Prediction and programming of microemulsion phase behavior simulation

In the process of microemulsion flooding,microemulsion phase state may be affected by the adsorption of core and the distribution of oil and water,and the upper phase,the middle phase or the lower phase microemulsion will appear.Accurate description of microemulsion phase state and quantitative discrimination of equilibrium phase composition are of great theoretical value and research significance for the design of microemulsion flooding system and improving the recovery efficiency of microemulsion.Therefore,in this paper,microemulsion phase model is deeply studied.Aiming at the difficulty of solving the existing Hand model and the unknown parameters of the improved HLD-NAC model,we introduce the binodal curve range parameter D and the asymmetric migration degree parameter B,and establish a new phase behavior description method of the binodal curve,the two-phase plait point lines and the III-phase node line.Then the phase discrimination programming of microemulsion was carried out,and the change laws of phase number,phase type and mass fraction of equilibrium phases of any microemulsion system under different salinity were revealed.The results show that the simulation results of microemulsions accord with the salinity scanning law,and can accurately identify the complex microemulsion phase states.

Experimental and numerical investigations on wave motions over platform reef-flat

Previous studies have shown that high-frequency(HF)waves,low-frequency(LF)waves and wave set-up coexist on shallow coral reef-flat and jointly contribute to potential floods and subsequent damages of infrastructures and islands on it.To better understand the reef-flat wave dynamics with incident waves and still water level,a wave-flume experiment was performed based on an idealized platform reef composed of a steep reef-face(1:4),a relatively mild reef-rim(1:14)and a horizontal reef-flat.Also,the non-hydrostatic phase-resolving model SWASH was validated against the experiment and then applied to further numerically investigate the effects of reef-rim topographic features on the reef-flat wave motions.The results show that incident waves of a larger wave height and a longer wave period can generate larger LF waves and wave set-up,thereby inducing greater HF waves on the reef-flat.Higher still water level can lead to larger HF waves but result in smaller wave set-up.In contrast to HF waves and wave set-up,LF waves are minimally affected by the still water level.A rim of milder slope and larger edge depth will induce smaller HF and LF waves and set-up on the reef-flat,and thus provide better protection for the reef-flat region.Furthermore,on the reef-flat,the ratio of HF significant wave height to water depth H_(s_h)/(h_(r)+η^(-))is approximately constant;the dimensionless LF significant wave height Hs_l/H_(s0) and the dimensionless wave set-upη^(-)/[T_(p)(gH_(s0))1/2]can be related to the inverse wave steepness parameter gT_(p)^(2)/H_(s0) and the relative reef-flat submergence(h_(r)+η^(-))/H_(s0) respectively.

A Heuristic Sampling Method for Maintaining the Probability Distribution

Sampling is a fundamental method for generating data subsets.As many data analysis methods are developed based on probability distributions,maintaining distributions when sampling can help to ensure good data analysis performance.However,sampling a minimum subset while maintaining probability distributions is still a problem.In this paper,we decompose a joint probability distribution into a product of conditional probabilities based on Bayesian networks and use the chi-square test to formulate a sampling problem that requires that the sampled subset pass the distribution test to ensure the distribution.Furthermore,a heuristic sampling algorithm is proposed to generate the required subset by designing two scoring functions:one based on the chi-square test and the other based on likelihood functions.Experiments on four types of datasets with a size of 60000 show that when the significant difference level,a,is set to 0.05,the algorithm can exclude 99.9%,99.0%,93.1%and 96.7%of the samples based on their Bayesian networks-ASIA,ALARM,HEPAR2,and ANDES,respectively.When subsets of the same size are sampled,the subset generated by our algorithm passes all the distribution tests and the average distribution difference is approximately 0.03;by contrast,the subsets generated by random sampling pass only 83.8%of the tests,and the average distribution difference is approximately 0.24.