Population-based Meta-heuristic Methods Research Articles

A Novel Hybrid Crow Search Arithmetic Optimization Algorithm for Solving Weighted Combined Economic Emission Dispatch with Load-Shifting Practice

The crow search arithmetic optimization algorithm (CSAOA) method is introduced in this article as a novel hybrid optimization technique. This proposed strategy is a population-based metaheuristic method inspired by crows’ food-hiding techniques and merged with a recently created simple yet robust arithmetic optimization algorithm (AOA). The proposed method’s performance and superiority over other existing methods is evaluated using six benchmark functions that are unimodal and multimodal in nature, and real-time optimization problems related to power systems, such as the weighted dynamic economic emission dispatch (DEED) problem. A load-shifting mechanism is also implemented, which reduces the system’s generation cost even further. An extensive technical study is carried out to compare the weighted DEED to the penalty factor-based DEED and arrive at a superior compromise option. The effects of CO2, SO2, and NOx are studied independently to determine their impact on system emissions. In addition, the weights are modified from 0.1 to 0.9, and the effects on generating cost and emission are investigated. Nonparametric statistical analysis asserts that the proposed CSAOA is superior and robust.

An enhanced meta-heuristic algorithm used for energy conscious priority-based task scheduling problems in heterogeneous multiprocessor systems

The Task Scheduling Problem (TSP) in Multiprocessor Systems (MPS) is an emerging research area in heterogeneous distributed computing environments. Managing complex tasks and achieving optimal efficiency while scheduling tasks in MPS presents challenges. Although adding more processors to a single system greatly increases its processing power, the energy these processors produce is the main drawback. Here, we employed a multi-objective optimization strategy to reduce the makespan and Energy Consumption (EC). To reduce processor EC, we considered an effective Dynamic Voltage and Frequency Scaling (DVFS) technique. Three distinct energy sources are considered, originating from the processors' communication, idle, and active states. The scheduling sequence of tasks and the assignment of tasks to processors are two vital aspects of TSP. Here, the tasks are arranged based on priority using the Upward Rank technique. To minimize the makespan and EC while allocating tasks to processors, we use the population-based metaheuristic method called Enhanced Honey Badger Optimisation (EHBO). We proposed three improvements to the HBO algorithm in EHBO. Initially, we addressed opposite learning-based population initialization to exclude the least suitable candidates and generate a candidate scheduling population that adheres to task precedence. Subsequently, the levy-flight technique is employed to improve the local and global search and preserve their ideal healthy balance. Finally, using dynamic values rather than constant value improves the ability to obtain maximum food. Several experiments are conducted on random task graphs and real-life data sets. Additionally, the results are compared with other upgraded meta-heuristic algorithms, demonstrating the superiority of EHBO.

Soft computing methods in the solution of an inverse heat transfer problem with phase change: A comparative study

Inverse heat transfer problems are ill-posed problems and their solution is challenging. Conventional (hard computing) solution methods were developed for this purpose in the past, but they are not well applicable in cases including phase change, which contain strong non-linearity and bring additional computational difficulties. Soft computing methods, which currently experience very rapid development, are a promising tool for the solution of such problems. This paper addresses an inverse heat transfer problem with phase change, in which the boundary heat flux is estimated. Four methods based on distinct mathematical principles are applied to this problem and thoroughly compared. These methods include a conventional Levenberg–Marquardt method (LMM), a predictive fuzzy logic (PFL)-based method, a population-based meta-heuristic method called LSHADE (a state-of-the-art differential evolution variant), and a recently developed surrogate-assisted method coupled with differential evolution, referred to as LSADE method. Furthermore, a reformulation of the problem was developed, utilising a dimension reduction scheme and a decomposition scheme that led to sub-problems with different time frames. This reformulation brought extensive computational improvements. Results of the comparison of the methods then showed that the LMM and the PFL behave well in case without phase change but their performance deteriorates substantially in case with phase change. The LSHADE and the LSADE showed superior performance in the solution of the inverse problem with the phase change. Moreover, their performance was rather stable and insensitive to the location of the temperature sensor, which was the source of data for the estimation.

