Parallel Genetic Algorithm Research Articles

Advances in the enumeration of foldable self-avoiding walks

Self-avoiding walks (SAWs) have been studied for a long time due to their intrinsic importance and the many application fields in which they operate. A new subset of SAWs, called foldable SAWs, has recently been discovered when investigating two different SAW manipulations embedded within existing protein structure prediction (PSP) software. Since then, several attempts have been made to find out more about these walks, including counting them. However, calculating the number of foldable SAWs appeared as a tough work, and current supercomputers fail to count foldable SAWs of length exceeding ≈ 30 steps. In this article, we present new progress in this enumeration, both theoretical (mathematics) and practical (computer science). A lower bound for the number of foldable SAWs is firstly proposed, by studying a special subset called prudent SAWs that is better known. The triangular and hexagonal lattices are then investigated for the first time, leading to new results about the enumeration of foldable SAWs on such lattices. Finally, a parallel genetic algorithm has been designed to discover new non-foldable SAWs of lengths ≈ 100 steps, and the results obtained with this algorithm are promising.

A real-time Decision Support System for Big Data Analytic: A case of Dynamic Vehicle Routing Problems

Recently, the explosion of large amounts of traffic data has guided data scientists to create models with big data for a better decision-making. Big Data applications process and analyze this huge amounts of data (collected from a variety of heterogeneous data sources) that cannot be processed with traditional technologies. In this paper, Big Data frameworks are used for solving an optimization problem known as Dynamic Vehicle Routing Problem (DVRP). Hence, due to the NP-Hardness of the problem and to deal with a large size of data, we develop a parallel Spark Genetic Algorithm named (S-GA). This parallelism aims to take the advantage of Spark’s in-memory computing ability (as a master-slave distribution computing) and GA’s iterations operations. Parallel operations were used for fitness evaluation and genetic operations. Based on the parallel S-GA a decision support system is developed for the DVRP in order to generate the best routes. The experiments show that our proposed architecture is improved due to its capacity when coping with Big Data optimization problems by interconnecting components and deploying on different nodes of a cluster.

A Parallel Genetic Algorithm Framework for Transportation Planning and Logistics Management

Small to medium sized transportation and logistics companies are usually constrained by limited computing and IT professional resources on implementing an efficient parallel metaheuristic algorithm for planning or management solutions. In this paper we extend the standard meta-description for genetic algorithms (GA) with a simple non-trivial parallel implementation. Our parallel GA framework is chiefly concerned with the development of a straightforward way for engineers to modify existing genetic algorithm implementations for real transportation and logistics problems to make use of commonly available hardware resources without completely reworking complex, useful and usable codes. The framework presented at its parallel base is a modification of the primitive parallelization concept, but if implemented as described it may be gradually extended to fit the qualities of any underlying problem better (via the adaptation of the merging and communications functions).We present our framework and computational results for a classical transportation related combinatorial optimization problem - the traveling salesman problem with a standard sequential genetic algorithm implementation. Our empirical analysis shows that this simple extension can lead to considerable solution improvements. We also tested our assumptions that the framework is easily implemented by an engineer not initially familiar with genetic algorithms to implement the framework for another minimum multiprocessor scheduling problem. These case studies verify that our framework is better than primitive parallelization because it gives empirically better results under equitable conditions. It also outperforms fine grained parallelization as it is easier and faster to implement.

A Hybrid Genetic Algorithm Based on Information Entropy and Game Theory

To overcome the disadvantages of traditional genetic algorithms, which easily fall to local optima, this paper proposes a hybrid genetic algorithm based on information entropy and game theory. First, a calculation of the species diversity of the initial population is conducted according to the information entropy by combining parallel genetic algorithms, including using the standard genetic algorithm (SGA), partial genetic algorithm (PGA) and syncretic hybrid genetic algorithm based on both SGA and PGA for evolutionary operations. Furthermore, with parallel nodes, complete-information game operations are implemented to achieve an optimum for the entire population based on the values of both the information entropy and the fitness of each subgroup population. Additionally, the Rosenbrock, Rastrigin and Schaffer functions are introduced to analyse the performance of different algorithms. The results show that compared with traditional genetic algorithms, the proposed algorithm performs better, with higher optimization ability, solution accuracy, and stability and a superior convergence rate.

