Real-Time Monitoring Model Of Streaming Data And Related Software Development

With the development of science and technology,a variety of real-time streaming data is emerged incessantly and recorded in the era of the data explosion.Statistical Process Control(SPC)can efficiently monitor the data of industrial production,daily life and emergencies,find anomalies as soon as possible,thereby improving the quality of system operation and reducing economic losses.Therefore,a new and more efficient control chart for real-time monitoring of streaming data is proposed and related software is developed.The development trend of new coronavirus disease is also monitored and analyzed based on the control chart.Firstly,using distribution-free control charts monitors the streaming data in real time.Considering the advantages of Wilcoxon rank sum test and Ansari-Bradley test,a new change point detection model that can simultaneously monitor changes in shape and scale parameters is proposed.By comparing with other control charts,the performance of this control chart is powerful.Secondly,in the process of monitoring the stream data based on the control chart,it is necessary to use the goodness-of-fit method to test data.The R package CPdtest of control chart and the R package Distribution Test of goodness-of-fit test are developed by R language.Compared with the traditional goodness-of-fit tests,the Distribution Test package has better performance.Finally,for the streaming data of the new coronavirus disease,regression analysis method and control charts are applied to analyze and monitor trend of development of pandemic.The data of the number of people under medical observation in Hubei province is collected and its growth rate is extrapolated linearly.It is predicted that the time for unblocking of Wuhan will be about April 5 to April 7.At the same time,The monitoring model for the future trend of development of the pandemic in China and the world is provided with applying the control chart.