Modified grey wolf optimizer based MPPT design and experimentally performance evaluations for wind energy systems

The method that aims to operate the wind energy system (WES) at the maximum power point (MPP) is called the maximum power point tracking (MPPT) method in the literature. The grey wolf optimization (GWO) is one of the newest population-based meta-heuristic methods, and its performance as an MPPT algorithm in WESs has not been extensively studied yet. In this study, the standard GWO algorithm has been modified considering the requirements of WES, so that the system can reach the MPP quickly and stably, thereby improving the system’s efficiency. Moreover, the performance of the proposed method is examined comparatively with the well-known MPPT methods via simulation and experimental studies for many possible scenarios. It is demonstrated that the proposed modified GWO (MGWO) performance is better than the classic and modified perturb and observe methods. The results have also been compared with the Fibonacci Search (FS) and Golden Section (GS) Search-based MPPT algorithms newly presented in the literature for WES. Although the results of FS, GS, and MGWO-based MPPT algorithms are very close to each other, it has been observed that FS has a slightly better performance.

A metaheuristic approach for multi-objective optimization of the Stirling cycle with internal irreversibilities and regenerative losses using artificial bee colony algorithm

This study aimed to investigate the multi-objective optimization of a Stirling cycle with a mathematical model that takes into account the dead volumes, effects of regenerative losses, and internal irreversibilities on the thermodynamic performance of the cycle. The optimization was carried out using the Artificial Bee Colony algorithm, which is a population-based metaheuristic method that mimics the foraging behavior of honeybees. The developed algorithm employs the concepts of Pareto frontier and ε-dominance to find the optimal solutions in the multi-objective space which is obtained with Artificial Bee Colony algorithm. The input parameters of the system are the maximum and minimum temperatures, compression and expansion volumes, and charge pressure which affect the thermodynamic variables of the cycle. The optimization process consisted of two stages: firstly, single-objective optimization was performed separately for each objective function to obtain the baseline results. Then, three different sets of triple-objective function groups were used to perform multi-objective optimization. These are Case 1 (net work output, thermal efficiency, irreversibility parameter), Case 2 (net work output, thermal efficiency, 2nd law efficiency), and Case 3 (net work output, thermal efficiency, entropy generation). The obtained results from the single and multi-objective optimizations were compared and analyzed. Since multi-objective optimization involves conflicting objectives, it does not result in a single optimal solution, but rather a set of optimal solutions that represent different trade-offs among the objectives. In order to achieve the optimum results with a good trade-off between solutions Pareto frontier method is used. In addition, to obtain a good distribution of solutions and filter very similar solution points in the solution space, ε-dominance was used to filter them. Finally, to select the final optimal solution from the Pareto frontier solution set, LINMAP was used as a decision-making tool which is a linear programming technique that assigns weights to each solution. The weights of the solutions are achieved according to the relative distance between objectives with their single optimal values achieved by single optimization. The best solution based on the net work output is achieved with Case 3 which includes entropy generation as different from other cases. Also, Case 3 has the lowest irreversibility parameter value even though the irreversibility parameter was optimized in Case 1. Entropy generation and mean effective pressure are quite sensitive due to the multiple solutions they have at each solution step when entropy generation is not an objective function.

A comparison performance analysis of eight meta-heuristic algorithms for optimal design of truss structures with static constraints

Metaheuristics have been successfully used for solving complex structural optimization problems. Many algorithms are proposed for truss structure size and shape optimization under some constraints. This study considers eight population-based meta-heuristic methods: African Vultures Optimization Algorithm (AVOA), Flow Direction Algorithm (FDA), Arithmetic Optimization Algorithm (AOA), Generalized Normal Distribution Optimization (GNDO), Stochastic Paint Optimizer (SPO), Chaos Game Optimizer (CGO), Crystal Structure Algorithm (CRY) and Material Generation Algorithm (MGO). These meta-heuristics methods are used to optimize the size of three aluminum truss structures. Optimization aims to reduce the weight of the truss members while meeting a set of displacement and stress constraints. The performance of these methods is assessed by solving and optimizing three well-known truss structure benchmarks under some constraints. The results show that the Stochastic Paint Optimizer (SPO) outperforms the other algorithms in terms of accuracy and convergence rate.