GA-Par: Dependable Microservice Orchestration Framework for Geo-Distributed Clouds

Recent advances in composing Cloud applications have been driven by deployments of inter-networking heterogeneous microservices across multiple Cloud datacenters. System dependability has been of the upmost importance and criticality to both service vendors and customers. Security, a measurable attribute, is increasingly regarded as the representative example of dependability. Literally, with the increment of microservice types and dynamicity, applications are exposed to aggravated internal security threats and externally environmental uncertainties. Existing work mainly focuses on the QoS-aware composition of native VM-based Cloud application components, while ignoring uncertainties and security risks among interactive and interdependent container-based microservices. Still, orchestrating a set of microservices across datacenters under those constraints remains computationally intractable. This paper describes a new dependable microservice orchestration framework GA-Par to effectively select and deploy microservices whilst reducing the discrepancy between user security requirements and actual service provision. We adopt a hybrid (both whitebox and blackbox based) approach to measure the satisfaction of security requirement and the environmental impact of network QoS on system dependability. Due to the exponential grow of solution space, we develop a parallel Genetic Algorithm framework based on Spark to accelerate the operations for calculating the optimal or near-optimal solution. Large-scale real world datasets are utilized to validate models and orchestration approach. Experiments show that our solution outperforms the greedy-based security aware method with 42.34 percent improvement. GA-Par is roughly 4x faster than a Hadoop-based genetic algorithm solver and the effectiveness can be constantly guaranteed under different application scales.

Minimizing makespan in a Flow Shop Sequence Dependent Group Scheduling problem with blocking constraint

Flow Shop Sequence Dependent Group Scheduling (FSDGS) problems gathered much attention from the body of literature in recent years. Nevertheless, the combination of blocking constraint and Group Technology (GT) principles has not been faced by academics so far. The aim of the present paper is to propose an original meta-heuristic approach for minimizing makespan in a FSDGS problem with blocking constraint. To this end, a novel Parallel Self-Adaptive Genetic Algorithm (PSAGA) which adaptively varies the genetic parameters along the evolutionary mechanism was devised. Validation of the proposed metaheuristics was performed by means of the global optima generated by a proper mixed integer linear programming model. An extended experimental campaign, also supported by a specific statistical analysis, demonstrates the effectiveness of the proposed approach compared to other meta-heuristics arising from the relevant literature.

Analysis of Parallel Genetic Algorithm and Parallel Particle Swarm Optimization Algorithm UAV Path Planning on Controller Area Network

Path planning of the unmanned aerial vehicle (UAV) is one of the complex optimization problems due to the model complexity and a high number of constraints. Evolutionary algorithms are a conventional method to solve complex optimization problems with multiple constraints. However, they are time-consuming algorithms which make their implementation difficult. UAV path planning is usually a real-time problem that is needed to solve in limited time steps. Parallelization is one of the most effective methods for reducing the computation time of evolutionary algorithms. In the industrial applications, usually there is a communication platform between devices which can be used for the implementation of the parallelization process. For example, controller area network (CAN) is a protocol used for reliable communication in aerospace systems. Due to the development of this protocol on UAVs and its characteristics, it can be used for parallel optimization of path planning problem. In this paper, an asynchronous distributed multi-master parallel genetic algorithm and a parallel particle swarm optimization algorithm are implemented on CAN bus to improve the speed and performance of the UAV path planning.

Triclustering of gene expression microarray data using coarse grained and dynamic deme based parallel genetic approach

