Multiple Graph Matching and Clustering via Decayed Pairwise Matching Composition

Jointly matching of multiple graphs is challenging and recently has been an active topic in machine learning and computer vision. State-of-the-art methods have been devised, however, to our best knowledge there is no effective mechanism that can explicitly deal with the matching of a mixture of graphs belonging to multiple clusters, e.g., a collection of bikes and bottles. Seeing its practical importance, we propose a novel approach for multiple graph matching and clustering. Firstly, for the traditional multi-graph matching setting, we devise a composition scheme based on a tree structure, which can be seen as in the between of two strong multi-graph matching solvers, i.e., MatchOpt (Yan et al. 2015a) and CAO (Yan et al. 2016a). In particular, it can be more robust than MatchOpt against a set of diverse graphs and more efficient than CAO. Then we further extend the algorithm to the multiple graph matching and clustering setting, by adopting a decaying technique along the composition path, to discount the meaningless matching between graphs in different clusters. Experimental results show the proposed methods achieve excellent trade-off on the traditional multi-graph matching case, and outperform in both matching and clustering accuracy, as well as time efficiency.

A Heuristic Algorithm for Optimal Service Composition in Complex Manufacturing Networks

Service composition in a Cloud Manufacturing environment involves the adaptive and optimal assembly of manufacturing services to achieve quick responses to varied manufacturing needs. It is challenged by the inherent heterogeneity and complexity of these services in terms of their diverse and complex functions, qualities of service, execution paths, etc. In this paper, a manufacturing network is constructed to explicitly identify and describe the relationships between individual services based on their attributes. On this basis, the service composition problem can be modeled as a multiple-constrained optimal path (MCOP) selection problem by taking into account different types of composition, namely, sequence, parallel, selection, and cycle. A novel Dual Heuristic Functions based Optimal Service Composition Path algorithm (DHA_OSCP) is proposed to solve the NP-Complete MCOP problem, which involves exploiting the backward search procedure with different search targets to obtain two heuristic functions for the forward search procedure. The proposed algorithm is evaluated through a set of computational experiments in which the proposed algorithm and other popular algorithms such as MFPB_HOSTP are applied to the same dataset, and the results obtained show that DHA_OSCP can efficiently find the optimal service composition path with better Quality of Service (QoS). The viability of DHA_OSCP is further proved in a case study of services composition on a Cloud Manufacturing platform.

Mass Transfer on Multiphase Transitions in Low-Temperature Carbon-Dioxide Floods

Summary Mixtures of reservoir oil and carbon dioxide (CO2) can exhibit complex multiphase behavior at temperatures typically less than 120°F, in which a third CO2-rich liquid (L2) phase can coexist with the oleic (L1) and gaseous (V) phases. The three-phase behavior is bounded by two types of critical endpoints (CEPs) in composition space. The lower CEP (LCEP) is a tie line in which the two liquid phases merge in the presence of the V phase, and the upper CEP (UCEP) is a tie line in which the L2 and V phases merge in the presence of the L1 phase. Slimtube tests reported in the literature show that low-temperature oil displacement by CO2 can result in the high displacement efficiency of more than 90% when three phases are present during the displacement. The nearly piston-like displacements can be quantitatively reproduced in numerical simulations when the CEP behavior is properly considered. However, it is uncertain how multicontact miscibility (MCM) is developed through the interaction of flow and three-hydrocarbon-phase behavior. This research presents a detailed analysis of mass conservation on multiphase transitions between two and three phases for the limiting three-phase flow, where the L1 phase is completely displaced by the L2 phase on the LCEP. The analysis indicates that interphase mass transfer on multiphase transitions occurs in the most-efficient way for MCM development. Simple analytical conditions derived for MCM through three phases are applied to 1D fine-scale simulations of CO2 floods by use of four and more components. Results show that the MCM conditions are nearly satisfied when the effect of numerical dispersion is small. MCM is likely developed through three hydrocarbon phases on the LCEP in the cases studied. This is consistent with analytical solutions of water and gas injection presented in the literature, in which MCM is developed on a CEP for the aqueous, V, and L1 phases. For MCM cases in this research, the L2-V two phases are present upstream of the miscible front if the composition path does not go through the UCEP tie line. However, they also can be miscible on the non-L1 edge of the UCEP tie line if the MCM composition path goes through it. Three-phase flow gradually changes to two-phase flow with varying pressure in the presence of numerical dispersion. It is shown that interphase mass transfer on multiphase transitions becomes less efficient during the change until the three-phase region completely disappears.

Analytical Solutions for Gas Displacements With Bifurcating Phase Behavior

Summary Minimum miscibility pressure (MMP) is one of the most important parameters in the design of a successful gasflooding process. The most-reliable methods to calculate the MMP are based on slimtube experiments, 1D slimtube simulations, mixing-cell calculations, and the analytical methods known as the method of characteristics (MOC). The calculation of MMP by use of MOC is the fastest method because it relies solely on finding the key tie lines in the displacement path. The MOC method for MMP estimation in its current form assumes that the composition path is a series of shocks from one key tie line to the next. For some oils, however, these key tie lines do not control miscibility, and the MMP calculated by use of the key-tie line approach can be significantly in error. The error can be as high as 5,000 psia for heavier oils or CO2 displacements at low temperature in which three-phase hydrocarbon regions can exist (L1–L2–V). At higher pressures, the two- or three-phase region can split (or bifurcate) into two separate two-phase regions (L1–L2 and L1–V regions). Thus, for the MMP calculation from MOC to be correct, we must calculate the entire composition path for this complex phase behavior, instead of relying on the shock assumption from one key tie line to the next. In this paper, the MOC-composition route is developed completely for the bifurcating phase-behavior displacement for pure CO2 injection by use of a simplified pseudoternary system that is analogous to the complex phase behavior observed for several real displacements with CO2. We develop the MOC analytical solutions by honoring all constraints required for a unique solution—velocity, mass balance, entropy, and solution continuity. The results show that a combination of shocks and rarefaction waves exists along the nontie-line path, unlike previous MOC solutions reported to date. We show that by considering the entire composition path, not just the key tie lines, the calculated MMP agrees with the mixing-cell method. We also show that, in this complex ternary displacement, the displacement mechanism has features of a both condensing and vaporizing (C/V) drive, which was thought to be possible only for gasfloods with four or more components. For pure CO2 injection, the solution also becomes discontinuous for oils that lie on the tie line envelope curve. Finally, we show that shock paths within the two-phase region are generally curved in composition space and that there is no MMP for some oil compositions considered in the displacements by CO2. Recovery can be large even though the MMP is not reached.