Standing wave optimization of SMB using a hybrid simulated annealing and genetic algorithm (SAGA)

Cauley, Fattaneh G.; Cauley, Stephen F.; Wang, Nien‐Hwa Linda

doi:10.1007/s10450-008-9119-8

Cited by 10 publications

(6 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Since the SWD equations are based on the solutions of the TMB model, this method works well for SMB systems with two or more columns in each zone, where the separation results approach those of a TMB system. The operating conditions determined from the SWD method always achieve the target product purity and yield specified in SWD, as shown in many previous studies [24][25][26][27][28][29][30][31][32][33][34][35][36]. For SMB systems with only one column per zone and severe mass transfer resistance, the operating parameters obtained from SWD are not expected to achieve high product purity or high yield.…”

Section: Introductionmentioning

confidence: 71%

“…They include grid search [31], genetic algorithms [32,33], simulated annealing [34,35], or combined simulated annealing and genetic algorithm (SAGA) [36]. The optimized variables include column configuration ( ), column length (L c ), yields (Y i ), and particle size (R p ) [36]. These techniques cannot guarantee finding of the global optima, and do not show how the material properties and the design parameters are related to the sorbent productivity or the solvent efficiency.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Speedy standing wave design, optimization, and scaling rules of simulated moving bed systems with linear isotherms

Weeden

Wang

2017

Journal of Chromatography A

Self Cite

View full text Add to dashboard Cite

Simulated Moving Bed (SMB) systems with linear adsorption isotherms have been used for many different separations, including large-scale sugar separations. While SMBs are much more efficient than batch operations, they are not widely used for large-scale production because there are two key barriers. The methods for design, optimization, and scale-up are complex for non-ideal systems. The Speedy Standing Wave Design (SSWD) is developed here to reduce these barriers. The productivity (P) and the solvent efficiency (F/D) are explicitly related to seven material properties and 13 design parameters. For diffusion-controlled systems, the maximum P or F/D is controlled by two key dimensionless material properties, the selectivity (α) and the effective diffusivity ratio (η), and two key dimensionless design parameters, the ratios of step time/diffusion time and pressure-limited convection time/diffusion time. The optimum column configuration for maximum P or F/D is controlled by the weighted diffusivity ratio (η/α). In general, high α and low η/α favor high P and F/D. The productivity is proportional to the ratio of the feed concentration to the diffusion time. Small particles and high diffusivities favor high productivity, but do not affect solvent efficiency. Simple scaling rules are derived from the two key dimensionless design parameters. The separation of acetic acid from glucose in biomass hydrolysate is used as an example to show how the productivity and the solvent efficiency are affected by the key dimensionless material and design parameters. Ten design parameters are optimized for maximum P or minimum cost in one minute on a laptop computer. If the material properties are the same for different particle sizes and the dimensionless groups are kept constant, then lab-scale testing consumes less materials and can be done four times faster using particles with half the particle size.

show abstract

Section: Introductionmentioning

confidence: 71%

Section: Introductionmentioning

confidence: 99%

Speedy standing wave design, optimization, and scaling rules of simulated moving bed systems with linear isotherms

Weeden

Wang

2017

Journal of Chromatography A

Self Cite

View full text Add to dashboard Cite

show abstract

“…For example, genetic algorithm (GA) has excellent ability of global search, but it is poor at hill-climbing; whereas simulated annealing (SA) is good at hill-climbing for optimum solutions, but its convergence speed is very slow. Thus, there has been much work on hybriding GA with SA for performance improvement (Cauley, Cauley, & Wang, 2008;He & Hwang, 2006;Jeong & Lee, 1996;Leung, Chan, & Troutt, 2003). Nevertheless, all of these hybrid algorithms are only designed for solving single-objective optimization problems, and to the best of our knowledge, there have been no other attempts, so far, to produce a hybrid algorithm with GA and SA which finds multiple Pareto-optimal solutions for a multi-objective nonlinear transportation problem.…”

Section: Multi-objective Simulated Annealing Genetic Algorithmmentioning

confidence: 96%

Multi-objective optimization matching for one-shot multi-attribute exchanges with quantity discounts in E-brokerage

Jiang

Lau

et al. 2011

Expert Systems with Applications

View full text Add to dashboard Cite

“…The SWD method has been incorporated into various optimization routines, based on grid search [9], genetic algorithms [20], simulated annealing [21,22], or combined simulated annealing and genetic algorithm (SAGA) [23]. Optimization variables include particle size (R p ), 6 column length (L c ), column configuration (N j ), and yields (Y i ) [24].…”

Section: Introductionmentioning

confidence: 99%

Speedy standing wave design of size-exclusion simulated moving bed: Solvent consumption and sorbent productivity related to material properties and design parameters

Weeden

Wang

2015

Journal of Chromatography A

Self Cite

View full text Add to dashboard Cite

Size-exclusion simulated moving beds (SEC-SMB) have been used for large-scale separations of linear alkanes from branched alkanes. While SEC-SMBs are orders of magnitude more efficient than batch chromatography, they are not widely used. One key barrier is the complexity in design and optimization. A four-zone SEC-SMB for a binary separation has seven material properties and 14 design parameters (two yields, five operating parameters, and seven equipment parameters). Previous optimization studies using numerical methods do not guarantee global optima or explicitly express solvent consumption (D/F) or sorbent productivity (PR) as functions of the material properties and design parameters. The standing wave concept is used to develop analytical expressions for D/F and PR as functions of 14 dimensionless groups, which consist of 21 material and design parameters. The resulting speedy standing wave design (SSWD) solutions are simplified for two limiting cases: diffusion or dispersion controlled. An example of SEC-SMB for insulin purification is used to illustrate how D/F and PR change with the dimensionless groups. The results show that maximum PR for both diffusion and dispersion controlled systems is mainly determined by yields, equipment parameters, material properties, and two key dimensionless groups: (1) the ratio of step time to diffusion time and (2) the ratio of diffusion time to pressure-limited convection time. A sharp trade off of D/F and PR occurs when the yield is greater than 99%. The column configuration for maximum PR is analytically related to the diffusivity ratio and the selectivity. To achieve maximum sorbent productivity, one should match step time, diffusion time, and pressure-limited convection time for diffusion controlled systems. For dispersion controlled systems, the axial dispersion time should be about 10 times the step time and about 50 times the pressure-limited convection time. Its value can be estimated from given yields, material properties, and column configuration. Among the material properties, selectivity and particle size have the largest impact on D/F and PR. Particle size and 14 design parameters can be optimized for minimum D/F, maximum PR, or minimum cost on a laptop computer.

show abstract

Standing wave optimization of SMB using a hybrid simulated annealing and genetic algorithm (SAGA)

Cited by 10 publications

References 29 publications

Speedy standing wave design, optimization, and scaling rules of simulated moving bed systems with linear isotherms

Speedy standing wave design, optimization, and scaling rules of simulated moving bed systems with linear isotherms

Multi-objective optimization matching for one-shot multi-attribute exchanges with quantity discounts in E-brokerage

Speedy standing wave design of size-exclusion simulated moving bed: Solvent consumption and sorbent productivity related to material properties and design parameters

Contact Info

Product

Resources

About