Genetic Algorithm Search Method Research Articles

Optimisation of Mixed Polarity Reed-Muller Functions

This paper presents a genetic algorithm (GA) search method in order to obtain better circuit implementation of the mixed polarity Reed-Muller functions. By combining global searching ability of genetic algorithm and local searching ability of simulated annealing, the proposed GA method could achieve fast convergence. It differs to traditional genetic algorithm, in which the proposed GA forms an intermediate population by using 2/3 population from previous generation and 2/3 population from current generation at the annealing stage. Annealing is then applied to the intermediate population to generate a new population. In the next generation selection, crossover and mutation operations are used for the newly generated population. The calculation of cost function of proposed algorithm is based on parallel tabular technique to overcome the disadvantage of the traditional tabular technique. The results of the tested benchmark indicated that this algorithm is highly effective for searching the best polarity and it could achieve 29% area reduction and 3.68X speedup.

Genetic Algorithm Based Finite State Markov Channel Modeling

Statistical properties of the error sequences produced by fading channels with memory have a strong influence over the performance of high layer protocols and error control codes. Finite State Markov Channel (FSMC) models can represent the temporal correlations of these sequences efficiently and accurately. This paper proposes a simple genetic algorithm (GA) based search for the optimum state transition matrix for a block diagonal Markov model. The burst error statistics of the GA based FSMC model with respect to Autocorrelation Function and error free interval distribution of the original error sequence are presented to validate the proposed method. The superiority of the GA approach over the semi-hidden Markov model (SHMM) based Fritchman model is exhibited in significant improvement of closeness of match and in the usage of shorter length of error sequences. Another Baum-Welch algorithm (BWA) based GA search method has been proposed and compared with the BWA and SHMM methods for the same error sequence. Again the superiority of GA approaches is recognized, especially for the smaller error lengths.

유전자 알고리즘 및 패턴 서치 방법을 이용한 풍력 터빈 블레이드의 형상 최적화

이 연구에서는 풍력 터빈 블레이드의 형상 최적화를 위한 직접탐색 기반의 최적화 기법을 적용하고, 최적화 기법간의 성능을 비교하여 효과적인 방법을 제안하고자 하였다. 이를 위하여 수평축 풍력 터빈의 최적설계 코드인 HARP_Opt(Horizontal Axis Rotor Performance Optimizer)을 기반으로 연간 발전량 평가 방법을 수정하고, HARP_Opt에서 적용하고 있는 기존의 유전자 알고리즘과 함께 패턴 서치 방법을 추가 적용하였다. 이를 1MW급 풍력 발전 터빈 블레이드의 단면 형상 최적 설계 문제에 적용하였으며, 기존의 유전자 알고리즘 및 마이크로 유전자 알고리즘, 그리고 패턴 서치 방법의 성능을 비교한 결과, 연간 발전량과 해의 일관성 면에 있어서는 패턴 서치 방법이 상대적으로 우수하였으며, 계산시간 측면에서는 마이크로 유전자 알고리즘이 상대적으로 우수한 것으로 분석되었다. In this study, direct-search based optimization methods are applied for blade shape optimization of wind turbines and the optimization performances of several methods including conventional genetic algorithm, micro genetic algorithm and pattern search method are compared to propose a more efficient method. For this purpose, the currently available version of HARP_Opt (Horizontal Axis Rotor Performance Optimizer) code is enhanced to rationally evaluate the annual energy production value according to control strategies and to optimize the blade shape using pattern search method as well as genetic algorithm. The enhanced HARP_Opt code is applied to obtain the optimal turbine blade shape for 1MW class wind turbines. The results from pattern search method are compared with the results from conventional genetic algorithm and also micro genetic algorithm and it is found that the pattern search method has a better performance in achieving higher annual energy production and consistent optimal shapes and the micro genetic algorithm is better for reducing the calculation time.

