Skip to main content

Über dieses Buch

This carefully edited book takes a walk through recent advances in adaptation and hybridization in the Computational Intelligence (CI) domain. It consists of ten chapters that are divided into three parts. The first part illustrates background information and provides some theoretical foundation tackling the CI domain, the second part deals with the adaptation in CI algorithms, while the third part focuses on the hybridization in CI.

This book can serve as an ideal reference for researchers and students of computer science, electrical and civil engineering, economy, and natural sciences that are confronted with solving the optimization, modeling and simulation problems. It covers the recent advances in CI that encompass Nature-inspired algorithms, like Artificial Neural networks, Evolutionary Algorithms and Swarm Intelligence –based algorithms.



Background Information and Theoretical Foundations of Computational Intelligence


Adaptation and Hybridization in Nature-Inspired Algorithms

The aim of this chapter is to familiarize readers with the basics of adaptation and hybridization in nature-inspired algorithms as necessary for understanding the main contents of this book. Adaptation is a metaphor for flexible autonomous systems that respond to external changing factors (mostly environmental) by adapting their well-established behavior. Adaptation emerges in practically all areas of human activities as well. Such adaptation mechanisms can be used as a general problem-solving approach, though it may suffer from a lack of problem-specific knowledge. To solve specific problems with additional improvements of possible performance, hybridization can be used in order to incorporate a problem-specific knowledge from a problem domain. In order to discuss relevant issues as general as possible, the classification of problems is identified at first. Additionally, we focus on the biological foundations of adaptation that constitute the basis for the formulation of nature-inspired algorithms. This book highlights three types of inspirations from nature: the human brain, Darwinian natural selection, and the behavior of social living insects (e.g., ants, bees, etc.) and animals (e.g., swarm of birds, shoals of fish, etc.), which influence the development of artificial neural networks. evolutionary algorithms, and swarm intelligence, respectively. The mentioned algorithms that can be placed under the umbrella of computational intelligence are described from the viewpoint of adaptation and hybridization so as to show that these mechanisms are simple to develop and yet very efficient. Finally, a brief review of recent developed applications is presented.

Iztok Fister, Damjan Strnad, Xin-She Yang, Iztok Fister

Adaptation in Computational Intelligence


Adaptation in the Differential Evolution

This chapter gives an overview of Differential Evolution (DE), then presents adaptive and self-adaptive mechanisms within the DE algorithm. They can be used in order to make a DE solver more robust, efficient, etc., and to overcome parameter tuning which is usually a time-consuming task needed to be done before the actual optimization process starts. Literature overviews of adaptive and self-adaptive mechanisms are mainly focused on mutation and crossover DE operations, but less on population size adaptation. Some experiments have been performed on benchmark functions to present both the advantages and disadvantages of using self-adaptive mechanisms.

Janez Brest, Aleš Zamuda, Borko Bošković

On the Mutation Operators in Evolution Strategies

Self-adaptation of control parameters is realized in classical evolution strategies (ES) using the appropriate mutation operators controlled by strategy parameters (i.e. mutation strengths) that are embedded into representation of individuals. The mutation strengths determine the direction and the magnitude of the changes on the basis of the new position of the individuals in the search space is determined. This chapter analyzes the characteristics of classical mutation operators, like uncorrelated mutation with one step size and uncorrelated mutation with n step sizes. In line with this, the uncorrelated mutation with


4-dimensional vectors is proposed that beside the mutation strengths utilizes two additional strategy parameters embedded in the 4-dimensional structure used for definition of the change, i.e., shifting the location of normal distribution for a small shift angle and reversing the sign of the change. The promising results conducted on the suite of ten benchmark functions taken from the publications shown that the ES despite their maturity serve as an interesting area of future research.

Iztok Fister, Iztok Fister

Adaptation in Cooperative Coevolutionary Optimization

Cooperative Coevolution (CC) is a typical divide-and-conquer strategy to optimize large scale problems with evolutionary algorithms. In CC, the original search directions are grouped in a suitable number of subcomponents. Then, different subpopulations are assigned to the subcomponents and evolved using an optimization metaheuristic. To evaluate the fitness of individuals, the subpopulations cooperate by exchanging information. In this chapter we review some of the most relevant adaptive techniques proposed in the literature to enhance the effectiveness of CC. In addition, we present a preliminary version of a new adaptive CC algorithm that addresses the problem of distributing efficiently the computational effort between the different subcomponents.

Giuseppe A. Trunfio

Study of Lagrangian and Evolutionary Parameters in Krill Herd Algorithm

Krill Herd (KH) is a novel swarm-based intelligent optimization method developed through the idealization of the krill swarm. In the basic KH method, all the movement parameters used are originated from real nature-driven data found in the literature. The parameter setting based on such data is not necessarily the best selection. In this work, a systematic method is presented for the selection of the best parameter setting for the KH algorithm through an extensive study of arrays of high-dimensional benchmark problems. An important finding is that the best performance of KH can be obtained by setting effective coefficient of the krill individual (



), food coefficient(C


), maximum diffusion speed (D


), crossover probability (C


) and mutation probability (



) parameters to 4.00, 4.25, 0.014, 0.225, and 0.025, respectively. This finding would eliminate the concerns regarding the optimal tuning of the KH algorithm for its most future applications.