Interpretation of microarray data is often crucial in various aspects of computational analysis. Dozens of techniques like clustering, subspace-clustering are often applied to these datasets to pluck-out significant results that provide solutions to drug discovery, disease identification like practical healthcare problems. Clustering techniques are used to group the genes exhibiting similar behavior under particular conditions. This classical method fails while the grouping of genes is done according to a subset of conditions as it performs globally. Biclustering addresses successfully this issue having constraints for evaluation of gene grouping only under a subset of the conditions. However, one of the limitations of the biclustering technique is, it is not capable to analyze the longitudinal experiments which consider different time points for the analysis of the gene expression profiles under a subset of conditions. This affair motivates to adopt triclustering on gene expression microarray data. Triclustering usually finds a set of genes of similar behaviors under a subset of conditions under certain time points. In this research article, two new frameworks based on different versions of parallel genetic algorithms are proposed to detect significant triclusters in gene expression profiles. In the first framework, the proposed algorithm is based on Coarse Grained parallel genetic approach and in the second framework the proposed algorithm is based on the Dynamic Deme parallel genetic approach. Both of them consider the experimental conditions and along with the time points with an advantage of paralleling the process and reducing the computational time. The proposed frameworks are tested on a standard yeast cell cycle(Saccharomyces cerevisiae) dataset and on its different synthetic versions which are widely used in the gene expression analysis. The performance analysis is done with respect to the aspects like the convergence speed of both the algorithms for different input sizes and with respect to the computation time. The statistical analysis is performed followed by the biological relevance of the simulation results is established with their functional annotations derived from the Gene Ontology and KEGG pathway analysis graph. The performance of the proposed frameworks demonstrate its effectiveness with the other state-of-the-art schemes. Experimental results reveal that the proposed architectures are efficient as they consume less computational time due to their inherent parallel behavior. Finally, the suggested architectures are considered as reliable frameworks and can be preferable over the traditional genetic approaches to analyze the gene expression microarray data from the triclustering prospective.

MapReduce-Based Parallel Genetic Algorithm for CpG-Site Selection in Age Prediction.

Genomic biomarkers such as DNA methylation (DNAm) are employed for age prediction. In recent years, several studies have suggested the association between changes in DNAm and its effect on human age. The high dimensional nature of this type of data significantly increases the execution time of modeling algorithms. To mitigate this problem, we propose a two-stage parallel algorithm for selection of age related CpG-sites. The algorithm first attempts to cluster the data into similar age ranges. In the next stage, a parallel genetic algorithm (PGA), based on the MapReduce paradigm (MR-based PGA), is used for selecting age-related features of each individual age range. In the proposed method, the execution of the algorithm for each age range (data parallel), the evaluation of chromosomes (task parallel) and the calculation of the fitness function (data parallel) are performed using a novel parallel framework. In this paper, we consider 16 different healthy DNAm datasets that are related to the human blood tissue and that contain the relevant age information. These datasets are combined into a single unioned set, which is in turn randomly divided into two sets of train and test data with a ratio of 7:3, respectively. We build a Gradient Boosting Regressor (GBR) model on the selected CpG-sites from the train set. To evaluate the model accuracy, we compared our results with state-of-the-art approaches that used these datasets, and observed that our method performs better on the unseen test dataset with a Mean Absolute Deviation (MAD) of 3.62 years, and a correlation (R2) of 95.96% between age and DNAm. In the train data, the MAD and R2 are 1.27 years and 99.27%, respectively. Finally, we evaluate our method in terms of the effect of parallelization in computation time. The algorithm without parallelization requires 4123 min to complete, whereas the parallelized execution on 3 computing machines having 32 processing cores each, only takes a total of 58 min. This shows that our proposed algorithm is both efficient and scalable.

Polygonal approximation based on coarse-grained parallel genetic algorithm

This paper proposes to apply coarse-grained parallel genetic algorithm (CGPGA) to solve polygonal approximation problem. Chromosomes are used to represent digital curves and genes correspond to points of curves. This method divides the whole population into several subpopulations, each of which performs evolutionary process independently. After every migration interval number of generations, these subpopulations exchange their information with each other. Inspired by the designing theory of ensemble learning in machine learning, this paper further improves the basic CGPGA through adopting different but effective genetic algorithms, respectively, in different subpopulations. Both the diversity among different subpopulations and the accuracy in each individual subpopulation are ensured. Experimental results, based on four benchmark curves and four real image curves extracted from the lake maps, show that the basic CGPGA outperforms the used genetic algorithm, and further the improved CGPGA (ICGPGA) is more effective than the basic CGPGA, in terms of the quality of best solutions, the average solutions, and the variance of best solutions. Especially for those larger approximation problems, the ICGPGA is more remarkably superior to some representative genetic algorithms.

Multi robot distance based formation using Parallel Genetic Algorithm

In this paper an alternative method to achieve distance based formation is presented. The method uses Genetic Algorithms to find a suitable solution based on angle and distance, and an appropriate constant velocity to avoid collisions. The designed algorithm is extended to a parallel scheme to improve its performance and achieve Artificial Distributed Intelligence, in which the robots share, through solution migration, the best ways to converge to desired distances while avoiding collisions, finally reaching consensus on the solution. The algorithm is tested using simulations and real robots experiments.