Levy Enhanced Cross Entropy-based Optimized Training of Feedforward Neural Networks

An Artificial Neural Network (ANN) is one of the most powerful tools to predict the behavior of a system with unforeseen data. The feedforward neural network is the simplest, yet most efficient topology that is widely used in computer industries. Training of feedforward ANNs is an integral part of an ANN-based system. Typically an ANN system has inherent non-linearity with multiple parameters like weights and biases that must be optimized simultaneously. To solve such a complex optimization problem, this paper proposes the Levy Enhanced Cross Entropy (LE-CE) method. It is a population-based meta-heuristic method. In each iteration, this method produces a "distribution" of prospective solutions and updates it by updating the parameters of the distribution to obtain the optimal solutions, unlike traditional meta-heuristic methods. As a result, it reduces the chances of getting trapped into local minima, which is the typical drawback of any AI method. To further improve the global exploration capability of the CE method, it is subjected to the Levy flight which consists of a large step length during intermediate iterations. The performance of the LE-CE method is compared with state-of-the-art optimization methods. The result shows the superiority of LE-CE. The statistical ANOVA test confirms that the proposed LE-CE is statistically superior to other algorithms.

The novel combination lock algorithm for improving the performance of metaheuristic optimizers

Nature-inspired population-based metaheuristics are promising search methods for solving optimization problems. In this paper, a novel systematic pre-processing approach, called the combination lock algorithm, for obtaining a good starting point for population-based algorithms is proposed. The proposed algorithm is tested on 32 benchmark unconstrained multidimensional optimization problems of different characteristics that are either unimodal or multimodal, continuous or non-continuous, separable or non-separable, differentiable or non-differentiable. The tests also include four engineering constrained optimization benchmark problems. The experimental results of applying the proposed algorithm for the Particle Swarm Optimization, the Ant Colony Optimization for Continuous Domain, and the Grey Wolf Optimization were compared with the results obtained from the conventional approach of initializing the starting population of population-based metaheuristic methods. The simulation results show the potential of the proposed algorithm as an efficient and reliable approach to enhance the performance of population-based optimization algorithms such that, overall, 50% and up to 100% of the tested problems “across various population size” settings, had either improved or equalled the optimal values when the proposed algorithm was applied.

Analiza i optimizacija glavnog nosača mosne dizalice sa asimetričnim kutijastim poprečnim presekom

The presented research deals with the optimal design of the main girder with an asymmetric box cross-section of the double-beam bridge crane. The Moth-Flame Optimization algorithm (MFO) is used for solving this multicriteria optimization problem. This algorithm is a relatively new population-based metaheuristic method. The paper takes the following criteria as the constraint functions: strength, local stability of the girder plates (webs and top flange), local stability of the longitudinal stiffeners, global stability of the main girder, deflections, and period of oscillation. The justification of the proposed procedure is shown in one example of a real solution of the double-beam bridge crane. Significant savings in material were achieved in this research, within the range of 19.42 to 25.49%. The use of this algorithm enables the application of a very large number of variables and constraint functions, whereby the optimal values are obtained in a relatively short period.

Population-based Metaheuristics for the Dynamic Minimum Cost Hybrid Berth Allocation Problem

This study considers the Dynamic Minimum Cost Hybrid Berth Allocation Problem (DMCHBAP) with fixed handling times of vessels. The objective function to be minimized consists of three components: costs of positioning, waiting, and tardiness of completion for all vessels. A mathematical formulation of DMCHBAP, based on Mixed Integer Linear Programming (MILP), is proposed and used within the framework of commercial CPLEX 12.3 solver. As the speed of finding high-quality solutions is of crucial importance for an efficient and reliable decision support system in container terminal, two population-based metaheuristic approaches to DMCHBAP are proposed: combined Genetic Algorithm (cGA) and improvement-based Bee Colony Optimization (BCOi). Both cGA and BCOi are evaluated and compared against each other and against state-of-the-art solution methods for DMCHBAP on five sets of problem instances. The conducted computational experiments and statistical analysis indicate that population-based metaheuristic methods represent promising approaches for DMCHBAP and similar problems in maritime transportation.

A two-individual based path-relinking algorithm for the satellite broadcast scheduling problem

Population-based metaheuristic algorithms normally manage a large number of ‘individuals’ to achieve diversity in the search process, which in turn lead to good quality solutions. Although the quality of the solutions produced by these algorithms is observed to be good for a variety of combinatorial optimization problems, managing a large number of individuals within the algorithm is often complex and time-consuming. In this paper, we develop a two-individual based path relinking (TPR) algorithmic framework harnessing the power of the solution-based tabu search and that of a distance-controlled relinking operator to solve the satellite broadcast scheduling problem (SBSP). The results of extensive computational experiments carried out demonstrate that our TPR algorithm outperforms state-of-the-art heuristic algorithms for SBSP with respect to various performance metrics, in a statistically significant way. The aim of this study is to compare the two-individual based search algorithm with the traditional population-based metaheuristic methods in the literature for the SBSP (i.e., ant colony optimization and differential evolution variant algorithms) to provide a highly competitive algorithm for solving this important practical problem with numerous applications, and to promote research on two-individual based search method, which received very little attention in the literature. Our results underscore that the two-individual based search algorithmic framework is a viable alternative for solving complex optimization problems and reconfirm the validity of analogous observations made by other researchers in the context of graph coloring and flexible job shop scheduling.