Heterogeneous ensemble approach with discriminative features and modified-SMOTEbagging for pre-miRNA classification

An ensemble classifier approach for microRNA precursor (pre-miRNA) classification was proposed based upon combining a set of heterogeneous algorithms including support vector machine (SVM), k-nearest neighbors (kNN) and random forest (RF), then aggregating their prediction through a voting system. Additionally, the proposed algorithm, the classification performance was also improved using discriminative features, self-containment and its derivatives, which have shown unique structural robustness characteristics of pre-miRNAs. These are applicable across different species. By applying preprocessing methods—both a correlation-based feature selection (CFS) with genetic algorithm (GA) search method and a modified-Synthetic Minority Oversampling Technique (SMOTE) bagging rebalancing method—improvement in the performance of this ensemble was observed. The overall prediction accuracies obtained via 10 runs of 5-fold cross validation (CV) was 96.54%, with sensitivity of 94.8% and specificity of 98.3%—this is better in trade-off sensitivity and specificity values than those of other state-of-the-art methods. The ensemble model was applied to animal, plant and virus pre-miRNA and achieved high accuracy, >93%. Exploiting the discriminative set of selected features also suggests that pre-miRNAs possess high intrinsic structural robustness as compared with other stem loops. Our heterogeneous ensemble method gave a relatively more reliable prediction than those using single classifiers. Our program is available at http://ncrna-pred.com/premiRNA.html.

Direct Torque Control of the Genetic Neural Network under the Low-Speed

For the direct torque control of asynchronous motor rapid response and speed ripple, this article proposes the genetic neural network algorithm based on the model of stator flux, to achieve the selecting of switch state under the low-speed. Using the global optimization and search method of genetic algorithm obtains global optimal solution, while the connection weights and network structure learning improve the training effectiveness of neural network, so that the BP network has better adaptive characteristic. The effectiveness of the design is verified by the simulation, and it shows that the speed control system has good dynamic performance and steady state performance under the low-speed.

Efficient Capacity-Based Joint Quantized Precoding and Transmit Antenna Selection Using Cross-Entropy Method for Multiuser MIMO Systems

The current paper considers the joint precoding and transmit antenna selection to reduce hardware cost, such as radio-frequency chains, associated with antennas in the downlink of multiuser multiple-input multiple-output systems with limited feedback. The joint precoding and transmit antenna selection algorithm requires an exhaustive search of all possible combinations and permutations to find the optimum solution at the transmitter, thus resulting in extremely high computational complexity. To reduce the computational load while still maximizing channel capacity, the cross-entropy (CE) method is adopted to determine the suboptimum solution. Compared with the conventional genetic algorithm and random search method, the CE method provides better performance under the same computational complexity, as shown by the simulation results.

Design of a single-DOF kinematic chain using hybrid GA-pattern search and sequential GA

With the aim of obtaining an optimal design for a single-DoF kinematic chain with the function of morphing aerofoils, two search methods are developed. The assessment criteria arethe length of the linkage and the accuracy of mapping. While the former is taken as the objective of optimization, the latter is treated as a constraint. The accuracy in mapping is measured by the aerodynamic performance deviation, a parameter defined by combining the geometric deviation and a weighting function based on the distribution of pressure coefficient. The first search method is based on a genetic algorithm (GA) with an embedded pattern search algorithm. The aim of the pattern search is to reduce the number of failed attempts in generating feasible solutions for the initial population of the GA as well as repairing infeasible individuals produced by reproduction operators. In the second search method, the set of design variables is determined in two consecutive steps by employing a sequential GA. Results of five runs of each search method reveal that while the best solution was produced by the sequential GA search method, hybrid GA-pattern search yielded more consistent results with shorter lengths and better mapping precision in average.

