Dierential Evolution Research Articles

Optimal design of fractional-order digital integrators:An evolutionary approach

This paper presents an optimal approach to design Fractional-Order Digital Integrators (FODIs) using a metaheuristic technique, called Hybrid Flower Pollination Algorithm (HFPA). HFPA is a hybrid approach which combines the exploitation and exploration capabilities of two di erent evolutionary optimization algorithms, namely, Particle Swarm Optimization (PSO) and Flower Pollination Algorithm (FPA). The proposed HFPA based designs are compared with the designs based on Real Coded Genetic Algorithm (RGA), PSO, Di erential Evolution (DE), and FPA. Simulation results demonstrate that HFPA based FODIs of all the di erent orders consistently achieve the best magnitude responses. The proposed technique yields FODIs which surpass all the designs based on both classical and evolutionary optimization approaches reported in recent literature.

Random adjustment - based Chaotic Metaheuristic algorithms for image contrast enhancement

Metaheuristic algorithm is a powerful optimization method, in which it can solve problemsby exploring the ordinarily large solution search space of these instances, that are believed tobe hard in general. However, the performances of these algorithms signicantly depend onthe setting of their parameter, while is not easy to set them accurately as well as completelyrelying on the problem's characteristic. To ne-tune the parameters automatically, manymethods have been proposed to address this challenge, including fuzzy logic, chaos, randomadjustment and others. All of these methods for many years have been developed indepen-dently for automatic setting of metaheuristic parameters, and integration of two or more ofthese methods has not yet much conducted. Thus, a method that provides advantage fromcombining chaos and random adjustment is proposed. Some popular metaheuristic algo-rithms are used to test the performance of the proposed method, i.e. simulated annealing,particle swarm optimization, dierential evolution, and harmony search. As a case study ofthis research is contrast enhancement for images of Cameraman, Lena, Boat and Rice. Ingeneral, the simulation results show that the proposed methods are better than the originalmetaheuristic, chaotic metaheuristic, and metaheuristic by random adjustment.

Optimization of ethylene oxychlorination fluidized-bed reactor using Differential Evolution (DE) method

The present work aims to employ Di erential Evolution (DE) algorithm to optimize ethylene oxychlorination process to produce 1,2-dichloroethane in a fluidized bed reactor as a feedstock of PVC production. A steady-state reactor model, based on twophase theory of fluidization, was developed to investigate the e ffects of various parameters on C2H4 and HCl conversions. The model's results were compared favorably with theindustrial data obtained from a pilot plant working in Italy. The feed temperature, pressure, HCl and O2 molar ow rates, and cooling medium temperature were selected as decision variables to minimize the objective function subject to the environmental constraints. The highest performance was found at HCl/C2H4 and O2/C2H4 molar ratios of 2 and 0.55, respectively; feed and cooling medium temperatures of 440 and 360 K, respectively; pressure of 367.6 kPa. The results show a decrease of 20C in the feed temperature, which leads to saving energy and reducing the size of the pre-heater.

Feature Selection for Improvement the Performance of an Electric Arc Furnace

Feature selection has as principal goal to nd a representa- tive space of minimal size from original set of larger size. Several research works have been developed on this problem. This paper presents Support Vector Machine-Recursive Feature Elimination (SVM-RFE), Genetic Al- gorithms (GA), and Dierential Evolution (DE) algorithms for feature selection from a database of an Electric Arc Furnace (EAF) for locating variables related to energy consumption. The proposal suggests merging the coecients generated by LDA and SVM, employing them in RFE to obtain the ranks for each discriminant variables. The measure of accuracy and error rate for each algorithm is presented like a decisive score for choosing the subset obtained by the algorithm with the best performance. The variables selected were adjusted for the EAF control system achieving the reduction of the energy consumption to 3.5% in a steel castings and 1 minute reduction of the connected EAF in a steel casting.