A parallel multi-objective swarm intelligence framework for Big Data analysis

Nowadays, data are generated from smart devices in huge volumes, different formats, and high pace, which comply with Big Data characteristics. Big Data led to the emergence of new technologies, such as Hadoop and Spark to provide both data management and analysis. Analysing Big Data is a time-consuming process. Particle swarm and ant colony optimisation are population-based meta-heuristic methods. They have been combined with data mining techniques to solve MultiObjective Problems (MOPs) of small and medium sized data, presenting good performance. However, when applying these methods to solve MOPs in Big data, an efficient scalable framework will be required. In this paper, we summarise new technologies proposed to manage and analyse Big Data. We present how meta-heuristics can be adapted with Big Data technologies. We characterise problems arose when analysing MO Big Data problems, in addition to proposed methods to overcome these problems, giving examples in Bioinformatics field.

A parallel multi-objective swarm intelligence framework for Big Data analysis

Nowadays, data are generated from smart devices in huge volumes, different formats, and high pace, which comply with Big Data characteristics. Big Data led to the emergence of new technologies, such as Hadoop and Spark to provide both data management and analysis. Analysing Big Data is a time-consuming process. Particle swarm and ant colony optimisation are population-based meta-heuristic methods. They have been combined with data mining techniques to solve MultiObjective Problems (MOPs) of small and medium sized data, presenting good performance. However, when applying these methods to solve MOPs in Big data, an efficient scalable framework will be required. In this paper, we summarise new technologies proposed to manage and analyse Big Data. We present how meta-heuristics can be adapted with Big Data technologies. We characterise problems arose when analysing MO Big Data problems, in addition to proposed methods to overcome these problems, giving examples in Bioinformatics field.

Energy-Based Acoustic Localization by Improved Elephant Herding Optimization

The present work proposes a new approach to address the energy-based acoustic localization problem. The proposed approach represents an improved version of evolutionary optimization based on Elephant Herding Optimization (EHO), where two major contributions are introduced. Firstly, instead of random initialization of elephant population, we exploit particularities of the problem at hand to develop an intelligent initialization scheme. More precisely, distance estimates obtained at each reference point are used to determine the regions in which a source is most likely to be located. Secondly, rather than letting elephants to simply wander around in their search for an update of the source location, we base their motion on a local search scheme which is found on a discrete gradient method. Such a methodology significantly accelerates the convergence of the proposed algorithm, and comes at a very low computational cost, since discretization allows us to avoid the actual gradient computations. Our simulation results show that, in terms of localization accuracy, the proposed approach significantly outperforms the standard EHO one for low noise settings and matches the performance of an existing enhanced version of EHO (EEHO). Nonetheless, the proposed scheme achieves this accuracy with significantly less number of function evaluations, which translates to greatly accelerated convergence in comparison with EHO and EEHO. Finally, it is also worth mentioning that the proposed methodology can be extended to any population-based metaheuristic method (it is not only restricted to EHO), which tackles the localization problem indirectly through distance measurements.

Investigations on Metaheuristic Algorithm for Designing Adaptive PID Controller for Continuous Stirred Tank Reactor

An integral part of industrial processes is the Continuous Stirred Tank Reactor (CSTR) whose dynamic characteristics are highly nonlinear causing the reactor to deviate from its set temperature point. For its efficient operation, specific parameters of the CSTR are required to be controlled. Hence, this paper presents designing of the Proportional-Integral-Derivative (PID) controller using conventional and metaheuristic methods for the temperature control of the CSTR. The conventional controller is tuned with Ziegler Nichols (Z-N) method. Global search population-based metaheuristic methods like Artificial Bee Colony (ABC), Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) have been implemented to optimize the adaptive controller parameters, and a comparative analysis is done taking the step response into consideration. The CSTR system is simulated by the proposed controller which improves the robustness, behaviour and tracking of the system. The simulation results present substantial enhancement of the time response parameters, i.e. settling time, rise time, peak overshoot, mean square error and integral time absolute error. Studies showed that the proposed adaptive methods with metaheuristic algorithms are fast and efficient in error reduction. Here, ABC irrespective of other optimization methods suitably optimized the controller parameters.

