| Environmental monitoring is an important part of smart cities,and it is also the prerequisite and foundation for pollution source control,environmental management and protection.Environmental monitoring applications usually need to deploy a large number of fixed sensor nodes to obtain sensing data.Not only the deployment speed is slow,but also the network cost is high.In order to solve these problems,researchers have proposed multiple data prediction and estimation methods,and introduced mobile nodes,to reduce the deployment of network nodes and the costs of network,and to improve the flexibility and robustness of the network.This thesis also researches these problems and does the following work.(1)Data prediction for unmonitored locations may reduce the deployment of sensing nodes and the cost of environmental sensing system.The traditional Kriging has the problem of inaccurate fitting parameters of the variogram,resulting in the inaccurate results.Based on the characteristics of perceptual data,this thesis introduces adaptive weighted least squares to fit the variogram,and proposes two improved Kriging algorithms: an improved Kriging based on residuals and an improved Kriging based on lag distances.Simulation results show that the prediction results of the two algorithms are better than the traditional Kriging.(2)In order to further reduce the number of sensor nodes needed,this thesis designs and implements a mobile sensing simulation system for environment information.The system uses Raspberry Pi 2 as the processing module.And it includes environment-aware nodes,GPS modules,and the power module.The system can collect the environment data of the target area in a mobile way and display them in a graphical interface.The system uses the improved Kriging to predict the unmonitored data.Comprehensive analysis shows that the improved Kriging proposed in this thesis is effective,and the mobile sensing simulation system for regional environment information is practical and feasible. |
PDF Full Text Request