Time filter

Source Type

Mirjalili S.M.,Zharfa Pajohesh System ZPS Co.
Applied Optics | Year: 2014

This work proposes a modularized framework for designing the structure of photonic crystal waveguides (PCWs) and reducing human involvement during the design process. The proposed framework consists of three main modules: parameters module, constraints module, and optimizer module. The first module is responsible for defining the structural parameters of a given PCW. The second module defines various limitations in order to achieve desirable optimum designs. The third module is the optimizer, in which a numerical optimization method is employed to perform optimization. As case studies, two new structures called Ellipse PCW (EPCW) and Hypoellipse PCW (HPCW) with different shape of holes in each row are proposed and optimized by the framework. The calculation results show that the proposed framework is able to successfully optimize the structures of the new EPCW and HPCW. In addition, the results demonstrate the applicability of the proposed framework for optimizing different PCWs. The results of the comparative study show that the optimized EPCW and HPCW provide 18% and 9% significant improvements in normalized delay-bandwidth product (NDBP), respectively, compared to the ring-shape-hole PCW, which has the highest NDBP in the literature. Finally, the simulations of pulse propagation confirm the manufacturing feasibility of both optimized structures. © 2014 Optical Society of America.

Mirjalili S.,Griffith University | Mirjalili S.,Queensland Institute of Business and Technology | Mirjalili S.M.,Zharfa Pajohesh System ZPS Co. | Hatamlou A.,Islamic Azad University at Khoy
Neural Computing and Applications | Year: 2016

This paper proposes a novel nature-inspired algorithm called Multi-Verse Optimizer (MVO). The main inspirations of this algorithm are based on three concepts in cosmology: white hole, black hole, and wormhole. The mathematical models of these three concepts are developed to perform exploration, exploitation, and local search, respectively. The MVO algorithm is first benchmarked on 19 challenging test problems. It is then applied to five real engineering problems to further confirm its performance. To validate the results, MVO is compared with four well-known algorithms: Grey Wolf Optimizer, Particle Swarm Optimization, Genetic Algorithm, and Gravitational Search Algorithm. The results prove that the proposed algorithm is able to provide very competitive results and outperforms the best algorithms in the literature on the majority of the test beds. The results of the real case studies also demonstrate the potential of MVO in solving real problems with unknown search spaces. Note that the source codes of the proposed MVO algorithm are publicly available at http://www.alimirjalili.com/MVO.html. © 2015, The Natural Computing Applications Forum.

Mirjalili S.,Griffith University | Mirjalili S.M.,Zharfa Pajohesh System ZPS Co. | Lewis A.,Griffith University
Information Sciences | Year: 2014

The Multi-Layer Perceptron (MLP), as one of the most-widely used Neural Networks (NNs), has been applied to many practical problems. The MLP requires training on specific applications, often experiencing problems of entrapment in local minima, convergence speed, and sensitivity to initialization. This paper proposes the use of the recently developed Biogeography-Based Optimization (BBO) algorithm for training MLPs to reduce these problems. In order to investigate the efficiencies of BBO in training MLPs, five classification datasets, as well as six function approximation datasets are employed. The results are compared to five well-known heuristic algorithms, Back Propagation (BP), and Extreme Learning Machine (ELM) in terms of entrapment in local minima, result accuracy, and convergence rate. The results show that training MLPs by using BBO is significantly better than the current heuristic learning algorithms and BP. Moreover, the results show that BBO is able to provide very competitive results in comparison with ELM. © 2014 Elsevier Inc. All rights reserved.

Abedifar V.,Shahid Beheshti University | Mirjalili S.M.,Zharfa Pajohesh System ZPS Co. | Eshghi M.,Shahid Beheshti University
Optical Fiber Technology | Year: 2015

In all-optical networks the ASE noise of the utilized optical power amplifiers is a major impairment, making the OSNR to be the dominant parameter in QoS. In this paper, a two-objective optimization scheme using Multi-Objective Particle Swarm Optimization (MOPSO) is proposed to reach the maximum OSNR for all channels while the optical power consumed by EDFAs and lasers is minimized. Two scenarios are investigated: Scenario 1 and Scenario 2. The former scenario optimizes the gain values of a predefined number of EDFAs in physical links. The gain values may be different from each other. The latter scenario optimizes the gains value of EDFAs (which is supposed to be identical in each physical link) in addition to the number of EDFAs for each physical link. In both scenarios, the launch powers of the lasers are also taken into account during optimization process. Two novel encoding methods are proposed to uniquely represent the problem solutions. Two virtual demand sets are considered for evaluation of the performance of the proposed optimization scheme. The simulations results are described for both scenarios and both virtual demands. © 2014 Elsevier Inc. All rights reserved.

Mirjalili S.Z.,Zharfa Pajohesh System ZPS Co. | Saremi S.,Griffith University | Saremi S.,Queensland Institute of Business and Technology | Mirjalili S.M.,Zharfa Pajohesh System ZPS Co.
Neural Computing and Applications | Year: 2015

Training feedforward neural networks (FNNs) is considered as a challenging task due to the nonlinear nature of this problem and the presence of large number of local solutions. The literature shows that heuristic optimization algorithms are able to tackle these problems much better than the mathematical and deterministic methods. In this paper, we propose a new trainer using the recently proposed heuristic algorithm called social spider optimization (SSO) algorithm. The trained FNN by SSO (FNNSSO) is benchmarked on five standard classification data sets: XOR, balloon, Iris, breast cancer, and heart. The results are verified by the comparison with five other well-known heuristics. The results prove that the proposed FNNSSO is able to provide very promising results compared with other algorithms. © 2015, The Natural Computing Applications Forum.

Discover hidden collaborations