Swarm Intelligence Volume 2: Innovation, new algorithms and methods
Swarm Intelligence (SI) is one of the most important and challenging paradigms under the umbrella of computational intelligence. It focuses on the research of collective behaviours of a swarm in nature and/or social phenomenon to solve complicated and difficult problems which cannot be handled by traditional approaches. Thousands of papers are published each year presenting new algorithms, new improvements and numerous real world applications. This makes it hard for researchers and students to share their ideas with other colleagues; follow up the works from other researchers with common interests; and to follow new developments and innovative approaches. This complete and timely collection fills this gap by presenting the latest research systematically and thoroughly to provide readers with a full view of the field of swarm. Students will learn the principles and theories of typical swarm intelligence algorithms; scholars will be inspired with promising research directions; and practitioners will find suitable methods for their applications of interest along with useful instructions. Volume 2 includes 17 chapters covering frontedge research with novel and newly proposed algorithms and methods. The companion volume 1 covers principles of swarm intelligence, current algorithms and methods; and volume 3 covers applications of swarm intelligence algorithms. With contributions from an international selection of leading researchers, Swarm Intelligence is essential reading for engineers, researchers, professionals and practitioners with interests in swarm intelligence.
Inspec keywords: swarm intelligence; optimisation; predatorprey systems
Other keywords: fireworks algorithm; interactive chaotic evolution; inclined planes system optimisation; binary mothflame optimisation algorithm; grey wolf optimisation; predatorprey optimisation; binary whale optimisation; firefly algorithm; ant lion optimiser; swarm intelligence; glowworm swarm optimisation
Subjects: Optimisation techniques; Expert systems and other AI software and techniques; Artificial intelligence
 Book DOI: 10.1049/PBCE119G
 Chapter DOI: 10.1049/PBCE119G
 ISBN: 9781785616297
 eISBN: 9781785616303
 Page count: 544
 Format: PDF

Front Matter
 + Show details

Hide details

p.
(1)
 + Show details


1 Standard fireworks algorithm 2017
 + Show details

Hide details

p.
1
–26
(26)
Fireworks algorithm (FWA) [1] is a novel evolution algorithm developed since 2010. Unlike other populationbased algorithms, individuals in FWA cooperate to control their behavior and allocation of computation resource. Instead, most algorithms like particle swarm optimization (PSO) concentrate on the moving of populations. Firework algorithm was well studied in these years. There has been theoretical analysis, engineering applications and algorithm improvements in FWA, making it a wellknown and competitive optimization method. Plenty of researchers are trying to test and adjust some old operators or introduce new ones in order to improve FWA's performance. However, these works' contribution to the further development of FWA is indeed limited for several reasons. The most common problem is the selection of benchmark functions. Some researchers chose oldfashioned functions, and some of them designed their own. Besides, some researchers did not explain how they finetuned the hyperparameters of the old algorithms in detail. In experiments, some researchers chose a relatively old version of FWA as the base of their new algorithm, even though some parts of these old algorithms have already been proved inefficient. So, their methods might not be valid for the later versions of FWA. When some researchers designed a set of operators, sometimes, they neglect to test each combination of the new and the old ones. Thus, we do not know how each operator works and how they cooperate with each other. So, we gathered some works on FWA and did a series of experiments on the latest benchmark functions of CEC2017 [2], hoping to get a set of operators that is both simple and effective (do not have to perform best), and form the standard FWA in 2017. On the one hand, this would help researchers to stop worrying about on which version of FWA they should implement their ideas and on which they should test them. And on the other hand, this is a summary of the development of FWA and the experiments can reveal some of the abilities of those methods. The remainder of this chapter is organized as follows. Section 1.1 presents the principle of FWA and some developments on it, which will be applied later in the experiments. Section 1.2 explains our experiment procedures and settings. Section 1.3 analyses the results of experiments. Section 1.4 concludes the chapter and discusses the further researches on FWA.
 + Show details


