Research On Time Series Data Computing And Visualization Of Sensor Networks

How to use these platforms to query and mine time series data has become a hot research topic.Combined with its own work,this paper conducts temperature monitoring and control based on time series related research.It is important to consider how to calculate the correlation coefficients on HBase.Correlation coefficients between long-term HBase sequences need to be calculated to communicate with the main I/O and network data,making it difficult to apply interactive queries,such as temperature data streams acquired by sensors in real time,through correlation analysis of acquired time series data,to achieve temperature monitoring and control.In order to solve this problem,a storage and computing platform based on Hadoop and HBase is set up.The storage and calculation framework of massive time series data is designed,and the distributed computing method is designed.Finally,the method is applied to engineering practice to mine value information.And realize the visualization of the data.In engineering applications,the efficiency and calculation methods of the platform have been verified.Specifically,the main tasks of this article are:(1)Establish a distributed storage and computing platform,design a distributed storage method for massive time series data,store the collected temperature data stream,and design HBase secondary index for nonprimary key query;(2)Designing an estimation method for quickly calculating the approximation of time series correlation coefficients,and studying the computational framework design and algorithm design of temperature data flow similarity query.In order to quickly derive the correlation coefficients of the upper and lower limits,the TSDC algorithm is proposed,and the I/O and network data are transmitted and considered.The ATSDC algorithm was obtained by the extension of the TSDC algorithm.The ATSDC algorithm mainly processes multilayer summary data and can transmit data while quickly estimating the correlation coefficient of the data;(3)Apply the storage framework and calculation method to a scientific research project(centralized heating system remote monitoring and optimization control application research)massive sensor data analysis and mining platform,and realize the visualization of value data,after field testing,this paper The proposed storage and calculation method can well analyze the mass sensor temperature data stream.(4)The storage framework and calculation method are applied to the data processing and mining platform of the Xinjiang Autonomous Region Science and Technology Plan Project(centralized heating system remote monitoring and optimization control application research),and the value data is visualized.After field testing,the paper design.The storage and calculation methods are good for managing massive sensor temperature data.