Search

This paper presents a Chaos–Infused Moth–Flame Optimizer (CI-MFO). The parent algorithm is modified to account for deviations in search agent (moth) flight trajectory and variations in the flame orientation for an enhanced balance between exploration and exploitation tendencies. Actual photographic evidence showing light traces of such phototactic moths in-flight has been used to model flight path deviations using Chaos Theory. This approach considers their intelligence and erratic flight behavior (when subjected to excessive crowding). The performance of the developed CI-MFO algorithm is investigated comprehensively using a suite of fifty-eight benchmarking functions, including seven unimodal, six multimodal, ten fixed-dimension multimodal, six CEC-2005 hybrid-composite, and twenty-nine CEC-2017 hybrid-composite functions. The proposed algorithm's effectiveness is tested against several classical algorithms and some modified metaheuristic optimization algorithms in terms of obtained mean optima and standard deviations, and scalability analysis is also performed. The paper concludes by solving several real-world problems and comparing the proposed algorithm's performance against several reported algorithms. The proposed algorithm exhibited a substantially better solution-finding ability. [ABSTRACT FROM AUTHOR]

A novel nature-inspired metaheuristic optimization algorithm, called the quantum firefly algorithm, is proposed in this paper. The algorithm imitates (a) the social behaviour of fireflies mating in nature, (b) laws of quantum physics, and (c) laws of natural evolution. The algorithm combines the powers of two well-known algorithms: the firefly algorithm and the quantum genetic algorithm. The proposed quantum firefly algorithm's performance is tested on 15 mathematical test functions and one structural design problem. The obtained results show that the quantum firefly algorithm is very competitive compared to the firefly algorithm and the quantum genetic algorithm. [ABSTRACT FROM AUTHOR]

Feature transformation methods can be used effectively to improve the performance of the machine learning techniques, such as classification and clustering, via enhancing the discrimination among the samples. Additionally, binary representation of features may be necessary in several circumstances. For example, distributed systems may have limited bandwidth, storage, and energy that necessitate rough quantization of the measurements. This paper proposes a random projection-based feature transformation method that maps the data points from the original space into a new binary space. In the proposed transformation method, the random projection process is formulated as an optimization problem. A modified chaotic version of the sine–cosine algorithm is used to find the optimal random projector. The performance of the proposed transformation method is validated using 15 standard UCI benchmark datasets against the linear discernment analysis. Moreover, it is compared to other metaheuristic-based feature reduction methods including the gray wolf optimizer, the genetic algorithm, and the antlion optimizer. The obtained results assure the superiority of the proposed transformation method in terms of classification accuracy and storage requirements. [ABSTRACT FROM AUTHOR]

This paper proposes a hybrid optimization algorithm based on the combination of the merits of the backtracking search (BSA) and sine-cosine algorithm (SCA) to achieve the optimal design of a shell and tube evaporator. To the author's best knowledge, this is the first application of the metaheuristic algorithms over shell and tube evaporator design problems. In order to test the accuracy of the proposed hybrid algorithm, 10 well-known optimization test functions have been solved. Numerical results obtained from the hybrid BSA-SCA have been compared with the literature optimizers including differential search, big bang-big crunch optimization, quantum-behaved particle swarm optimization, bat algorithm, intelligent tuned harmony search algorithm, and backtracking search algorithm. Comparison results reveal that solutions obtained from the BSA-SCA are better than those of the results acquired by the aforementioned optimizers with respect to statistical analysis. Proposed optimization procedure is then utilized to obtain optimum values of the two heat exchanger design objectives including total cost and overall heat transfer coefficient. Six decision variables such as tube outer diameter, shell diameter, baffle spacing, tube length, number of tube passes, and tube bundle configuration are selected to be iteratively optimized. It is found that BSA-SCA provides better results than the compared literature optimizers for both objective functions. In addition, a sensitivity analysis is performed for the design parameters at the optimal point. Results show that variation of the design parameters at the optimum point has considerable effect on the objective function rates. [ABSTRACT FROM AUTHOR]

This study proposes a brand new practical multi-agent optimization framework based on an intelligent collaborative interaction between some prevalent metaheuristic algorithms available in the literature. Proposed optimization architecture is built on widely known and reputed master-slave model assisted with some useful and promising modifications. Conventional stochastic-based optimization algorithms including Particle Swarm Optimization, Crow Search Algorithm, Differential Evolution, and Global Best Algorithm are structurally coordinated to form the slave populations, while the best solutions obtained from these slave subpopulations are forming the master individuals. Contrary to the traditional master-slave approach, master individuals in this proposed framework becomes more functional by performing an extensive local search over numerical results of the best slave individuals. Main aim in constructing such a highly devised multi-agent algorithm is to maintain an effective communication domain between slave individuals (agents) as well as to enhance the capabilities of the cooperative search mechanism through the systematic combination of metaheuristics. Optimization performance of the proposed framework is tested on a suite of 29 optimization benchmark functions. Proposed optimization method surpasses the compared optimization algorithms in 27 out of 29 problems and proves its solution efficiency in multidimensional optimization problems. Then, proposed strategy is applied on single and multi-objective thermal design of a shell and tube evaporator operated with R134a/Al2O3 nanorefrigerant. It is seen that maximum overall heat transfer coefficient is increased by 18.1% and minimum total cost of heat exchanger is reduced by 5.1% in the case of incorporating Al2O3 nanoparticles into R134a. [ABSTRACT FROM AUTHOR]