Inversion of marine controlled source electromagnetic data using a parallel non-dominated sorting genetic algorithm

SUMMARY We developed a multi-objective optimization method for inverting marine controlled source electromagnetic data using a fast-non-dominated sorting genetic algorithm. Deterministic methods for inverting electromagnetic data rely on selecting weighting parameters to balance the data misfit with the model roughness and result in a single solution which do not provide means to assess the non-uniqueness associated with the inversion. Here, we propose a robust stochastic global search method that considers the objective as a two-component vector and simultaneously minimizes both components: data misfit and model roughness. By providing an estimate of the entire set of the Pareto-optimal solutions, the method allows a better assessment of non-uniqueness than deterministic methods. Since the computational expense of the method increases as the number of objectives and model parameters increase, we parallelized our algorithm to speed up the forward modelling calculations. Applying our inversion to noisy synthetic data sets generated from horizontally stratified earth models for both isotropic and anisotropic assumptions and for different measurement configurations, we demonstrate the accuracy of our method. By comparing the results of our inversion with the regularized genetic algorithm, we also demonstrate the necessity of casting this problem as a multi-objective optimization for a better assessment of uncertainty as compared to a scalar objective optimization method.

Комбінування генетичних алгоритмів в елементах штучної нейронної мережі

Models of genetic algorithms and their combinations are considered in order to improve their functional capabilities in systems of artificial intelligence. The features of training of artificial neural networks with the help of parallel genetic algorithms for making decisions in systems of artificial intelligence are considered.

Solving University Course Timetabling Problem Using Parallel Genetic Algorithm

Solving University Course Timetabling Problem Using Parallel Genetic Algorithm

Real-time energy optimization of irrigation scheduling by parallel multi-objective genetic algorithms

The present work is motivated by the need to reduce the energy costs arising from the pressure demands of drip and sprinkling irrigation, compounded by the increase in the energy price in recent years. Researchers have demonstrated that proper operation of the irrigation network reduces associated pumping costs. The main challenge was to obtain the optimal operation parameters on near real-time due to the fact that the high complexity of the optimization problem requires a great computational effort. The classic approach to the problem imposes a strict fulfilment of minimum pressures as a restriction. This study, however, presents a new methodology for the reordering of irrigation scheduling, incorporating the constraint of daily volume requests for each hydrant. The methodology is capable of minimizing the cost of energy while maximizing pressures at the critical hydrants. Cost reductions of about 6–7% were reached for scenarios without pressure deficit for the case study. Greater computational efficiency was achieved by posing the problem from a multi-objective approach, on the one hand, and by establishing the parallel evaluation of the objective function, on the other. The speed-up obtained by combining a reduction in the number of function evaluations thanks to the faster convergence of the multi-objective approach and the reduction of the computational time due to the parallelization of the algorithm achieved results about 10 times faster. This improvement allowed the tool to be implemented for the daily optimization of irrigation requests.

Optimization Difficulty Indicator and Testing Framework for Water Distribution Network Complexity

In the last three decades, benchmark water distribution networks (WDNs) have provided a common testbed for new optimization algorithms and design approaches. However, deriving generalized and reliable conclusions from such benchmark WDNs is difficult because their optimization difficulty levels (ODLs) are either too low or too high (i.e., biased). Final solutions do not consistently converge to a global optimum for a WDN problem with a high ODL. In addition, little effort has been given to quantifying and comparing the ODLs of WDNs with different characteristics and conditions. In this study, an ODL indicator was developed for WDNs: the coefficient of variation of the final solution fitness values. An ODL quantification framework was also developed with two phases: (1) generating network layouts with various topological characteristics, and (2) quantifying the statistics of the final solution quality and ODL by using a global parallel genetic algorithm. The proposed indicator and framework were applied to the design of a dense-grid B-city network and large C network, and the results demonstrated their applicability to generating a WDN benchmark problem with the target ODL.