Optimal integrated planning of MV–LV distribution systems using DPSO

A new technique for optimal planning of MV and LV segments of a distribution system is presented in this paper. The main goal is to find optimally distribution transformer and substation locations and ratings, as well as, the route and type of Medium Voltage (MV) and Low Voltage (LV) feeders. The proposed technique is applicable to both uniform and non-uniform load densities areas. In this method, the planning area is divided into regions with relatively uniform load density such as urban, semi-urban, sub-urban. Each of regions is divided into zones, called LV zone. Each LV zone is supplied by an MV/LV transformer. The dimensions of LV zones are found based on the average load of each region. The placement and rating of MV/LV transformers, the type and route of LV conductors in an LV zone all depend on its loads’ location and power. Regarding the placement and rating of MV/LV transformers in planning area and the space of regions, the dimensions of a zone which is supplied by a HV/MV transformer, called MV zone, is determined. Additionally, the location and rating of HV/MV transformers as well as the feeder's routes and types are calculated. Since the dimensions of an LV zone influence the associated length of MV feeder, the MV feeder cost needs to be included in the total cost associated with the LV zone. This requires the MV feeder type to be known to calculate the corresponding cost. However, the MV feeder type is determined as an output from MV zone planning. As a result, an iterative based method is proposed to consider this common element in computations to develop integrated planning of both LV and MV zones. It is observed that the iterative technique quickly converges to the same results as the exhaustive search method. Discrete particle swarm optimization (DPSO) method is employed for solving the planning problem. The results are compared with nonlinear programming, genetic algorithm and exhaustive search methods. It is observed that DPSO is as accurate as the exhaustive search method for integrated planning of MV–LV distribution systems while its computation time is significantly lower.

Optimization of Radiant Heater Power for Heating of Flat Plates Using Metaheuristic Methods

In this article, simulated annealing (SA), genetic algorithm (GA), and tabu search (TS) methods are employed separately and together for finding optimal radiant heater settings for obtaining a specific temperature profile on a design surface. Results show that all methods yield to very acceptable results for various desired temperature profiles but after different number of iterations. The GA and TS methods progress very rapidly before slowly stabilizing after a certain number of iterations, while the SA method progresses for longer periods. In addition, hybrid combinations of GA, SA, and TS proved to be better than separate ones.

Sensitivity analysis and thermo-economic optimisation of Steam Injected Gas Turbine cycle design parameters

The purpose of the present work is to apply thermo-economic concepts to a Steam Injected Gas Turbine (STIG) electricity generation system and optimise it to indicate its economical advantages over a simple gas turbine. The system incorporates a combustion chamber, gas turbine, a compressor and a Heat Recovery Steam Generator (HRSG). A detailed sensitivity analysis was also performed to examine the impacts of the variation of the operating parameters. Genetic algorithm and direct search methods were used to find optimum values. The system is optimised, using thermo-economic concepts. The results demonstrate that the steam injected system shows a better thermodynamic and thermo-economic performance.

Genetic algorithm and Tabu search based methods for molecular 3D-structure prediction

The search for the global minimum of a potential energy function is very difficult since the number of local minima grows exponentially with the molecule size. The present work proposes the application of genetic algorithm and tabu search methods, which are called GAMCP (Genetic Algorithm with Matrix Coding Partitioning) [7], and TSVP (Tabu Search with Variable Partitioning) [8], respectively, for minimizing the molecular potential energy function. Computational results for problems with up to 200 degrees of freedom are presented and are favorable compared with other four existing methods from the literature. Numerical results show that the proposed two methods are promising and produce high quality solutions with low computational costs.

Optimal design of multilayered composite plate using bio-inspired optimisation techniques

Fibre reinforced composite materials are widely used in construction of components of aircraft, ships and marine structures that are subjected to high compression loads. One of the important problems in fibre reinforced composite markets is to find the optimal stacking sequence which tailors the properties of the composite structures. The motivation of the study is to apply the newly developed bio-inspired technique particle swarm optimisation (PSO) and compare its effectiveness with other optimisation techniques genetic algorithm (GA) and pattern search method in the design of composite structures. As a case study, a square and rectangular thick plate subjected to uniaxial or biaxial loading is considered and it is optimised to find the layer orientation which maximises the buckling load. This work showed that GA performs better than PSO and pattern search method and so is the preferable method in design of composite plate structures.

