and filtering time performance of key data in surveillance data.(2)Automatic code generation engine for operation monitoring system is designed and implemented.The engine needs to create a set of relatively complete code templates for the target operation monitoring system project,and provide the data model information required by the system,and implement the automatic generation of the target code file through the Free Marker template engine technology.At the same time,based on the data model information,the code generation technique in this study can also automate the construction of the target project database and project deployment.The experimental comparison with the manual coding method proves that the automatic code generation of the operation monitoring system through the automatic code generation technique has more advantages.This can not only greatly reduce the development cycle,but also effectively improve the maintainability of the system.