2 Guided fireworks algorithm applied to multilevel image thresholding
 + Show details

Hide details

p.
27
–58
(32)
Digital image segmentation is often used as a preprocessing step in image analysis where different objects in the image need to be separated. One of the most used techniques for image segmentation is multilevel thresholding. Different criteria are used for determination of optimal thresholds, but in all cases, multilevel thresholding is a hard optimization problem that cannot be solved by deterministic methods. Swarm intelligence stochastic optimization metaheuristics have been proven to be very successful for such problems. In this chapter, we propose adjusted guided fireworks algorithm (GFWA) for multiple optimal threshold determination based on Kapur's entropy, Otsu's criteria and Tsallis' entropy. Comparison with other stateoftheart techniques shows that the GFWA exhibits superior performance.
 + Show details


3 Credit card number encryption using fireworkbased key generation
 + Show details

Hide details

p.
59
–82
(24)
Encryption of credit card numbers (CCNs) is essential since it is vulnerable to eavesdropping when stored in the database or during transfer in the network. Security of data is the necessity of the hour since communications over open network occur frequently. A novel algorithm termed credit card number encryption (CCE) is proposed to encrypt the CCN. The algorithm works in two phases. In the first phase, the CCN is represented as a binary image and the binary image is encoded as digits. The second phase involves the design of a stream cipher method to encrypt the encoded image using a fireworks key generation credit card number encryption (FWKGCNE) algorithm for key generation. One outstanding advantage of CCE algorithm is the reduction in the number of keys to be stored and distributed. Experimental results demonstrating encryption of CCNs using CCE algorithm and the comparison with the existing methods are presented. Simulation and analysis results show that CCE algorithm is greatly sensitive to the keys and the algorithm has a large space of keys. Security analysis of CCE algorithm was performed to demonstrate that the algorithm is resistant to image attacks.
 + Show details


4 ST (ShafiabadyTeshnehlab) optimization algorithm
 + Show details

Hide details

p.
83
–110
(28)
ShafiabadyTeshnehlab (ST) optimization algorithm is a local swarm intelligence algorithm that has been inspired from the motion of the molecules in the air. Similar to all the other swarm optimization algorithms, the mentioned algorithm uses iterative approach by updating the values of the cells in each particle. This method is superior to conventional optimization algorithms because of its capability in finding the local minimum in very few and incomparably less numbers of iterations relative to other local optimization methods; hence, ST optimization algorithm leads to faster decisionmaking speed. The other specification of this algorithm is the precision and accuracy of the results in comparison with the algorithms in its own group. In addition, this algorithm has the ability to perform the optimization task accurately when dealing with several unknown values simultaneously; hence, increasing the dimensions of the search space does not deteriorate the optimization results like the other conventional algorithms. The only shortcoming of ST optimization algorithm is its local nature that makes it sensitive to the initial values that represent the particles in the search space. The various advantages of ST optimization method make it an appropriate local optimization algorithm.
 + Show details


5 Predatorprey optimization with heterogeneous swarms
 + Show details

Hide details

p.
111
–144
(34)
Predatorprey optimization algorithms use the interaction between predator and prey particles to control the balance between local and global search in particle swarm optimization. Since their introduction in 2002, predatorprey optimizers have been successfully applied to many practical problems, frequently outperforming other particle swarm algorithms. In this chapter, we will start by presenting the original predatorprey optimizing algorithm and to review some of its applications. We will then describe the most recent version of the algorithm, the scouting predator prey optimizer, where scout particles are proposed as a mechanism to introduce new exploratory behaviors in the new heterogeneous swarm. Scout particles can be used to improve the predatorprey algorithm in different ways, from integrating previous knowledge to increase performance in specific problems to introducing new heuristics that globally improve the algorithm. We illustrate the effect of using different scout particles by empirically comparing the performance of several variants of the scouting predatorprey optimizer on a large set of benchmark problems, carefully chosen to present the algorithms with different challenges. Finally, the scouting predatorprey algorithm will be compared with several particle swarm optimizers and differential evolution algorithms to investigate how competitive the algorithm is with stateoftheart particle swarm and evolutionary optimizers.
 + Show details