Gai-Ge Wang, Amir H. Gandomi, Amir H. Alavi

Solutions of Non-smooth Economic Dispatch Problems by Swarm Intelligence

The increasing costs of fuels and operations of power generating units necessitate the development of optimization methods for economic dispatch (ED) problems. Classical optimization techniques such as direct search and gradient methods often fail to find global optimum solutions. Modern optimization techniques are often meta-heuristic, and they are very promising in solving nonlinear programming problems. This chapter presents a novel method to determine the feasible optimal solutions of the ED problems utilizing the newly developed Bat Algorithm (BA). The proposed BA is based on the echolocation behavior of bats. This technique is adapted to solve non-convex ED problems under different nonlinear constraints such as transmission losses, ramp rate limits, multi-fuel options and prohibited operating zones. Parameters are tuned to give the best results for these problems. To describe the efficiency and applicability of the proposed algorithm, we will use four ED test systems with non-convexity. We will compare our results with some of the most recently published ED solution methods. Comparing with the other existing techniques, the proposed approach can find better solutions than other methods. This method can be deemed to be a promising alternative for solving the ED problems in real systems.

Seyyed Soheil Sadat Hosseini, Xin-She Yang, Amir H. Gandomi, Alireza Nemati

Hybridization in Computational Intelligence


Hybrid Artificial Neural Network for Fire Analysis of Steel Frames

Tuning parameters of artificial neural networks (ANN) is a very complex task that typically demands a lot of experimental work performed by developers. In order to avoid this hard work, the automatic tuning of these parameters is proposed. A real-coded genetic algorithm (GA) was developed for this purpose. This, so-called meta-GA, algorithm acts as a meta-heuristic that searches for the optimal values of ANN parameters using the genetic operators of crossover and mutation and evaluates quality of solutions, obtained after applying the ANN for fire analysis of steel frames. As matter of fact, steel exhibits very unusual wavy behavior which is a very difficult to model by a close form empirical models when heated to the temperatures between 250°C and 600°C. Therefore, the use of ANN was one of the possible solutions which proved to be very promising. However, the results of this ANN with manual parameter setting by an expert can significantly be improved when using the meta-GA for automatic searching the optimal parameter setting of the original ANN algorithm.

Tomaž Hozjan, Goran Turk, Iztok Fister

A Differential Evolution Algorithm with a Variable Neighborhood Search for Constrained Function Optimization

In this paper, a differential evolution algorithm based on a variable neighborhood search algorithm (DE_VNS) is proposed in order to solve the constrained real-parameter optimization problems. The performance of DE algorithm depends on the mutation strategies, crossover operators and control parameters. As a result, a DE_VNS algorithm that can employ multiple mutation operators in its VNS loops is proposed in order to further enhance the solution quality. We also present an idea of injecting some good dimensional values to the trial individual through the injection procedure. In addition, we also present a diversification procedure that is based on the inversion of the target individuals and injection of some good dimensional values from promising areas in the population by tournament selection. The computational results show that the simple DE_VNS algorithm was very competitive to some of the best performing algorithms from the literature.

M. Fatih Tasgetiren, P. N. Suganthan, Sel Ozcan, Damla Kizilay

A Memetic Differential Evolution Algorithm for the Vehicle Routing Problem with Stochastic Demands

This chapter introduces a new hybrid algorithmic approach based on the Differential Evolution (DE) algorithm for successfully solving a number of routing problems with stochastic variables. More precisely, we solve one problem with stochastic customers, the Probabilistic Traveling Salesman Problem and one problem with stochastic demands, the Vehicle Routing Problem with Stochastic Demands. The proposed algorithm uses a Variable Neighborhood Search algorithm in order to increase the exploitation abilities of the algorithm. The algorithm is tested on a number of benchmark instances from the literature and it is compared with a hybrid Genetic Algorithm.

Yannis Marinakis, Magdalene Marinaki, Paraskevi Spanou

Modeling Nanorobot Control Using Swarm Intelligence for Blood Vessel Repair: A Rigid-Tube Model

The future nanorobots for diagnosis and treatment purposes in nano-medicine may exhibit only simple behaviors and work together in their early stage. Through exploring the existing swarm intelligence techniques, the canonical particle swarm optimization was selected to employ for adaptively controlling the locomotion of a swarm system of early-stage nanorobots with only essential characteristics for self-assembly into a structure in a simulations system. In this study, we demonstrated nanorobots operating as artificial platelets for repairing wounds in a simulated human small vessel, which may be used to treat platelet diseases. In a rigid-tube model, we investigated how artificial platelet capabilities including the perception range, maximum velocity and response speed have impacts on wound healing effectiveness. It was found that the canonical particle swarm optimization is an efficient algorithm for controlling the early-stage nanorobots with essential characteristics in both Newtonian and non-Newtonian flow models. The demonstration could be beneficial as guidelines of essential characteristics and useful locomotion control for the realization of nanorobots for medical applications in the future.

Boonserm Kaewkamnerdpong, Pinfa Boonrong, Supatchaya Trihirun, Tiranee Achalakul


Weitere Informationen

Premium Partner

Neuer Inhalt

BranchenIndex Online

Die B2B-Firmensuche für Industrie und Wirtschaft: Kostenfrei in Firmenprofilen nach Lieferanten, Herstellern, Dienstleistern und Händlern recherchieren.



Product Lifecycle Management im Konzernumfeld – Herausforderungen, Lösungsansätze und Handlungsempfehlungen

Für produzierende Unternehmen hat sich Product Lifecycle Management in den letzten Jahrzehnten in wachsendem Maße zu einem strategisch wichtigen Ansatz entwickelt. Forciert durch steigende Effektivitäts- und Effizienzanforderungen stellen viele Unternehmen ihre Product Lifecycle Management-Prozesse und -Informationssysteme auf den Prüfstand. Der vorliegende Beitrag beschreibt entlang eines etablierten Analyseframeworks Herausforderungen und Lösungsansätze im Product Lifecycle Management im Konzernumfeld.
Jetzt gratis downloaden!