Application of Genetic Algorithm for Optimization of Important Parameters of Magnetically Biased Microstrip Circular Patch Antenna

The application of Genetic Algorithm (GA) to the optimization of important parameters (Directivity, Radiated Power, Impedance etc.) of magnetically biased microstrip antenna, fabricated on ferrite substrate, is reported. The fitness functions for the GA program have been developed using cavity method for the analysis of microstrip antenna. The effect of external magnetic biasing has also been incorporated in the fitness function formulation as effective propagation constant. Using stochastic based search method of GA the common characteristics of electro-magnetic were entertained which cannot be handled by other optimization techniques. The genetic algorithm was run for 500 generations. The computed results are in good agreement with the results obtained experimentally.

Search target band gap with inverse band structure approach for complex superlattices

Because it is too difficult to pick out the needed structures from the vast possible configurations, complex superlattices have not been studied well for a long time. In this paper, an inverse band structure (IBS) approach which combined genetic algorithm search method with an empirical spds* tight-binding energy band calculation to address this problem is presented. Needed direct energy band gaps of Ga(Al)As complex superlattices are found by using this approach. It can be found that the band gap value can be same for different superlattice structures, but other properties of these complex superlattices could be different.

Docking Study and Three‐Dimensional Quantitative Structure‐Activity Relationship (3D‐QSAR) Analyses of Transforming Growth Factor‐β Type I Receptor Kinase Inhibitors

AbstractTransforming growth factor‐β Type I Receptor Kinase (TGF‐β RI) is an important and novel drug target for the treatment of fibrosis and cancer. In this study, docking and comparative molecular field analysis (CoMFA) were performed and the structural characteristics of TGF‐β RI ligands were rationalized. The genetic algorithm search method in the docking program GOLD 3.0.1 was employed to determine the likely binding mode conformations of 70 inhibitors in the active site of TGF‐β RI. Based on the binding mode conformations, a highly predictive 3D‐QSAR model was developed with a q2 value of 0.589 for CoMFA. The predictive ability of this model was validated with a number of compounds that were not included in the original training set. Furthermore, the 3D‐QSAR model was mapped back to the binding site of the TGF‐β RI to obtain a better understanding of the essential interactions between the inhibitors and TGF‐β RI. The robustness, predictive ability and automated alignment generation of this model make it a potential tool for the design and development of new drug leads to inhibit TGF‐β RI.

Digital Pattern Search and Its Hybridization with Genetic Algorithms for Bound Constrained Global Optimization

In this paper, we present a recently developed pattern search method called Genetic Pattern Search algorithm (GPSA) for the global optimization of cost function subject to simple bounds. GPSA is a combined global optimization method using genetic algorithm (GA) and Digital Pattern Search (DPS) method, which has the digital structure represented by binary strings and guarantees convergence to stationary points from arbitrary starting points. The performance of GPSA is validated through extensive numerical experiments on a number of well known functions and on robot walking application. The optimization results confirm that GPSA is a robust and efficient global optimization method.

A framework to integrate history matching and geostatistical modeling using genetic algorithm and direct search methods

History matching is an inverse problem where the reservoir model is modified in order to reproduce field observed data. Traditional history matching processes are executed separately from the geological and geostatistical modeling stage due to the complexity of each area. Changes made directly on the reservoir properties generally yield inconsistent geological models. This work presents a framework to integrate geostatistical modeling and history matching process, where geostatistical images are treated as matching parameters. The traditional optimization methods normally applied in history matching generally use gradient information. The treatment of geostatistical images as matching parameters is difficult for these methods due to the strong non-linearities in the solution space. Therefore, another objective of this work is to investigate the application of two optimization methods: genetic algorithm and direct search method in the proposed framework. In order to accelerate the optimization process, two additional techniques are used: upscaling and distributed computing. Results are presented showing the viability of the genetic algorithm in the type of problem addressed in this work and also that direct search method can be used with some restriction. Finally, the benefits of distributed computing and the consistence of the upscaling process are shown.