6 A novel modified ant lion optimizer algorithm: extension to proposed 4DTC
 + Show details

Hide details

p.
145
–184
(40)
In this chapter, a novel algorithm named as modified ant lion optimizer (MALO) algorithm has been suggested for the improvement of convergence behavior of the existing ALO algorithm. Moreover, a new form of turbo code (TC) termed as 4DTC has also been projected for the enhancement of the minimum Hamming distance of the existing TCs. Thereafter, the MALO algorithm has been judiciously used for the design of power efficient 4DTC. From the simulated outcomes, it is obvious that the highgrade act of MALO algorithm is pretty remarkable than the existing ALO as well as other optimization procedures like particle swarm optimization, Cuckoo search and Harmony search in the matter of exactness and confluence standpoint. Additionally, it has been witnessed that the projected modification on ALO algorithm provides the best optimum value close to theoretical limit by adopting fast exploration and exploitation process over the other algorithms. At the same time, it has also been established that the projected 4DTC offers substantial improvement in BER act over the others like parallel concatenated convolution turbo code (PCCTC), serially concatenated convolution turbo code (SCCTC), and 3DTC. The supremacy of MALO has been observed for eight states as well as 16 states 4DTCs over the other variance like PCCTC, SCCTC, 3DTC and proposed 4DTC by means of bit error rate (BER) performance by allocating the optimum power to the systematic and parity bits. Analogous enhancements in BER act have been detected with puncturing as well.
 + Show details


7 Pushpull glowworm swarm optimization algorithm for multimodal functions
 + Show details

Hide details

p.
185
–219
(35)
Glowworm swarm optimization (GSO) is a wellestablished swarmintelligencebased optimization technique mainly used for identifying peaks of all local optima of multimodal functions rather than just the global optima as done by most other similar algorithms. The main application of GSO is in cases where the optima are created by sources of certain signals, which interact to create these peaks in the measured signal profile. As there is a greater likelihood of the peaks being close to the individual sources, GSO can identify such sources. It has been shown that GSO can be used to implement signal source seeking behavior in multirobot systems. The name GSO comes from the specific characteristics of glowworms in nature, which is used in the algorithm, to get attracted to other glowworms which glow brighter. Several researchers have used GSO for various applications and some of them have modified GSO to obtain better performance. In this chapter, we present a detailed study of the pushpull glowworm swarm optimization algorithm, which is a variant based on an extension of the basic philosophy of attraction in GSO. Each glowworm encodes the fitness of its current location, evaluated using the given objective function, into a luciferin value that is accessible to its neighbors. A glowworm in the proposed pushpull version of the algorithm identifies two disjoint adaptive neighborhoods, containing glowworms with higher luciferin value and lower luciferin value, for deciding its direction of movement by probabilistically selecting one neighbor from each of the neighborhoods. The lower luciferin value neighbor provides the “push” and the higher luciferin one provides the “pull.” In the originally proposed basic GSO, only the pull action was considered. This exploitation of local information and strategic neighbor interaction from two adaptive neighborhoods enables the swarm of glowworms to partition into disjoint subgroups and converge to the multiple optima of a given multimodal function with a much smoother trajectory and faster convergence. Experimental results are presented to demonstrate the efficacy of the proposed algorithm.
 + Show details


8 Firefly algorithm and its applications
 + Show details

Hide details

p.
221
–250
(30)
In recent years, swarm intelligence optimization algorithm is a research hotspot in the area of computational intelligence and artificial intelligence. And its application has already penetrated into many fields. In many kinds of intelligent algorithms, the firefly algorithm (FA) is a relatively novel algorithm and shows excellent performance. In this chapter, we introduced FA algorithm and its applications. First, we described the basic concepts of the FA algorithm. Then, we improved the FA by combining the characteristics of specific problems to solve various problems such as numerical optimization, clustering analysis and protein complexes discovering on proteinprotein interaction network. We gave the detailed implement steps and the comparing results to show the feasibility of FA and extensive of its applications.
 + Show details


9 The optimization dialectical method for the multiple sequences alignment problem
 + Show details

Hide details

p.
251
–263
(13)
Multiple sequence alignment (MSA) of biological sequences like DNA and proteins is one of the most important problems in bioinformatics, fundamental for the construction of phylogenetic trees, which are useful to establish evolutionary relationships among individuals and species. From the use of MSA methods, phylogenetic analysis could be conducted in order to reveal shared evolutionary origins. However, it is a very complex computational problem. Dialectical optimization is an evolutionary method designed to solve optimization and search problems using a socialevolutionary metaheuristic, based on the interaction of poles in a generation of solution candidates. Poles interact with each other along historical and crisis stages, in such a way that population sizes vary from one historical phase to another. Herein this work, we propose a dialectical approach to solve iteratively MSA problems, considering these problems as optimization tasks. We also propose an objective function based on some biological and computational constraints, in order to obtain feasible and biologically significant alignments. The results were compared with Clustal, a stateoftheart MSA method, and proved to be reasonably useful, once alignment performances were comparable and, in some cases, our approach reaches superior scores. Our proposed method is also able to improve Clustal results using them as seeds for the dialectical optimization method.
 + Show details