A REVIEW OF SINGLE AND POPULATION-BASED METAHEURISTIC ALGORITHMS SOLVING MULTI DEPOT VEHICLE ROUTING PROBLEM

Multi-Depot Vehicle Routing Problem (MDVRP) arises with rapid development in the logistics and transportation field in recent years. This field, mainly, faces challenges in arranging their fleet efficiently to distribute the goods to customers by minimizing distance and cost. Therefore, the decision maker needs to specify the vehicles to reach the particular depot which, serves the customers with the predetermined capacity. Hence, to solve the stated problems, there is a need to apply metaheuristic methods to get minimal transportation costs. This article reviews on single and population-based metaheuristic methods solving MDVRP from the year 2013 until 2018. The methods discussed were simulated annealing (SA), variable neighborhood search (VNS), ant colony algorithm (ACO), particle swarm optimization (PSO) and genetic algorithm (GA). From the previous works, it can be concluded that the application of population based metaheuristic gives better solutions in solving MDVRPs. Keywords: Metaheuristic, Multi Depot, MDVRP

Associating optical measurements of MEO and GEO objects using Population-Based Meta-Heuristic methods

Currently several thousands of objects are being tracked in the MEO and GEO regions through optical means. The problem faced in this framework is that of Multiple Target Tracking (MTT). The MTT problem quickly becomes an NP-hard combinatorial optimization problem. This means that the effort required to solve the MTT problem increases exponentially with the number of tracked objects. In an attempt to find an approximate solution of sufficient quality, several Population-Based Meta-Heuristic (PBMH) algorithms are implemented and tested on simulated optical measurements. These first results show that one of the tested algorithms, namely the Elitist Genetic Algorithm (EGA), consistently displays the desired behavior of finding good approximate solutions before reaching the optimum. The results further suggest that the algorithm possesses a polynomial time complexity, as the computation times are consistent with a polynomial model. With the advent of improved sensors and a heightened interest in the problem of space debris, it is expected that the number of tracked objects will grow by an order of magnitude in the near future. This research aims to provide a method that can treat the association and orbit determination problems simultaneously, and is able to efficiently process large data sets with minimal manual intervention.

A Parthenogenetic Algorithm for the Founder Sequence Reconstruction Problem

The maximum fragment length (MFL) is an important computational model for solving the founder sequence reconstruction problem. Benedettini et al. presented a meta-heuristic algorithm BACKFORTH based on iterative greedy method. The BACKFORTH algorithm starts with a single initial solution, and iteratively alternates between a partial destruction and reconstruction in order to obtain a final solution. The kind of optimization mechanism, which is based on a single initial solution, may make the performance of the BACKFORTH algorithm sensitive to the quality of the initialization. In this paper, a practical parthenogenetic algorithm PGMFL, which is a population-based meta-heuristic method, is proposed. The PGMFL algorithm can search multiple regions of a solution space simultaneously. A novel genetic operator is introduced based on the presented heuristic algorithm HF, which takes advantage of look-ahead mechanism and some potential information, i.e., the proportions of 0 and 1 entries in a column of recombinant matrix and those in the corresponding column of the founder matrix, and some other heuristic information, to compute the column values. The PGMFL algorithm can get fewer breakpoints and longer fragment average length than the BACKFORTH algorithm, which are proved by a number of experiments.

Dynamic task scheduling with load balancing using parallel orthogonal particle swarm optimisation

This paper presents a novel approach for dynamic task scheduling using particle swarm optimisation. Particle swarm optimisation (PSO) is a population-based meta-heuristic method which can be used to solve np-hard problems. The algorithm has been developed to dynamically schedule heterogeneous tasks on to heterogeneous processors in a distributed setup. Load balancing which is a major issue in task scheduling is also considered. The nature of the tasks are independent and non pre-emptive. Different approaches using PSO has been tried namely PSO with fixed inertia, PSO with variable inertia, PSO with elitism, MPSO, parallel PSO, hybrid PSO, orthogonal PSO and parallel orthogonal PSO. The performance of PSO and its variants is also compared with the genetic algorithm concept. The objective of the algorithms is to minimise the make-span of the entire schedule. Benchmark problems have been taken and validated. The result depicts that the dynamic task scheduling implemented using parallel orthogonal particle swarm optimisation technique is cost-effective in nature when compared to the other algorithms tested.