Robustness Of Genetic Algorithm Research Articles

Thermodynamic representation of ionic liquids phase equilibrium with PDH-ASOG and PDH-UNIFAC models

The systematic study of thermodynamic properties for understanding liquid-liquid equilibrium (LLE) behavior of Ionic Liquids (ILs) is an important research direction. The predictive thermodynamic models are useful tools in this area. However, the optimization problem in this kind of predictive modelling can be very challenging. This work proposes a hybrid optimization method for handling such problems. This method benefits from the robustness of Genetic Algorithm (GA) as a global search technique while complementing it with local search abilities of Nelder-Mead Algorithm (NMA). The proposed algorithm showed quite promising results in estimating the interaction parameters of four thermodynamic models: UNIFAC, ASOG, PDH-UNIFAC and PDH-ASOG. In these models, the long-range contribution is accounted by Pitzer–Debye–Hückel (PDH) equation while the short-range contribution is considered by nonelectrolyte UNIFAC and ASOG. Particularly, LLE phase behavior of two ILs comprising of sulfate-based anions and imidazolium cations is examined for 44 ternary (421 tie-line) systems and the full set of computed parameters by the proposed hybrid method is reported.

Uncertain multi-objective programming model: a genetic algorithm approach

This paper aims at describing an uncertain multi-objective programming model involving uncertain variables with genetic algorithm approach. In this paper, the uncertain multiobjective programming model is converted into an equivalent crisp mathematical programming model. Then, a genetic algorithm is proposed to search the Stackelberg-Nash equilibrium of the uncertain multiobjective programming model with supporting numerical illustrations. Finally, sensitivity analysis study is carried out over parameters of algorithm and solution obtained to show efficiency and robustness of genetic algorithm for uncertain multiobjective programming model.

A New Method to Generate Test Cases by Genetic Algorithm

Because of the powerful capability of global searching and robustness of genetic algorithm, so it can be well applied in the automated generation of test data.Establish a module to generate test cases automatically by genetic algorithms.We can find that genetic algorithm guide the generation of the test data by the evaluation function according to the constraints of the path.It will not only avoid the blindness of the data searching process with higher efficiency,but also need not to consider the matching problem of the generated test data and test sequences.It can even run effectively relying only on the fitness of the point in the searching space supplied by the evaluation function.This advantage makes it a powerful searching algorithm on the type of random searching.

Optimization of reaction parameters of acidolysis reaction between mustard oil and capric acid by using Thermomyces lanuginosus lipase.

Structured lipids were prepared from mustard oil by enzymatic acidolysis reaction with capric acid (C10) using lipase enzyme TLIM from Thermomyces lanuginosus as biocatalyst. Parameters such as substrate molar ratio, enzyme concentration, reaction temperature, stirring speed and time of maximum incorporation, were studied for the optimization of the reaction. The optimized set of process conditions was predicted by response surface methodology (RSM) and genetic algorithm (GA). The robustness of GA and RSM was evaluated using regression coefficient and p value. The R(2) found out by GA was 0.996 while from RSM was 0.973. The results proved that GA models have better performance than RSM models. From the result, it could be concluded that optimal conditions for synthesis of capric acid rich mustard oil were: Temperature = 39.5°C ; time = 21.1hr; Substrate ratio = 3.5; Enzyme content = 8.8%; Speed = 570.8rpm.

Implementation of different encoding types on structural optimization based on adaptive genetic algorithm

In this study, the use of new alternatives of encoding types such as quaternary encoding and octal encoding is reviewed along with that of binary and value encodings and how they provide contribution to efficiency and robustness of genetic algorithm is discussed in structural problems. In addition to new alternatives of encoding types, a genetic algorithm with adaptive manner is presented so that an adaptive approach including adaptive mutation and adaptive crossover operators is employed. It is concluded that it is not possible to tell that one of the encoding types is exactly dominant over the others in all aspects such as convergence, finding the optimum solution, and iteration number. However, it is worth to say that to run genetic process with different encoding types should be considered since one of the runs can give more appropriate solution than that of the others.

Searching for critical failure surface in slope stability analysis by using hybrid genetic algorithm

The radius and coordinate of sliding circle are taken as searching variables in slope stability analysis. Genetic algorithm is applied for searching for critical factor of safety. In order to search for critical factor of safety in slope stability analysis efficiently and in a robust manner, some improvements for simple genetic algorithm are proposed. Taking the advantages of efficiency of neighbor-search of the simulated annealing and the robustness of genetic algorithm, a hybrid optimization method is presented. The numerical computation shows that the procedure can determine the minimal factor of safety and be applied to slopes with any geometry, layering, pore pressure and external load distribution. The comparisons demonstrate that the genetic algorithm provides a same solution when compared with elasto-plastic finite element program.

Application of the Genetic Algorithm Joint with the Powell Method to Nonlinear Least-Squares Fitting of Powder EPR Spectra

The application of the stochastic genetic algorithm (GA) in conjunction with the deterministic Powell search to analysis of the multicomponent powder EPR spectra based on computer simulation is described. This approach allows for automated extraction of the magnetic parameters and relative abundances of the component signals, from the nonlinear least-squares fitting of experimental spectra, with minimum outside intervention. The efficiency and robustness of GA alone and its hybrid variant with the Powell method was demonstrated using complex simulated and real EPR data sets. The unique capacity of the genetic algorithm for locating global minima, subsequently refined by the Powell method, allowed for successful fitting of the spectra. The influence of the population size, mutation, and crossover rates on the performance of GA was also investigated.