10 A new binary mothflame optimization algorithm (BMFOA)  development and application to solve unit commitment problem
 + Show details

Hide details

p.
265
–291
(27)
A binary variant of mothflame optimizer, namely the binary mothflame optimizer algorithm (BMFOA) is developed in this chapter and is applied to solve unit commitment (UC) problem in power system operation. The mothflame algorithm is a bioinspired optimization algorithm that mimics the traverse navigation mechanism of moth around flames. The navigation mechanism is modelled as a spirally converging approach of moth towards flame. However, the direct application of realvalued mothflame optimization algorithm (MFOA) to binary matured problems such as UC problem is not possible considering the binary search space attributes. Thus, a binary variant BMFOA is developed via modified sigmoidal transformation of realvalued MFOA. The efficacy of proposed BMFOA is demonstrated through numerical experiments using test systems of different sizes ranging from smalltomedium and large scale. The simulation results are presented, discussed and compared to various existing approaches to solve UC problems. In addition, the statistical significance of BMFOA with respect to other existing approaches is established by performing standard statistical tests such as Friedman, Friedman aligned ranks test, Wilcoxon pairwise test and Quade test. The comparison of statistical test results confirms the statistical significance of proposed BMFOA for solving UC problem of different scales.
 + Show details


11 Binary whale optimization algorithm for unit commitment problem in power system operation
 + Show details

Hide details

p.
293
–328
(36)
This chapter discusses the development and application of an intelligent computational technique called binary whale optimization algorithm (BWOA) and its application to solve the unit commitment (UC) problem. The whale optimization is a heuristic approach that mimics the intelligence associated with hunting and feeding behaviour of whales. The two distinct properties of location updates of whales, namely shrinking approach and spiral update approach are used for optimizing the position of prey. To improvise the realvalued whale optimization algorithm for binary UC problem, update process of whale position is mapped to binary search space using various transfer functions. The binary variants include three sigmoidal transformations and two tangent hyperbolic transformations. The binary variants presented are evaluated using extensive numerical experiments on various test systems and operating conditions. The simulation results are presented and compared to various existing classical/traditional, heuristic and metaheuristic approaches. In addition, the statistical significance of proposed BWOA approaches among other binary approaches and within themselves is verified using a series of standard statistical tests. The same demonstrates the effectiveness of proposed BWOA to solve UC problem of small, medium and large scale.
 + Show details


12 Realcoded grey wolf optimisation algorithm for progressive thermal power system unit commitment
 + Show details

Hide details

p.
329
–361
(33)
The thermal unit commitment (TUC) in regulated and competitive electric power markets aims to minimise the systemwide operational costs of generators by providing an optimal power generation schedule so that the forecasted power demand should be equalled. This process is formulated mathematically as a nonlinear, largescale, mixedinteger combinatorial optimisation problem, which is quite difficult due to its inherent highdimensional, nonconvex, discrete and nonlinear nature. Nowadays, the inclusion of pollutant emission, ramp rate limits and forced outage rate in the TUC problem is being an appreciative insight in the fossilfuel scarcity scenario. To address all these aspects, the progressive TUC (PTUC) process is developed which increases further the complexity in finding the best feasible solution. Recently, in the field of evolutionary computations, an innovative optimisation algorithm, namely grey wolf optimisation (GWO), has been developed by inspiring the behaviour of grey wolves. GWO does not have any affinity to stick in local optimum points in the complex multimodal optimisation problem, and it provides a more diverse search of the solution space. In order to handle the operational constraints of PTUC problem effectively, the realcoded scheme is introduced in GWO, and in this context, real coded GWO (RCGWO) is implemented to solve the various PTUC problems under single and multiobjective frameworks. The realcoded scheme is developed using the load curve, and it seems to be particularly natural when tackling optimisation problems of parameters with variables in continuous domains. The standard and practical systems are used for demonstration, and the performance analysis of RCGWO confirms that the RCGWO is robust and consistent in finding the ever reported best feasible TUC solution. The RCGWO provides more reliable unit commitment (UC) decisions on fuel consumption, emission allowance, reliable operation of power system and longterm utilisation of generating units. This could devise a set of corrective/preventive control actions for the secure, reliable, social welfare and economical operation of power generation systems. Hence, the RCGWObased UC solutions enable the utility to obtain an extra value and cope easier with the demands of energy economics.
 + Show details


