Soft sensor model derived from Wiener model structure: modeling and identification

The processes of building dynamic and static relationships between secondary and primary variables are usually integrated in most of nonlinear dynamic soft sensor models. However, such integration limits the estimation accuracy of soft sensor models. Wiener model effectively describes dynamic and static characteristics of a system with the structure of dynamic and static submodels in cascade. We propose a soft sensor model derived from Wiener model structure, which is an extension of Wiener model. Dynamic and static relationships between secondary and primary variables are built respectively to describe the dynamic and static characteristics of system. The feasibility of this model is verified. Then the expression of discrete model is derived for soft sensor system. Conjugate gradi-ent algorithm is applied to identify the dynamic and static model parameters alternately. Corresponding update method for soft sensor system is also given. Case studies confirm the effectiveness of the proposed model, alternate identification algorithm, and update method.     

Output feedback robust model predictive control with unmeasurable model parameters and bounded disturbance☆

The output feedback model predictive control (MPC), for a linear parameter varying (LPV) process system including unmeasurable model parameters and disturbance (all lying in known polytopes), is considered. Some previously developed tools, including the norm-bounding technique for relaxing the disturbance-related constraint handling, the dynamic output feedback law, the notion of quadratic boundedness for specifying the closed-loop stability, and the el ipsoidal state estimation error bound for guaranteeing the recursive feasibility, are merged in the control design. Some previous approaches are shown to be the special cases. An example of continuous stirred tank reactor (CSTR) is given to show the effectiveness of the proposed approaches.     

A comprehensive fractal char combustion model☆

The char combustion mechanisms were analyzed and a comprehensive fractal char combustion model was developed to give a better understanding and better predictions of the char combustion characteristics. Most of the complex factors affecting the char combustion were included, such as the coupling effects between the pore diffusion and the chemical reactions, the evolution of the char pore structures and the variation of the apparent reaction order during combustion, the CO/CO2 ratio in the combustion products and the correction for oxy-char combustion. Eleven different chars were then combusted in two drop tube furnaces with the conversions of the partly burned char samples measured by thermogravimetric analysis. The combustion processes of these chars were simulated with the predicted char conversions matching very well with the measured data which shows that this char combustion model has good accuracy. The apparent reaction order of the char combustion decreases, stabilizes and then increases during the combustion process. The combustion rates in the oxy-mode are general y slower than in the air-mode and the effect of the char-CO2 gasification reac-tion becomes obvious only when the temperature is relatively high and the O2 concentration is relatively low.     

Surface interaction model ofγ-alumina-supported metal oxides

An incorporation model has been proposed and used to discuss the interaction between various metal oxides and -alumina. The dispersion capacities of various metal oxides are predicted and the surface hydroxyl group density on -alumina estimated, based on the assumption that the (110) plane is preferentially exposed on the surface of -alumina. These results are in good agreement with the available experimental data.     

Molecular thermodynamic model for fluids containing associated molecules

With the statistical mechanical theory for square-well-chain fluids as the basic princi-ples,a molecular thermodynamic model including expressions for the Helmholtz function and thecompressibility factor for associated square-well-chain fluids has been developed.Employment ofthe shield-sticky concept enables the corporation of association bonds between in the model.Good agreement with computer-simulation results for dimer-linear quadrimer systems is obtained.Satisfactory correlation for vapor pressures and saturated liquid volumes of pure substances con-taining associated molecules(such as carboxylic acids,alcohols,amines,water,etc.)with four mo-lecular parameters indicates the applicability of the model.     

Metal–support interactions on rhodium model catalysts

Metal–support interactions on supported rhodium catalysts were studied by using specially prepared Rh/TiO₂/Mo model systems. For their characterization and the analysis of modifications due to various heat treatments several surface analytical methods were applied: low-energy ion scattering, X-ray photoemission spectroscopy and thermal desorption spectroscopy. Heating in ultrahigh vacuum to 670 K leads to Rh agglomeration followed (above 720 K) by encapsulation including the formation of reduced titanium oxide species. These morphological and chemisorption changes are reversible upon reoxidation and low-temperature reduction and thus exhibit the characteristic features of strong metal–support interactions. For the effective mechanism a reaction is suggested that involves oxygen chemisorption on the Rh clusters and partial reduction of the surrounding support oxide.     

Anscombe’s model for sequential clinical trials revisited

In Anscombe’s classical model, the objective is to find the optimal sequential rule for learning about the difference between two alternative treatments and subsequently selecting the superior one. The population for which the procedure is optimized has size N and includes both the patients in the trial and those who are treated with the chosen alternative after the trial. We review earlier work on this problem and give a detailed treatment of the problem itself. In particular, whereas previous work has mainly focused on the case of conjugate normal priors for the incremental effect, we demonstrate how to handle the problem for priors of a general form. We also discuss methods for numerical solutions and the practical implications of the results for the regulation of clinical trials.

