This thesis is about the tuning and simplification of black-box (direct-search, derivative-free) optimization methods, which by definition do not use gradient information to guide their search for an optimum but merely need a fitness (cost, error, objective) measure for each candidate solution to the optimization problem. Such optimization methods often have parameters that infuence their behaviour and efficacy. A Meta-Optimization technique is presented here for tuning the behavioural parameters of an optimization method by employing an additional layer of optimization. This is used in a number of experiments on two popular optimization methods, Differential Evolution and Particle Swarm Optimization, and unveils the true performance capabilities of an optimizer in different usage scenarios. It is found that state-of-the-art optimizer variants with their supposedly adaptive behavioural parameters do not have a general and consistent performance advantage but are outperformed in several cases by simplified optimizers, if only the behavioural parameters are tuned properly.

Tuning & simplifying heuristical optimization

In this paper, a new memetic algorithm, which combines PSO and local search technique, is proposed for mul-timodal optimization problems. In the investigated algorithm, a local PSO model is used to disperse the individuals into different sub-regions, an adaptive local search method is employed to refine the quality of individuals and a triggered re-initialization scheme is introduced to enhance the algorithm's capacity of solving functions with numerous optima. Experimental results based on a set of benchmark functions show that the proposed memetic algorithm is a good optimizer in multimodal optimization domain.

A memetic particle swarm optimization algorithm for multimodal optimization problems

This work presents a comprehensive insight about simulated moving bed (SMB) technology, from concept to unit operation and process demonstration. A particular relevance is given to SMB nonconventional modes of operation, modeling, and design of SMB-based separations as well as to the construction and operation of SMB units. A conventional separation of a racemic mixture of (S,R)-α-tetralol by means of the FlexSMB-LSRE® unit is addressed to better exemplify a state-of-the-art SMB lab-scale process development and demonstration.

Simulated Moving Bed Chromatography: From Concept to Proof‐of‐Concept

The phase equilibrium modeling for multi-component systems is essential in process systems engineering. In particular, phase stability analysis, Gibbs free energy minimization and estimation of parameters in thermodynamic models are challenging global optimization problems involved in phase equilibrium calculations and modeling for both reactive and non-reactive systems. To date, many significant works have been performed in the area of global optimization, and several algorithms and computational contributions of global optimization have been used for solving these problems; global optimization methods used include both deterministic and stochastic algorithms. To the best of our knowledge, there is no review in the literature that focuses on the global optimization methods and their applications to phase equilibrium modeling and calculations. In this paper, we briefly describe selected deterministic and stochastic optimization algorithms, and then review their use for phase stability analysis, Gibbs free energy minimization and parameter estimation in phase equilibrium models. In short, we provide a general overview of global optimization for modeling and calculating the phase behavior of systems with and without chemical reactions including the prediction of azeotropes and critical points.

/pdf/a-review-on-global-optimization-methods-for-phase-4pjzf8kxts.pdf

Grzegorz Ziomek

Papers

Random search optimization approach for highly multi-modal nonlinear problems

Stochastic optimization of simulated moving bed process: Sensitivity analysis for isocratic and gradient operation