Insights on the March 1998 eruption at Piton de la Fournaise volcano (La Réunion) from microgravity monitoring

We investigate the temporal gravity changes associated with one of the major recent eruptions at Piton de la Fournaise volcano (March 1998) that occurred after an unusual five‐year quiet period and initiated a new eruptive cycle. Repeated microgravity surveys allowed us to measure residual gravity changes up to 100 μGal within the Enclos Fouqué caldera for four months before the start of the eruption. We first analyzed the temporal gravity changes and the height changes, also measured at the gravity benchmarks, on the basis of an intrusive dyke model previously proposed for this eruption from InSAR and tilt data. This analysis reveals that such simple model (finite rectangular shaped tensile dislocation) cannot fit both gravity and geodetic data and leads us to propose a dual source gravity model for this eruption. The gravity data inversion using a genetic algorithm search method allows us to quantify the mass change produced by the intrusive dyke (3.9 to 8.7 × 109kg) and to suggest an additional mass increase at sea level depth (4.6 to 7.2 × 1010kg). We interpret the latter as the effect of magma ascent into a reservoir that occurred within the four months before the eruption. We suggest that high concentrations of gas content in the magma reservoir account for a higher compressibility of the intruded magma and produce significant mass change with moderate surface deformation. The 1996 seismic crisis, interpreted as an upward migration of magmatic fluids into a reservoir located at sea level, might support such a hypothesis.

The Minimum Capacity of HVAC Secondary Systems (with Capacity Reduction by Interzonal Airflow) (RP-1049)

To date, the performance of alternative HVAC secondary systems has been compared using either the systems' energy use directly or by a life-cycle cost analysis. This paper introduces the concept of rating a system's performance by comparing its capacity at a particular operating point to the thermodynamic minimum capacity. A simple ratio, termed the “system effectiveness,” is also introduced to indicate the extent to which the system operates with the minimum possible capacity. This paper describes the calculation of the minimum system capacity as a nonlinear, single-criterion, constrained optimization problem. In particular, it describes the case for the minimization of the system capacity by use of interzonal airflow (the interzonal airflow rates and zone thermal conditions being variables of the optimization). This optimization problem is multi-modal in that the same system capacity can result from more than one interzonal airflow configuration. The optimization problem has been solved here using a genetic algorithm (GA) search method. This paper illustrates the use of the minimum system capacity as a benchmark for the performance of a typical multizone heating, ventilating, and air-conditioning (HVAC) system. The example also illustrates the potential benefit of using interzonal airflow to reduce the required capacity of a system. It can be concluded from the example that the arrangement of the system components is a significant factor in determining the thermal effectiveness of HVAC systems.

Knowledge-Based Evolution Method for Optimizing Refrigerant Circuitry of Fin-and-Tube Heat Exchangers

Fin-and-tube heat exchangers are widely used in HVAC applications, and the optimization of the refrigerant circuitry (RC) is important to improve their performance. The genetic algorithm (GA) is a commonly used optimization method but cannot be directly used for RC optimization. This paper presents a new optimization method based on the GA, which is simpler and more effective for RC optimization than existing methods. The optimization method for RC optimization presented in this paper is a knowledge-based evolution method (KBEM) that includes an improved genetic algorithm (IGA) and domain knowledge-based RC search methods. The IGA, as the basis of the KBEM, uses a tailor-made RC coding method, RC generation method, and improved genetic operators. The knowledge-based RC search methods increase the optimization efficiency of the IGA by reasonably reducing the search space according to the domain knowledge. Case studies show that the KBEM can generate RC designs better than those generated by the existing RC optimization methods, and the KBEM is suitable for various types of fin-and-tube heat exchangers.