Genetic Algorithm-based Search Research Articles

Genetic algorithm-based search heuristic for turbine balancing problem

Turbine balancing problem (TBP) can be defined as arranging blades of each stage of turbine engines in such a way that provides a minimum residual unbalance to eliminate the undesired vibration. This problem is regarded as an NP-hard problem and no exact solution has been found for such problems up to now. In each stage of the turbine, an optimal arrangement of blades around a disk by considering the difference in weight of blades to achieve a minimum residual unbalance is essential for preventing the early failure of turbine engines. In this article, a genetic algorithm (GA) has been proposed and applied to TBP for defining an optimal blade arrangement of a turbine stage. This algorithm proposes a suitable choice for operators of GA, based on permutation representation. By eliminating the shortcomings of previous researchers’ attempts with a simple form of GA, our proposed GA is capable of attaining an optimal arrangement of blades with reasonable computational efforts. Moreover, suitable values for parameters of GA including population size, crossover, and mutation rates of the proposed GA have been determined using practical sets of blades belonging to a Siemens gas turbine engine. The proposed GA with tuned parameters has been successfully employed for a variety of TBPs, and obtained results show a good convergence to an optimal value of residual unbalance with reasonable computational efforts.

An optimal sensor placement strategy for reliable expansion of mode shapes under measurement noise and modelling error

Modal expansion techniques are typically used to expand the experimental modal displacements at sensor positions to all unmeasured degrees of freedom. Since in most cases, sensors can be attached only at limited locations in a structure, an expansion is essential to determine mode shapes, strains, stresses, etc. throughout the structure which can be used for structural health monitoring. Conventional sensor placement algorithms are mostly aimed to make the modal displacements at sensor positions of different modes as linearly independent as possible. However, under the presence of modelling errors and measurement noise, an optimal location based on this criterion is not guaranteed to provide an expanded mode shape which is close to the real mode shape. In this work, the expected value of normal distance between the real mode shape and the expanded mode shape is used as a measure of closeness between the two entities. Optimal sensor locations can be determined by minimizing this distance. This new criterion is applied on a simple cantilever beam and an industrial milling tower. In both cases, by using an exhaustive search of all possible sensor configurations it was possible to find sensor locations which resulted in a significant reduction in the distance when compared to a conventional optimal sensor placement strategy. Sufficiently accurate sub-optimal sequential sensor placement algorithm is also suggested as an alternative to the exhaustive search which is then compared with a genetic algorithm-based search. The efficiency of this new sensor placement criterion is further verified using Monte Carlo simulations for some realistic modelling error conditions.

Specificity Analysis of Genome Based on Statistically Identical K-Words With Same Base Combination.

Goal: Individual characteristics are determined through a genome consisting of a complex base combination. This base combination is reflected in the k-word profile, which represents the number of consecutive k bases. Therefore, it is important to analyze the genome-specific statistical specificity in the k-word profile to understand the characteristics of the genome. In this paper, we propose a new k-word-based method to analyze genome-specific properties. Methods: We define k-words consisting of the same number of bases as statistically identical k-words. The statistically identical k-words are estimated to appear at a similar frequency by statistical prediction. However, this may not be true in the genome because it is not a random list of bases. The ratio between frequencies of two statistically identical k-words can then be used to investigate the statistical specificity of the genome reflected in the k-word profile. In order to find important ratios representing genomic characteristics, a reference value is calculated that results in a minimum error when classifying data by ratio alone. Finally, we propose a genetic algorithm-based search algorithm to select a minimum set of ratios useful for classification. Results: The proposed method was applied to the full-length sequence of microorganisms for pathogenicity classification. The classification accuracy of the proposed algorithm was similar to that of conventional methods while using only a few features. Conclusions: We proposed a new method to investigate the genome-specific statistical specificity in the k-word profile which can be applied to find important properties of the genome and classify genome sequences.

Stock Price Prediction Using Fuzzy Time-Series Population Based Gravity Search Algorithm

The main motive of this research is to predict the future stock value of the particular day with minimum variation from the actual value of stock. In this research, a genetic algorithm-based gravity search algorithm is proposed for stock market prediction. It will be helpful for short-term investors in the National stock market. Some important factors that affect the value of stock are total stocks traded, total turnover of the company, gross domestic product (GDP) of the country, GDP per capita and political or external factors are some of the main factors that affect the stock value of that particular day. Opening and closing values of the stock market were predicted with the help of the above factors. Each factor will be considered as an object with mass, the mass of every object will be based on the importance. With the help of a Gravitational Search Algorithm (GSA) [1], the converging point of the entire object is determined and it is said to be the optimal output of the algorithm. The input considered are opening, closing, low and high values for a period of one year.

