Search

LBSs have been prospering over recent years. Although many LPPMs are proposed to solve specific problems under certain settings, it is still hard to satisfy the different demands of users due to the diverse requirements for quality of service (QoS) and privacy preservation. In this article, we propose an adaptive scheme to achieve fine-grained QoS for privacy-aware users in LBSs. Specifically, we construct a Bayesian-based classification model to identify the QoS requirements of users under different circumstances. QoS and privacy degrees provided by LBSs are carefully defined based on the concept of rate-distortion theory, which plays an important role in bridging the LPPMs with users' requirements. When users request the location service from the built-in GPS module in smart devices, we first specify their circumstances and requirements, then reply with the modified locations. We also design an efficient algorithm to dynamically choose the appropriate privacy strategy according to the QoS requirements of users in different circumstances. Experimental results show that our scheme can identify QoS requirements accurately and protect location privacy adaptively.

The micromechanics of multiphase flow in fractured porous media are explored through the construction of microstructures to the modelling of micro-physics and macro-behaviors. Accordingly, four parts of studies are conducted so as to give a scientific explanation about the mechanism of two phase flow in fractured porous media. The quantitative characterization of the microstructure and statistical description of the porous media are discussed in the first part. The microstructure reconstruction of sandstone from limited surface images is carried out by using simulated annealing method. The effectiveness of the reconstructed sandstone microstructure is tested and verified by comparing with available X-ray computed tomography (CT) experimental data. The characterization of the multiphase distribution in fractured porous media is investigated in the second part. The single component multiphase (SCMP) lattice Boltzmann method (LBM) is selected among three commonly used numerical approaches through a comparison against the available results of micro X-ray CT. The two-phase fluid distribution in porous media and the effect of fracture are investigated. In the third part, the single phase flow in natural fracture network is numerically investigated. The influence of fracture roughness and deformation on the fracture transmissivity/permeability is numerically studied. Meanwhile, the ability of LBM in the study of fluid flow in fracture network is validated through the comparison to the pipe network model. The influence of fracture length, fracture density, and deformation on the fluid flow is also studied preliminarily from coupling LBM with the Discrete Fracture Network (DFN) model and Discrete Element Model (DEM). The multiphase flow through fractured porous media is studied in the fourth part. The SCMP LBM is found effectively in simulating the two phase flow problems through the comparison with the multi-component multiphase (MCMP) LBM. Moreover, the body force for the simulation is calibrated and effect of wettability on two phase flow is studied. After that, the influence of fracture on two phase flow is investigated. The trial of two phase flow in 3D reconstructed porous media is conducted and some suggestions are given at the end. Based on the comprehensive studies and numbers of numerical simulations, a clear understanding of macroscopic behaviors and their dependence on microstructure and microscopic physics process at pore scale are obtained.