Two extensions of the model are proposed and analyzed. The first breaks the symmetry of the treatments, giving one the role of the current standard being administered in parallel with the trial. We show how certain asymptotic results due to Chernoff can be adapted to this asymmetric case. The other extension assumes that N is a random variable instead of a known constant.     

A dual-scale turbulence model for gas-liquid bubbly flows☆

A dual-scale turbulence model is applied to simulate cocurrent upward gas–liquid bubbly flows and validated with available experimental data. In the model, liquid phase turbulence is split into shear-induced and bubble-induced turbulence. Single-phase standard k-εmodel is used to compute shear-induced turbulence and another transport equation is added to model bubble-induced turbulence. In the latter transport equation, energy loss due to interface drag is the production term, and the characteristic length of bubble-induced turbulence, simply the bubble diameter in this work, is introduced to model the dissipation term. The simulated results agree well with experimental data of the test cases and it is demonstrated that the proposed dual-scale turbulence model outperforms other models. Analysis of the predicted turbulence shows that the main part of turbulent kinetic en-ergy is the bubble-induced one while the shear-induced turbulent viscosity predominates within turbulent vis-cosity, especially at the pipe center. The underlying reason is the apparently different scales for the two kinds of turbulence production mechanisms:the shear-induced turbulence is on the scale of the whole pipe while the bubble-induced turbulence is on the scale of bubble diameter. Therefore, the model reflects the multi-scale phe-nomenon involved in gas–liquid bubbly flows.     

Numerical simulation of gas–liquid flow in the bubble column using Wray–Agarwal turbulence model coupled with population balance model

In this paper, an improved computational fluid dynamic(CFD) model for gas–liquid flow in bubble column was developed using the one-equation Wary–Agarwal(WA) turbulence model coupled with the population balance model(PBM). Through 18 orthogonal test cases, the optimal combination of interfacial force models, including drag force, lift force, turbulent dispersion force. The modified wall lubrication force model was proposed to improve the predictive ability for hydrodynamic behavior near the wall ...     

Determination of parameters in a two-dimensional model for fixed-bed adsorbers

A new method is proposed for the determination of the parameters in a two-dimensionalmodel which characterizes the properties of axial and radial mixing and mass transport in afixed-bed adsorber.Parameter estimation for the model is carried out with methane-air-5A molecularsieve in a bed under the condition of step injection of tracer from a point on the main axis of thebed by the curve fitting method in the time domain.     

How to conceptualize catalytic cycles? The energetic span model

A computational study of a catalytic cycle generates state energies (the E-representation), whereas experiments lead to rate constants (the k-representation). Based on transition state theory (TST), these are equivalent representations. Nevertheless, until recently, there has been no simple way to calculate the efficiency of a catalytic cycle, that is, its turnover frequency (TOF), from a theoretically obtained energy profile. In this Account, we introduce the energetic span model that enables one to evaluate TOFs in a straightforward manner and in affinity with the Curtin-Hammett principle. As shown herein, the model implies a change in our kinetic concepts. Analogous to Ohm's law, the catalytic chemical current (the TOF) can be defined by a chemical potential (independent of the mechanism) divided by a chemical resistance (dependent on the mechanism and the nature of the catalyst). This formulation is based on Eyring's TST and corresponds to a steady-state regime. In many catalytic cycles, only one transition state and one intermediate determine the TOF. We call them the TOF-determining transition state (TDTS) and the TOF-determining intermediate (TDI). These key states can be located, from among the many states available to a catalytic cycle, by assessing the degree of TOF control (X(TOF)); this last term resembles the structure-reactivity coefficient in classical physical organic chemistry. The TDTS-TDI energy difference and the reaction driving force define the energetic span (δE) of the cycle. Whenever the TDTS appears after the TDI, δE is the energy difference between these two states; when the opposite is true, we must also add the driving force to this difference. Having δE, the TOF is expressed simply in the Arrhenius-Eyring fashion, wherein δE serves as the apparent activation energy of the cycle. An important lesson from this model is that neither one transition state nor one reaction step possess all the kinetic information that determines the efficiency of a catalyst. Additionally, the TDI and TDTS are not necessarily the highest and lowest states, nor do they have to be adjoined as a single step. As such, we can conclude that a change in the conceptualization of catalytic cycles is in order: in catalysis, there are no rate-determining steps, but rather rate-determining states. We also include a study on the effect of reactant and product concentrations. In the energetic span approximation, only the reactants or products that are located between the TDI and TDTS accelerate or inhibit the reaction. In this manner, the energetic span model creates a direct link between experimental quantities and theoretical results. The versatility of the energetic span model is demonstrated with several catalytic cycles of organometallic reactions.     

RESEARCH NOTES Unified model of purification units in hydrogen networks

Purification processes are widely used in hydrogen networks of refineries to increase hydrogen reuse. In refineries, hydrogen purification techniques include hydrocarbon, hydrogen sulfide and CO removal units. In addition, light hydrocarbon recovery from the hydrogen source streams can also result in hydrogen purification. In order to simplify the superstructure and mathematical model of hydrogen network integration, the models of different purification processes are unified in this paper, including mass balance and the expressions for hydrogen recovery and impurity removal ratios, which are given for all the purification units in refineries. Based on the proposed unified model, a superstructure of hydrogen networks with purification processes is constructed.     