A multi-objective approach for supply chain design considering disruptions impacting supply availability and quality

Abstract We develop a multi-objective stochastic programming model to explore tradeoffs between costs and risk in the supply chain in the event a disruption occurs. We explicitly consider network configuration and operating cost under normal conditions, cost of unsatisfied demand, cost of shipping tainted products to a customer, and quality inspection cost as conflicting goals to be minimized simultaneously. We analyze different disruption scenarios to determine the best supplier selection and inspection strategies to mitigate the effect of disruptions on supply availability and quality. Even the single-objective version of this problem is NP-hard; thus, we propose a genetic algorithm-based search method to identify Pareto-optimal supply chain configurations. We use data envelopment analysis for calculating the fitness value of various supply chain configurations. The proposed approach efficiently yields high-quality supply chain designs, allowing the decision maker to determine an appropriate tradeoff between various costs.

Amber-Compatible Parametrization Procedure for Peptide-like Compounds: Application to 1,4- and 1,5-Substituted Triazole-Based Peptidomimetics.

Peptidomimetics are molecules of particular interest in the context of drug design and development. They are proteolytically and metabolically more stable than their natural peptide counterparts but still offer high specificity toward their biological targets. In recent years, 1,4- and 1,5-disubstituted 1,2,3-triazole-based peptidomimetics have emerged as promising lead compounds for the design of various inhibitory and tumor-targeting molecules as well as for the synthesis of peptide analogues. The growing popularity of triazole-based peptidomimetics and a constantly broadening range of their application generated a demand for elaborate theoretical investigations by classical molecular dynamics simulations and molecular docking. Despite this rising interest, accurate and coherent force field parameters for triazole-based peptidomimetics are still lacking. Here, we report the first complete set of parameters dedicated to this group of compounds, named TZLff. This parametrization is compatible with the latest version of the AMBER force field (ff14SB) and can be readily applied for the modeling of pure triazole-based peptidomimetics as well as natural peptide sequences containing one or more triazole-based modifications in their backbone. The parameters were optimized to reproduce HF/6-31G* electrostatic potentials as well as MP2/cc-pVTZ equilibrium Hessian matrices and conformational potential energy surfaces through the use of a genetic algorithm-based search and least-squares fitting. Following the standards of AMBER, we introduce residue building units, thus allowing the user to define any given sequence of triazole-based peptidomimetics. Validation of the parameter set against ab initio- and NMR-based reference systems shows that we obtain fairly accurate results, which properly capture the conformational features of triazole-based peptidomimetics. The successful and efficient parametrization strategy developed in this work is general enough to be applied in a straightforward manner for parametrization of other peptidomimetics and, potentially, any polymeric assemblies.

A genetic algorithm-based search space splitting pattern and its application in hydraulic and coastal engineering problems

This article reports a search space splitting pattern that can be applied to genetic algorithms in order to ensure that the entire search space is investigated. Hence, by keeping the genetic algorithm simple, in a reasonable time and with a high degree of accuracy, the initial solutions can be improved toward the global optimum point. The simplicity of the presented method is an advantage that makes it useful for applied hydraulic and coastal engineering problems. The performance of the proposed method was evaluated by a benchmark optimization problem, Levy No. 5, and three hydraulic and coastal engineering problems: inverse problem of Manning’s equation, the equation of equilibrium beach profiles, and the settling velocity equation of natural sediment particles. The results indicated that the nonlinear complex problems can be solved by the proposed method with a high degree of accuracy. The proposed genetic algorithm-based search space splitting pattern can either be used exclusively or alternatively it can be combined with improved operators in the literature.

Energy-efficient resource allocation for heterogeneous cognitive radio network based on two-tier crossover genetic algorithm

Cognitive radio (CR) is considered an attractive technology to deal with the spectrum scarcity problem. Multi-radio access technology (multi-RAT) can improve network capacity because data are transmitted by multiple RANs (radio access networks) concurrently. Thus, multi-RAT embedded in a cognitive radio network (CRN) is a promising paradigm for developing spectrum efficiency and network capacity in future wireless networks. In this study, we consider a new CRN model in which the primary user networks consist of heterogeneous primary users (PUs). Specifically, we focus on the energy-efficient resource allocation (EERA) problem for CR users with a special location coverage overlapping region in which heterogeneous PUs operate simultaneously via multi-RAT. We propose a two-tier crossover genetic algorithm-based search scheme to obtain an optimal solution in terms of the power and bandwidth. In addition, we introduce a radio environment map to manage the resource allocation and network synchronization. The simulation results show the proposed algorithm is stable and has faster convergence. Our proposal can significantly increase the energy efficiency.

