Research On Sensor And Network Data Preprocessing In CPS Based On Entropy

Cyber-Physical Systems(CPS)is an open embedded system that integrates precision computing technology,communication technology,and control technology to accurately and in real time respond to the real world event.In the environment of CPS monitoring system,the data collected by each monitoring point sensor can be divided into two categories: the first type is real-time,continuous,time-series based data,which changes steadily with the change of the environment;the second type is used to the data describing the state of the physical equipment and the monitoring object,which is catastrophic,can indicate the presence or absence of an emergency.For CPS systems,the second type of data with greater uncertainty can trigger actuators.However,subject to various factors,CPS networks often exhibit limited network conditions such as low network bandwidth and frequent congestion.At this time,the second type of data cannot be guaranteed to be transmitted to the data collection end in real time,which makes the CPS monitoring system difficult to deal with in the real world.The emergency response responds.Therefore,it is urgent to apply the data pre-processing method under the limited network conditions to improve the emergency response rate at this time.The main work of this paper is as follows:(1)In order to better preprocess the data in CPS,a data classification model based on information entropy under CPS is proposed,and the representation of monitoring scenes and monitoring data based on CPS is described.In the CPS-based monitoring system,abnormal data that can trigger actuators are often abnormal data.The uncertainty of these abnormal data is greater.The greater the uncertainty,the greater the probability that the actuator will be triggered.Therefore,according to the monitoring data the size of the uncertainty divides the monitoring data into two types of models,namely the low entropy data model and the entropy high data model.(2)In order to reduce the congestion caused by the limited network bandwidth in the CPS,improve the ability to identify abnormal data with great uncertainty,and fully guarantee the response rate of the CPS to emergencies,the complexity of the CPS is analyzed from the perspective of information theory,according to the CPS.The average dynamic complexity of the system is set as a threshold to determine the entropy of the sensor data in a certain period of time.The CPS selects the information entropy high data to be sent first,and its effectiveness is analyzed through experiments.(3)In order to complete the data of the CPS and facilitate the analysis and processing of data in the later period,when the system is idle,the entropy-low data of the non-transmitted information is filled in,and part of the entropy-low data of the information is uploaded according to the frequency of the time interval T,and then based on the IMM algorithm.The information entropy of the complete department was supplemented with low data,and it was verified through experiments that it could well complement the low information entropy data.The LZ4 compression algorithm is improved.When the information entropy has a low data volume,it is compressed,and through experiments,it has a good compression effect,and further solves the congestion problem caused by the limited network bandwidth in the CPS.A dynamic optimization method based on entropy is proposed.That is,with the change of the monitoring environment,the time interval T of the sensor data collection must be adjusted accordingly to meet the system requirements and improve the effect of low data entropy.