13 Application of grey wolf optimization in fuzzy controller tuning for servo systems
 + Show details

Hide details

p.
363
–387
(25)
This chapter presents aspects concerning the tuning of fuzzy controllers (FCs) by grey wolf optimization (GWO) algorithms with focus on costeffective TakagiSugeno proportionalintegral fuzzy controllers (TS PIFCs). GWO is one of the latest swarm intelligence algorithms, which has been developed by mimicking grey wolf social hierarchy and hunting habits. TS PIFCs are applied to servo systems, represented as nonlinear processes characterized by secondorder dynamics with an integral component, variable parameters, a saturation and deadzone static nonlinearity. The variable parameters of the processjustify the need to design fuzzy control systems with a reduced process parametric sensitivity. Four optimization problems are defined with this regard, with the tuning parameters ofTS PIFCs considered as vector variables and with objective functions that include the weighted output sensitivity function of the state sensitivity model with respect to process parametric variations. GWO is next employed in the minimization of these objective functions. Simulation and experimental results are given for a case study that deals with the optimal tuning of TS PIFCs for the angular position control of a laboratory nonlinear servo system. The process gain is variable, and fuzzy control systems with a reduced process gain sensitivity are offered.
 + Show details


14 Smart swarm inspired algorithms for microwave imaging problems
 + Show details

Hide details

p.
389
–416
(28)
In the last decade smart swarm inspired optimization algorithms, such as particle swarm optimizer (PSO) [1], ant colony optimizer (ACO) [2], firefly optimizer (FFO) [3], bacteria foraging optimizer (BFO) [4], honey bees optimizer (HBO) [5] and others algorithms [6,7], have been successfully adopted as a powerful optimization tools in several areas of applied engineering, and they demonstrated their advantages and superiority with respect to optimizers based on natural competition such as genetic algorithms (GAs) [8], different evolution (DE) [9] and their customizations. The development of CAD tools based on smart swarm optimizer (SSO) could provide the researchers, engineers and industrial designers with powerful tools that can be the solution for the industrial market since they permit to reduce the time to market of a specific device keeping the commercial predominance. It is worth noticing that these tools do not require expert engineers and they can reduce the computational time typical of the standard trial errors methodologies. This chapter is aimed at presenting an overview of smart swarm inspired optimization algorithms (SSOs) as applied to the solution of complex engineering problems. The overview starts from the wellknown GAs up to recent collaborative optimizers based on intelligent swarms and inspired by nature. In particular, SSOs are mimic the behaviour of insects, birds, bats or flock of fishes, searching for food. The goal of this chapter is on the use, calibration and the capabilities assessment of different kind of smart swarms based optimization algorithms for the solution of complex engineering problems. In order to apply a smart swarm inspired algorithm an engineering problem is usually recast into a global optimization problem. Formulated in such a way, complex problems can be efficiently handled by smart swarm inspired optimizer by defining a suitable cost function that represent the distance between requirements and the trial solution. The effectiveness of a given trial solution can be analysed with numerical methodologies, such as finite element method (FEM), finite difference time domain (FDTD), method of moment (MoM) simulator, and then compared with the initial requirements. As a common feature, these environments usually integrate an optimizer and commercial numerical simulators. In particular, this chapter describes how to solve a set of electromagnetic problems, typically characterized by high unknown number, and strong nonlinearities. An example of complex electromagnetic problem is a typical microwave imaging application [10] or the control [11] and design [12] of complex radiating structures, the calibration of microwave systems and other interesting practical applications [13]. The first two section deal with an accurate description of competitive algorithms [such as GAs, differential evolution (DE) and their customization] versus collaborative algorithms [namely PSO [1], ACO [2], FFO [3], BFO [4], artificial honey bees (AHB) [5] and others]. In the first two sections, the key of force and the limitations of both competitive as well as collaborative will be reported and discussed. Theoretical discussions concerned convergence issues, parameters sensitivity analysis and computational burden estimation are reported as well. Successively, a brief review on how different research groups have applied or customized different nature inspired optimizations (NIOs) approaches for the solution of complex practical electromagnetic problems ranging from industrial up to biomedical applications. The chapter ends with open problems and discussion on future applications.
 + Show details