Heterogeneous Ensemble Combination Search Using Genetic Algorithm for Class Imbalanced Data Classification.

Classification of datasets with imbalanced sample distributions has always been a challenge. In general, a popular approach for enhancing classification performance is the construction of an ensemble of classifiers. However, the performance of an ensemble is dependent on the choice of constituent base classifiers. Therefore, we propose a genetic algorithm-based search method for finding the optimum combination from a pool of base classifiers to form a heterogeneous ensemble. The algorithm, called GA-EoC, utilises 10 fold-cross validation on training data for evaluating the quality of each candidate ensembles. In order to combine the base classifiers decision into ensemble’s output, we used the simple and widely used majority voting approach. The proposed algorithm, along with the random sub-sampling approach to balance the class distribution, has been used for classifying class-imbalanced datasets. Additionally, if a feature set was not available, we used the (α, β) − k Feature Set method to select a better subset of features for classification. We have tested GA-EoC with three benchmarking datasets from the UCI-Machine Learning repository, one Alzheimer’s disease dataset and a subset of the PubFig database of Columbia University. In general, the performance of the proposed method on the chosen datasets is robust and better than that of the constituent base classifiers and many other well-known ensembles. Based on our empirical study we claim that a genetic algorithm is a superior and reliable approach to heterogeneous ensemble construction and we expect that the proposed GA-EoC would perform consistently in other cases.

Optimization of Cyclic-Delay Diversity Aided Frequency-Selective Scheduling in OFDMA Downlink Systems

Cyclic-delay diversity (CDD)-aided frequency-selective scheduling has been known as an effective technique for increasing the system capacity of orthogonal frequency-division multiple-access (OFDMA) systems. By increasing the channels' frequency selectivity with different cyclic delays applied on transmit antennas, the merit of multiuser diversity can be more effectively exploited in the system. In this paper, we aim to optimize the design of the CDD-aided frequency-selective scheduling for OFDMA downlink systems. Optimization is done from two aspects: cyclic-delay search and scheduling-information feedback. In the optimization of cyclic-delay search, three new methods are proposed, including antenna-wise sequential search for frequency-nonselective (AWSS-FNS) channels, AWSS for frequency-selective (AWSS-FS) channels, and a genetic-algorithm-based search (GAS). AWSS-FNS achieves the optimal performance if a channel is frequency nonselective over a subchannel, whereas AWSS-FS and GAS provide good tradeoffs between system performance and complexity for channels that are frequency selective over a subchannel. In the optimization of scheduling-information feedback, a new channel-dependent feedback (CDF) method is proposed, where the feedback resource is allocated to users according to their frequency-selectivity rates and a principle of proportional fairness. The optimized scheduler is shown to provide significant improvement in both system sum rate and user fairness over previous methods under fixed feedback overhead.

A single-objective genetic-fuzzy approach for multi-objective fuzzy problems

In this paper, a genetic algorithm-based search method, which builds ideal rule set for fuzzy rule-based classification systems FRBCSs, is developed. In FRBCSs, ideal rule set means a set of rules which ensure high classification accuracy with small rule count and small rule length. The related studies in the literature point out that rule set grows exponentially with input attribute count. This growth complicates the searching process and lowers the success rate. Through the proposed method, successive results are obtained for datasets with large input attribute counts using a special coding technique. The proposed method is tested for various datasets and results are compared against the method which uses candidate rule set.

Process parameters optimisation of CNC drilling operation using genetic algorithm

Drilling is an important operation in sheet metal shop. Mathematically correlating the input control variables with responses and subsequently optimising them will lead to better machining performance. The efficiency of machining processes can be improved by process parameters optimisation which identifies and determines the regions of critical process control factors leading to desired responses. Genetic algorithm (GA) is a global optimisation technique and can be applied without recourse to domain-specific heuristics. Considering the multifacet advantages of genetic algorithm, the optimisation of CNC drilling operation has been carried out in this research using this technique. Through genetic algorithm, minimum axial force obtained is 17.48 N which is 2.52 N less in magnitude than response surface methodology (RSM) predicted result. Also, the minimum torque obtained through genetic algorithm-based search is 0.208 Nm which is far less than the RSM predicted result. This clearly shows the advantages of using GA-based approach over other techniques.

