Pravesh Kumar

Bi-level thresholding using PSO, Artificial Bee Colony and MRLDE embedded with Otsu method

Image segmentation is required to be studied in detail some particular features (areas of interest) of a digital image. It forms an important and exigent part of image processing and requires an exhaustive and robust search technique for its implementation. In the present work we have studied the working of MRLDE, a newly proposed variant of differential evolution combined with Otsu method, a well known image segmentation method for bi-level thresholding. The proposed variant, termed as Otsu+MRLDE, is tested on a set of 10 images and the results are compared with Otsu method and some other well known metaheuristics.

Enhanced mutation strategy for differential evolution

It is well known that mutation plays a very important role in the successful performance of Differential Evolution (DE) algorithm. The proposed scheme named Modified Random Localization (MRL) is based on strategically selecting the individuals from the entire search space rather than choosing them randomly as in basic DE. The corresponding DE variant named MRL-DE is analyzed on a set of 8 traditional benchmark functions and 6 nontraditional shifted functions. Numerical and statistical results indicate the competence of the proposed MRL-DE for solving unconstrained global optimization problems.

Noisy source recognition in multi noise plants by differential evolution

Since last few decades differential evolution algorithm (DE) has been successfully applied for solving many real life optimization problems. In this paper DE is applied to identifying the location of noisy sources in a multi noise plants. A trail noise technique is used to obtain the variation between trial sound pressure level (SPL) and exact SPL at monitoring points and then DE is employed in conjunction with the method of minimized variation square in seeking for the best locations and sound power level (SWLs). The results reveal that the significant locations and SWLs of noises can be precisely identified by DE.

Modified random localisation-based DE for static economic power dispatch with generator constraints

In the present study, a modified DE variant called MRLDE is used for solving economic dispatch problem. It is a highly constrained non linear optimisation problem which can be a challenge for traditional gradient based methods. Numerical results and comparison show the efficiency and robustness of the MRLDE algorithm for dealing with such problems.

Two enhanced Differential Evolution variants for solving global optimization problems

Differential Evolution (DE) algorithms are very robust, effective and highly efficient in solving the global optimization problems. Thus, they are usually able to mitigate the drawback of long computation times commonly associated with Evolutionary algorithms. However, in certain cases the performance of DE is observed not to be completely flawless. In this paper we have proposed the two enhanced variants of DE using a modified mutation operator. The DE versions named as EDE-1 and EDE-2 are tested on six benchmark problems and a real time molecular potential energy problem. The simulation results prove the efficiency as well as the effectiveness of the proposed variants.

Information preserving selection strategy for Differential Evolution algorithm

Differential Evolution (DE) is a popular technique for solving real parameter global optimization problems. Several variants of DE are proposed in literature which aims at further strengthening its performance for solving complex problems. In the present study we suggest a simple and efficient modification in the selection strategy of basic DE. The proposed strategy is named Information Preserving (IP) selection strategy. It makes use of most of the information that is generated during the different phases of DE. The proposed IP scheme is embedded in the structure of basic DE and also in DERL, another variant of DE. The numerical results indicate that the inclusion of proposed scheme significantly improves the performance in terms of convergence rate while maintaining the solution quality.

A Self Adaptive Differential Evolution Algorithm for Global Optimization

This paper presents a new Differential Evolution algorithm based on hybridization of adaptive control parameters and trigonometric mutation. First we propose a self adaptive DE named ADE where choice of control parameter F and Cr is not fixed at some constant value but is taken iteratively. The proposed algorithm is further modified by applying trigonometric mutation in it and the corresponding algorithm is named as ATDE. The performance of ATDE is evaluated on the set of 8 benchmark functions and the results are compared with the classical DE algorithm in terms of average fitness function value, number of function evaluations, convergence time and success rate. The numerical result shows the competence of the proposed algorithm.

Modified Single Array Selection Operation for DE Algorithm

In this study, a modified selection operation is proposed for differential evaluation (DE) algorithm. The proposed selection strategy called information utilization (IU) strategy and the proposed DE variant called IUDE reuse redundant trial vectors embedded with single array selection strategy. The proposed selection strategy is implemented on DERL and MRLDE, the enhanced DE variants and the corresponding algorithms are termed IU-DERL and IU-MRLDE. Six traditional functions are taken for experiments. Results confirm that the proposed selection strategy is helpful in amplifying the convergence speed.

Control parameters and mutation based variants of differential evolution algorithm.

Differential Evolution (DE) is considered as a simple yet influential search engine used for optimization of realvalued, multimodal and nonlinear functions. Here two new variants of the parent DE are presented with self-tuned control parameters and a modified mutation scheme. The First variant, designed by applying self-adaptive control parameters to the parent DE, is named as CPDE and the second one, the enhanced version of CPDE in which CPDE is embedded with a modified mutation scheme, is named as CPMDE. The crucial role of the self-adaptive control parameters can’t be abandoned, which is used to improve the quality of the solution. The next application of self-adapted control parameters is drawn with a modified mutation operation in which the whole search space is divided into three equal parts for the sake of maximum exploration of the search space. A set of 14 traditional and 12 non-traditional (6 shifted and 6 hybrid) test problems is chosen for validation of the performance of the proposed algorithms which is then compared with the parent DE and some other variants of DE in terms of number of function evaluations, CPU time, error and standard deviation. Numerical and statistical results show that the proposed algorithm helps in providing a better trade-off between convergence rate and efficiency.

MRLDE for solving engineering optimization problems

MRLDE is an enhanced version of Differential Evolutionary (DE) algorithm which adapts a modified approach for selecting individuals for mutation operation. MRLDE has been successfully tested on various benchmark problems and some real life applications. The objective of this paper is to solving constrained engineering optimization problems using MRLDE. The algorithm is validated using four engineering optimization problems, taken from literature. Pareto-ranking method is used to handle constrained with proposed variant. Results and comparison show the efficiency and effectiveness of the proposed MRLDE for solving such types of problems.