Research And Simulation Of Sensor Data Model In Microclimate Environment

Indoor environment is a small environment for people's daily life,this particular environment,we call the microclimate.It refers to a different phenomenon with the surrounding environment climate in a small range.Microclimate not only affects people's production,life and health of all aspects,and to a large extent determines the quality of life.In this dissertation,three aspects of sensor data acquisition,modeling and simulation are studied for the application of Internet of things(IoT)in microclimate environment,and the following work is completed:(1)In the microclimate monitoring,we generally use the equal interval data acquisition method: according to the pre-set time interval,the periodic collection of data.This method is simple,but does not take the changes into account.Small interval will often cause a lot of data redundancy,and large interval cannot accurately reflect the law and characteristics of data changes.In this dissertation,a self-adaptive frequency conversion(SAFC)data acquisition strategy based on swing door trending algorithm is proposed.The strategy is innovative and simple,and it is not only feasible but also efficient.(2)In view of the fact that the data collected by the SAFC data acquisition strategy is unequal interval,this dissertation then proposes an unequal interval data modeling method based on correlation coefficient.Firstly,the piecewise linear interpolation method is used to convert the unequal interval data into equal interval data.And then the correlation analysis and regression analysis of these data are carried out.Finally,the regression model based on the correlation coefficient is established.After the test and residual analysis,the fitting effect of the model is excellent.(3)This dissertation realized a sensor data simulation system,and the virtual sensor in the system can generate the necessary sensor data in real time.The data generation strategy is based on the correlation coefficient of the regression model which is very close to the data collected by the real sensor group.Massive sensor data will take up a lot of storage resources,which provides a great challenge to the development of IoT applications.IoT applications need to collect and store sensor data for a long time,and a large number of storage resource consumption increases the development costs of IoT applications.In this dissertation,the SAFC data acquisition strategy based on swing door trending algorithm can reduce the data redundancy in the sensor data acquisition stage,at the same time,it is the basis of further research on the unequal interval data modeling method based on correlation coefficient.The unequal interval data regression model reveals the nature of the general relationship between the different sensor data,so the simulation data generated in the simulation system is reliable and effective.