The use of a genetic algorithm-based search strategy in geostatistics: application to a set of anisotropic piezometric head data

In any geostatistical study, an important consideration is the choice of an appropriate, repeatable, and objective search strategy that controls the nearby samples to be included in the location-specific estimation procedure. Almost all geostatistical software available in the market puts the onus on the user to supply search strategy parameters in a heuristic manner. These parameters are solely controlled by geographical coordinates that are defined for the entire area under study, and the user has no guidance as to how to choose these parameters. The main thesis of the current study is that the selection of search strategy parameters has to be driven by data—both the spatial coordinates and the sample values—and cannot be chosen beforehand. For this purpose, a genetic-algorithm-based ordinary kriging with moving neighborhood technique is proposed. The search capability of a genetic algorithm is exploited to search the feature space for appropriate, either local or global, search strategy parameters. Radius of circle/sphere and/or radii of standard or rotated ellipse/ellipsoid are considered as the decision variables to be optimized by GA. The superiority of GA-based ordinary kriging is demonstrated through application to the Wolfcamp Aquifer piezometric head data. Assessment of numerical results showed that definition of search strategy parameters based on both geographical coordinates and sample values improves cross-validation statistics when compared with that based on geographical coordinates alone. In the case of a variable search neighborhood for each estimation point, optimization of local search strategy parameters for an elliptical support domain—the orientation of which is dictated by anisotropic axes—via GA was able to capture the dynamics of piezometric head in west Texas/New Mexico in an efficient way.

GA-based multiresolution fusion of segmented brain images using PD-, T1- and T2-weighted MR modalities

Medical image fusion has been used to derive the useful complimentary information from multimodality imaging. The proposed methodology introduces fusion approach for robust and automatic extraction of information from segmented images of different modalities. This fusion strategy is implemented in multiresolution domain using wavelet transform- and genetic algorithm-based search technique to extract maximum complementary information. The analysis of input images at multiple resolutions is able to extract more fine details and improves the quality of the composite fused image. The proposed approaches are also independent of any manual marking or knowledge of fiducial points and start the fusion procedure automatically. The performance of fusion scheme implemented on segmented brain images has been evaluated computing mutual information as similarity measuring matrix. Prior to fusion process, images are being segmented using different segmentation techniques like fuzzy C-mean and Markov random field models. Experimental results show that Gibbs- and ICM-based segmentation approaches related to Markov random field perform over the fuzzy C-mean and which are being used prior to GA-based fusion process for MR T1, MR T2 and MR PD images of section of human brain.

Metabolic Profiling Uncovers a Phenotypic Signature of Small for Gestational Age in Early Pregnancy

Being born small for gestational age (SGA) confers increased risks of perinatal morbidity and mortality and increases the risk of cardiovascular complications and diabetes in later life. Accumulating evidence suggests that the etiology of SGA is usually associated with poor placental vascular development in early pregnancy. We examined metabolomic profiles using ultra performance liquid chromatography-mass spectrometry (UPLC-MS) in three independent studies: (a) venous cord plasma from normal and SGA babies, (b) plasma from a rat model of placental insufficiency and controls, and (c) early pregnancy peripheral plasma samples from women who subsequently delivered a SGA baby and controls. Multivariate analysis by cross-validated Partial Least Squares Discriminant Analysis (PLS-DA) of all 3 studies showed a comprehensive and similar disruption of plasma metabolism. A multivariate predictive model combining 19 metabolites produced by a Genetic Algorithm-based search program gave an Odds Ratio for developing SGA of 44, with an area under the Receiver Operator Characteristic curve of 0.9. Sphingolipids, phospholipids, carnitines, and fatty acids were among this panel of metabolites. The finding of a consistent discriminatory metabolite signature in early pregnancy plasma preceding the onset of SGA offers insight into disease pathogenesis and offers the promise of a robust presymptomatic screening test.