A Master Slave Parallel Genetic Algorithm for Feature Selection in High Dimensional Datasets

Feature Selection in High Dimensional Datasets is a combinatorial problem as it selects the optimal subsets from N dimensional data having 2N possible subsets. Genetic Algorithms are generally a good choice for feature selection in large datasets, though for some high dimensional problems it may take varied amount of time - few seconds, few hours or even few days. Therefore, it is important to use Genetic Algorithms that can give quality results in reasonably acceptable time limit. For this purpose, it is becoming necessary to implement Genetic Algorithms in an efficient manner. In this paper, a Master Slave Parallel Genetic Algorithm is implemented as a Feature Selection procedure to diminish the time intricacies of sequential genetic algorithm. This paper describes the speed gains in parallel Master-Slave Genetic Algorithm and also discusses the theoretical analysis of optimal number of slaves required for an efficient master slave implementation. The experiments are performed on three high-dimensional gene expression data. As Genetic Algorithm is a wrapper technique and takes more time to find the importance of any feature, Information Gain technique is used first as pre-processing task to remove the irrelevant features.

Using Parallel Genetic Algorithms for Estimating Model Parameters in Complex Reactive Transport Problems

In this study, we present the details of an optimization method for parameter estimation of one-dimensional groundwater reactive transport problems using a parallel genetic algorithm (PGA). The performance of the PGA was tested with two problems that had published analytical solutions and two problems with published numerical solutions. The optimization model was provided with the published experimental results and reasonable bounds for the unknown kinetic reaction parameters as inputs. Benchmarking results indicate that the PGA estimated parameters that are close to the published parameters and it also predicted the observed trends well for all four problems. Also, OpenMP FORTRAN parallel constructs were used to demonstrate the speedup of the code on an Intel quad-core desktop computer. The parallel code showed a linear speedup with an increasing number of processors. Furthermore, the performance of the underlying optimization algorithm was tested to evaluate its sensitivity to the various genetic algorithm (GA) parameters, including initial population size, number of generations, and parameter bounds. The PGA used in this study is generic and can be easily scaled to higher-order water quality modeling problems involving real-world applications.

Coarse grained parallel quantum genetic algorithm for reconfiguration and service restoration of electric power networks

In this paper, a Coarse Grained Parallel Quantum Genetic Algorithm (CGPQGA) is proposed to solve the network reconfiguration and restoration problems in distribution networks with the objective of reducing network losses, balancing load and improving the quality of voltage in the system. Based on t he parallel evolutionary concept and the insights of quantum theory, a model of parallel quantum computation was simulated. In this frame, there are some demes (sub-populations) and some universes (groups of populations), which are structured in super star-shaped topologies. A new migration scheme based on penetration theory is developed to control both the migration rate and direction adaptively between demes and a coarse grained quantum crossover strategy is devised among universes. The proposed approach is tested on 33-bus distribution networks in two cases the first case without distributed generators and the second case with distributed generators, with the aim of minimizing the losses of reconfigured network, and minimize number of operated switches, deviation of bus voltages, and loading of transformer (power loss) where the choice of the switches to be opened is based on the calculation of voltages at the system buses, real and reactive power flow through lines, real power losses and voltage deviations, using distribution load flow (DLF) program. Simulation results prove the effectiveness of the proposed methodology in solving the current challenges in this phase.

Experimental study on vibration instability model of rock slope based on parallel chaotic algorithms

Abstract In view of the irrationality of sensor arrangement scheme in the model test method of anti-dip rock slope vibration instability, the accuracy and efficiency of test data acquired are reduced, which cannot meet the requirements of real-time vibration simulation process. There are some limitations in the authenticity and scientificity of test results. A model test of rock slope vibration instability based on parallel chaotic algorithm is proposed. The method uses cement, sand, iron powder, clay and admixture to prepare rock mass materials, sets Teflon cloth with very low surface friction coefficient to simulate discontinuous joints of rock slope according to certain rules in the interior of slope, and adopts parallel chaotic genetic algorithm to optimize sensor placement scheme based on the corresponding relationship between algorithm and optimal parameters of sensor placement. The experimental results show that the failure modes of rock slope under vibration are as follows: crack development, slope spalling and collapse sliding; acceleration response enlarges obviously along the slope face, and horizontal acceleration dominates; the optimization effect and operation time of sensor placement scheme are significantly improved to avoid premature convergence and fall into local optimum solution.