Micro-environment Data Perception And Processing Analysis On Mobile IoT Platform

With the development of Internet of Things technology and sensor equipment,sensors have been widely used in agriculture,industry and daily life.For environmental quality monitoring,sensor networks play an important role in the signal sampling stage.However,for fine-grained environmental quality monitoring,due to the large number of sensors required,the cost and accuracy of the sensors have been hindering the development of micro-environmental quality monitoring.How to form a high-quality IoT platform at low cost to perceive,process and analyze the quality of the microenvironment has become an urgent problem to be solved.This thesis designs a mobile IoT platform to monitor the quality of the micro-environment.The main contents include the following:(1)This thesis proposes a sensor calibration model for low-precision sensors.The training of this model uses a two-stage learning method.In order to reduce the cost of sensors,low-cost,low-precision sensors are often used to form a sensor network.Due to the large error between the data measured by low-cost and low-precision sensors and the true value,a calibration model needs to be trained to make the measured value of the sensor close to the true value.By using the two-stage learning method to learn the relationship between the measured value and the true value,a practical sensor calibration model can be obtained.After using the calibration model,the accuracy of the data collected by the sensor can be significantly improved,and the calibrated data can be close to the true value.(2)Based on Bayesian compressive sensing theory(BCS)and adaptive Bayesian compressive sensing technology,a low-cost fine-grained environmental monitoring technology is proposed.By gridding the monitoring space,calculating the information gain on each grid,and deploying the sensors to the appropriate grid,it is possible to recover the environmental data of all grids with high accuracy with a small number of sensors.This technology can make the sensor have mobile deployment capability and realize the rapid collection of environmental data at the required location by combining the group intelligence perception technology,and improve the accuracy of data reconstruction.(3)This thesis designs a mobile IoT platform prototype,which uses the sensor calibration model proposed in this paper to calibrate the sensor data,and the proposed micro-environment monitoring technology based on Bayesian compressed sensing to sense the micro-environment data,processing and analysis.The mobile internet of things platform realizes the movement of the sensor by using the method of mobile group intelligence perception,increases the reuse rate of the sensor,and further reduces the cost of the platform.The platform has completed experimental tests on campus,and the experimental results show that the platform is a lightweight,low-cost,and accurate micro-environment quality monitoring platform.