Genetic algorithm-based search on the role of variables in the work hardening process of multiphase steels

Abstract Multi-objective optimizations of strength and ductility of multiphase steels are conducted using genetic algorithms (GAs), to investigate the role of the composition and process variables in their complicated work hardening process. Neural network-based computational models, describing the complex correlations between the decision parameters for processing and materials chemistry of such steels, are developed using existing data and are used for the fitness functions. The cases of both high-strength low-alloy steel (HSLA) and the transformation-induced plasticity (TRIP)-aided steel are separately studied, and the findings are compared and contrasted. The Pareto solutions are used successfully to study the role of the parameters at different combinations of strength and ductility. The findings are also utilized for qualitative assessment of the dominant mechanisms behind the work hardening of the steels.

Enhanced performance of overlap flow-shop scheduling involving reworking and a time buffer

This study presents an overlap scheduling solution to enhance system performance for branched-line flow-shop operations that involve reworking and a time buffer. In scheduling the branched-line operations involving reworking and a time buffer, two mutual inter-dependent problems must be solved: one is to determine the appropriate job schedule, and the other is to find a suitable time buffer at the right position of the system. Analysis results indicate that arranging a job sequence using a meta-search, of which the genetic algorithm-based best-fit search is the simplest, most effective, and least costly way for improving overlap flow-shop scheduling performance. In determining the job schedule involving reworking and a time buffer for branched-line operations, we propose a genetic algorithm-based overlap scheduling heuristic to perform the best-fit search using a pre-determined time buffer. Having found a suitable time buffer, we build a scheduling simulator to measure the performance for the job schedule that is determined by the genetic algorithm when a time buffer is added.

FLIPDock: Docking flexible ligands into flexible receptors

Conformational changes of biological macromolecules when binding with ligands have long been observed and remain a challenge for automated docking methods. Here we present a novel protein-ligand docking software called FLIPDock (Flexible LIgand-Protein Docking) allowing the automated docking of flexible ligand molecules into active sites of flexible receptor molecules. In FLIPDock, conformational spaces of molecules are encoded using a data structure that we have developed recently called the Flexibility Tree (FT). While the FT can represent fully flexible ligands, it was initially designed as a hierarchical and multiresolution data structure for the selective encoding of conformational subspaces of large biological macromolecules. These conformational subspaces can be built to span a range of conformations important for the biological activity of a protein. A variety of motions can be combined, ranging from domains moving as rigid bodies or backbone atoms undergoing normal mode-based deformations, to side chains assuming rotameric conformations. In addition, these conformational subspaces are parameterized by a small number of variables which can be searched during the docking process, thus effectively modeling the conformational changes in a flexible receptor. FLIPDock searches the variables using genetic algorithm-based search techniques and evaluates putative docking complexes with a scoring function based on the AutoDock3.05 force-field. In this paper, we describe the concepts behind FLIPDock and the overall architecture of the program. We demonstrate FLIPDock's ability to solve docking problems in which the assumption of a rigid receptor previously prevented the successful docking of known ligands. In particular, we repeat an earlier cross docking experiment and demonstrate an increased success rate of 93.5%, compared to original 72% success rate achieved by AutoDock over the 400 cross-docking calculations. We also demonstrate FLIPDock's ability to handle conformational changes involving backbone motion by docking balanol to an adenosine-binding pocket of protein kinase A.

Search for Initial Conditions for Sustained Hopping of Passive Springy-Leg Offset-Mass Hopping Robot

A springy-leg, offset-mass or SLOM hopper configuration is considered here, wherein the center of gravity of the body is offset from the line of action of spring force. It is observed that this feature tends to extend the passive hopping motion. A heuristics-based search strategy and a genetic algorithm-based search strategy are implemented for finding the initial conditions that result in extended hopping motion.

Genetic Algorithm-Assisted Design of Adaptive Predictive Filters for 50/60 Hz Power Systems Instrumentation

We introduce a genetic algorithm-based method for structural optimization of multiplicative general parameter (MGP) finite impulse response (FIR) filters. These computationally efficient reduced-rank adaptive filters are robust, suitable for predictive configurations, and they have numerous applications in 50/60 Hz power systems instrumentation. The design process of such filters has three independent stages: Lagrange multipliers-based optimization of the sinusoid-predictive basis filter, genetic algorithm-based search of optimal FIR tap cross-connections and, finally, the online MGP-adaptation phase guided by variations in signal statistics. Thus, our multistage design procedure is a complementary fusion of hard computing (HC) and soft computing (SC) methodologies. Such advantageous fusion (or symbiosis) thinking is emerging among researchers and practicing engineers, and it can potentially lead to competitive combinations of individual HC and SC methods.