15 Interactive chaotic evolution
 + Show details

Hide details

p.
417
–443
(27)
In this chapter, we propose a new interactive evolutionary computation (IEC) algorithm, we call it interactive chaotic evolution (ICE), which fuses the optimization capability of chaotic evolution and the subjective evaluation of human. We make a brief review on the research of IEC optimization and present the philosophy and the implementation of chaotic evolution and some chaotic systems that can be involved in ICE optimization framework. For a comparison study of ICE, we introduce the interactive differential evolution algorithm (IDE) as a comparative algorithm. We investigate the optimization performance of ICE using some benchmark functions as pseudoIEC evaluators, and several statistical tests are applied. We analyse and discuss the subjects on paired comparison mechanism of ICE, the distribution characteristic of chaotic system and optimization capability of ICE, the characteristics of combining a chaotic system and a uniform random system, and fusion of ICE and IDE. In this work, we do not only pursue to analyse and discuss the algorithm optimization mechanism of ICE but also induce the philosophy and methodology behind it. We hope that many people realize that the capability of ICE is not only optimization from these efforts, and ICE can make benefit to both chaos theory and evolutionary optimization, perspectively.
 + Show details


16 Symbiotic organisms search algorithm for static and dynamic transmission expansion planning
 + Show details

Hide details

p.
445
–465
(21)
Transmission expansion planning (TEP) is a conventional problem of electric power systems. The main objective of TEP is to govern the optimum expansion plan of the electrical power networks. This chapter proposes symbiotic organisms search (SOS) algorithm (a novel metaheuristic optimization technique) for the solution of TEP problem of power systems. SOS algorithm is motivated by the interactions among the organisms in the ecosystem. Both static and dynamic TEP problem have been modeled in this chapter using DC power flow model and are effectively solved by the SOS algorithm. Several constraints such as rightofway's validity, maximum number of lines addition, power flow of the network lines have been taken into consideration. To authenticate the capability of the proposed method, Garver's 6bus system, IEEE 25bus system and Colombian 93bus system are tested for TEP problem. The efficacy of the proposed SOS algorithm, while dealing with different case studies of the studied power networks, is established in terms of higher quality results (i.e., lower investment cost), lower competitive computational burden and quicker (also stable) convergence mobility.
 + Show details


17 Inclined planes system optimisation (IPO) and its applications in data mining and system identification
 + Show details

Hide details

p.
467
–495
(29)
In this chapter, a swarm intelligence algorithm [called inclined planes system optimisation (IPO)] is described, and its applications in different aspects of data mining and system modelling are investigated. IPO is a new swarm intelligence technique that is inspired by the dynamic of sliding motions along frictionless inclined planes. In fact, a swarm of agents (balls) cooperate with each other and move towards better positions in the search space by employing Newton's second law and equations of motion on inclined planes. After introducing IPO, its applications on data clustering, decision function estimation, circuit design and image processing (which are three branches of data mining) are described. Also, its performance on infiniteimpulseresponse model identification as an instance of system identification is investigated. IPO shows comparable results to other optimisation methods in different applications. A lot of control parameters in IPO give freedom to designers but sometimes make it need a lot of try and errors for finding optimal parameters which lead to complexity. A modified version of IPO (MIPO) can solve this problem. While in MIPO, the parameters are more independent from user, it shows better, more reliable results than normal IPO. Another concern is optimising multiobjective problems, which can be done using multiobjective IPO. MIPO shows its robustness in designing LC_VCO circuits.
 + Show details


Back Matter
 + Show details

Hide details

p.
(1)
 + Show details