Layout Optimization of Braced Frames Using Differential Evolution Algorithm and Dolphin Echolocation Optimization

In this study, topology optimization is applied to concentrically braced frames in order to find economical solutions for conventional structural steel frames. Dierential Evolution Algorithm and Dolphin Echolocation Optimization are applied for structural optimization. Numerical examples are studied and results of comparison with other meta-heuristic algorithms, including Genetic Algorithm, Ant colony optimization, Particle Swarm, and Big Bang-Big Crunch are presented.

Induction motor parameter estimation using metaheuristic methods

The steady-state equivalent circuit parameters of an induction motor can be estimated using the operation characteristics that are provided by manufacturers. The characteristics of the motor used in estimation methods are the starting, maximum, and nominal torque values; the power factor; and eciency. The operation characteristics of a motor given in data sheets are generally based on design parameters and are not suitable with real values. For this reason, in this paper, the data used in the parameter estimation for induction motors are taken from the literature. Using an optimization method for parameter estimation is useful for comparing the manufacturer values and values at the end of estimation, as well as minimizing the error in between. There are many methods in the literature for the parameter estimation of induction motors. In this study, the estimation is made using the charged system search (CSS), dierential evolution algorithm (DEA), particle swarm optimization, and genetic algorithm optimization techniques. The CSS algorithm is rst applied for estimation of the parameters of an induction motor. The results obtained from all of the methods show that the CSS algorithm is suitable with the DEA. From the obtained results, it is understood that an exact approach can be made to equivalent circuit parameters in case the values given by the manufacturer model the motor properly.

SECURING INTERPRETABILITY OF FUZZY MODELS FOR MODELING NONLINEAR MIMO SYSTEMS USING A HYBRID OF EVOLUTIONARY ALGORITHMS

In this study, a Multi-Objective Genetic Algorithm (MOGA) is utilized to extract interpretable and compact fuzzy rule bases for modeling nonlinear Multi-input Multi-output (MIMO) systems. In the process of non- linear system identication, structure selection, parameter estimation, model performance and model validation are important objectives. Furthermore, se- curing low-level and high-level interpretability requirements of fuzzy models is especially a complicated task in case of modeling nonlinear MIMO systems. Due to these multiple and conicting objectives, MOGA is applied to yield a set of candidates as compact, transparent and valid fuzzy models. Also, MOGA is combined with a powerful search algorithm namely Dierential Evolution (DE). In the proposed algorithm, MOGA performs the task of membership function tuning as well as rule base identication simultaneously while DE is utilized only for linear parameter identication. Practical applicability of the proposed algorithm is examined by two nonlinear system modeling prob- lems used in the literature. The results obtained show the eectiveness of the proposed method.

On a nonlinear integrodifferential evolution inclusion with nonlocal initial conditions in Banach spaces

In this paper, we discuss the existence results for a class of nonlinear integro- dierential evolution inclusions with nonlocal initial conditions in Banach spaces. Our results are based on a fixed point theorem for condensing maps due to Martelli and the resolvent operators combined with approximation techniques.

Differential Evolution to Solve the Lot Size Problem in Stochastic Production Systems

An Advanced Resource Planning model is presented to support optimal lot size decisions for performance improvement of a production system in terms of either delivery time or setup related costs. Based on a queueing network, a model is developed for a mix of multiple products following their own specic sequence of operations on one or more resources, while taking into account various sources of uncertainty, both in demand as well as in production characteristics. In addition, the model includes the impact of parallel servers and dierent time schedules in a multi-period planning setting. The corrupting inuence of variabilities from rework and breakdown is explicitly modeled. As a major result, the dierential evolution algorithm is able to nd the optimal lead time as a function of the lot size. In this way, we add a conclusion on the debate on the convexity between lot size and lead time in a complex production environment. We show that dierential evolution outperforms a steepest descent method in the search for the global optimal lot size. For problems of realistic size, we propose appropriate control parameters for the dierential evolution in order to make its search process more ecient.