Dynamic analysis on methanation reactor using a double-input-multi-output linearized model☆

A double-input–multi-output linearized system is developed using the state-space method for dynamic analysis of methanation process of coke oven gas. The stability of reactor alone and reactor with fee...     

Adaptive partitioning PCA model for improving fault detection and isolation☆

In chemical process, a large number of measured and manipulated variables are highly correlated. Principal com-ponent analysis (PCA) is widely applied as a dimension reduction technique for capturing strong correlation un-derlying in the process measurements. However, it is difficult for PCA based fault detection results to be interpreted physical y and to provide support for isolation. Some approaches incorporating process knowledge are developed, but the information is always shortage and deficient in practice. Therefore, this work proposes an adaptive partitioning PCA algorithm entirely based on operation data. The process feature space is partitioned into several sub-feature spaces. Constructed sub-block models can not only reflect the local behavior of process change, namely to grasp the intrinsic local information underlying the process changes, but also improve the fault detection and isolation through the combination of local fault detection results and reduction of smearing effect. The method is demonstrated in TE process, and the results show that the new method is much better in fault detection and isolation compared to conventional PCA method.     

Char characterisation and its application in a coal burnout model☆

In this study, char image analysis techniques have been employed to investigate the morphology of chars obtained from a Drop-Tube furnace. Char image analysis results have been incorporated as inputs to a char burnout model based on Hurt's CBK model. It has been observed that the char combustion rate was strongly affected by char structural parameters and the inclusion of char morphology has led to a better prediction of char burnout. It has also been suggested by the model that the inclusion of ash inhibition overestimates the resistance attributed by ash film and the consideration of ash film resistance should be undertaken in a different way to give a better prediction at the later stages of char combustion.     

A continuum model for gas–liquid flow in packed towers

Gas–liquid flow in packed towers is commonly encountered in the chemical and processing industry. A continuum model is developed based on the volume-and-time averaging of multiphase flows in isotropic rigid porous media/packed columns. Closures are presented for the evaluations of the extra surface/intrinsic phase integral terms. Both inertia and inter-phase interactions are retained in the volume averaged (Navier–Stokes) equations. These governing equations are solved for fully-developed axi-symmetric single and gas–liquid two phase flows in highly porous packed towers. It is found that the dispersion term is present in the continuity equation as well as the momentum equations. Numerical simulations with the models show that the volume-and-time averaged equations can predict the velocity, phase hold-up and pressure drop quite well for up to the loading point for gas–liquid counter-current flows.     

A grain size distribution model for non-catalytic gassolid reactions

A new model to describe the non-catalytic conversion of a solid by a reactant gas is proposed. This so-called grain size distribution (GSD) model presumes the porous particle to be a collection of grains of various sizes. The size distribution of the grains is derived from mercury porosimetry measurements. The measured pore size distribution is converted into a grain size distribution through a so-called pore-tosphere factor whose value is also derived from the porosimetry measurements. The grains are divided into a number of size classes. For each class the conversion rate is calculated either according to the shrinking core model, involving core reaction and product layer diffusion as rate-determining steps or according to a new model in which some reaction at the grain surface is assumed to be limiting. The GSD model accounts for the phenomenon of pore blocking by calculating the maximum attainable conversion degree for each size class. In order to verify the model, two types of precalcined limestone particles with quite different microstructures were sulphided as well as sulphated. Furthermore, a single sample of sulphided dolomite was regenerated with a mixture of carbon dioxide and steam. For each reaction good agreement was attained between measured and simulated conversion vs. time behaviour.     

Blending silica bodies in centrifugal runner mills, “model 115”

Conclusions The output of model 115 centrifugal runner mills for mixing dinas bodies is 1.7 times higher than that of runner mills with heavy rollers. Goods pressed from bodies prepared in centrifugal runner mills are no different in quality from those pressed from bodies made in runner mills with heavy rollers. The difference in output of acceptable product is slight.In the preparation of bodies in model 115 runner mills there are few fines in the batch. When formulating the batch it is necessary to introduce 20–30°% quartzite ground in a tube mill with a content of not less than 80% fraction finer than 0.088 mm. With this, the fraction finer than 0.5 and 3–2 mm should be considered; air blowing of the body should be started 2–2.5 min after the start of blending.To observe the norms for alkalinity and moisture content of the batch, with a moisture content of 2–2.4% the total content of CaO in the lime-iron slip is 23.5–24.5% and with a wet weight of 700–750 kg the amount of slip varies from 43 to 50 liters, the blending time lies in the 4–5 min range.Scrapers and the lining of the runner mills should be made from manganese steel and the bottom of the mills from carbon steel.A drawback of model 115 runner mills is the sticking of the lime-iron slip and the need to clean them more than once during 24 h.