Genetic Algorithm Control Parameters Research Articles

Comparative evaluation of the performance of improved genetic algorithms and differential evolution for in-core fuel management of a research reactor

This paper presents a comparative evaluation of the performance of Genetic Algorithm (GA) and Differential Evolution (DE) algorithm applied to in-core fuel management of the DNRR research reactor. Two GA variants corresponding to two selection operators, i.e., tournament (GA1) and roulette wheel (GA2) selections, respectively, with two-point crossover and scramble mutation were implemented for the ICFM problem. A comprehensive survey of the GA control parameters such as population size, crossover-type, mutation probability, and elitist archive size has been conducted to optimize the performance of the GAs. The basic DE was implemented with a standard mutation strategy DE/rand/1/bin. Numerical computations were performed based on the DNRR research reactor core loaded with 100 highly enriched uranium fuel (HEU) bundles for evaluating the performance of the GA and DE algorithms. Two main objectives were included in the fitness function to maximize the fuel cycle length and flatten the power distribution. The performance of the two GA variants and the basic DE was investigated with the same population size, fitness function, and convergence criterion. Each method was performed with 50 independent runs, and the best fitness values were collected for statistical analysis using Kruskal–Wallis and Mann–Whitney tests in comparison among the three methods. The statistical analysis shows that the performance of GA1 with tournament selection and DE are not significantly different and are better than GA2 with roulette wheel selection. DE is stable and efficient in exploring the search space to approach the global optimal solution in most runs. While, GA1 and GA2 were trapped at local optima by about 26% and 38%, respectively. However, the best solutions obtained with GA1 and GA2 after 50 independent runs are better than that obtained with DE in term of fitness values. This suggests an improvement of the basic DE is needed to maintain the potential good solutions during the evolution process.

Effects of Particle Swarm Optimization and Genetic Algorithm Control Parameters on Overcurrent Relay Selectivity and Speed

Distribution systems continue to grow and becoming more complex with increasing operational challenges such as protection miscoordination. Initially, conventional methods were favoured to solve overcurrent relay coordination problems; however, the implementation of these methods is time-consuming. Therefore, recent studies have adopted the utilisation of particle swarm optimization and genetic algorithms to solve overcurrent relay coordination problems and maximise system selectivity and operational speed. Particle swarm optimization and genetic algorithms are evolutionary algorithms that at times suffer from premature convergence due to poor selection of control parameters. Consequently, this paper presents a comprehensive sensitivity analysis to evaluate the effect of the discrete control parameters on particle swarm optimizer and genetic algorithms performance, alternatively on the behaviour of overcurrent relays. Optimization algorithms aim to minimise overcurrent relay time multiplier settings and accomplish optimal protection coordination. The findings indicate that particle swarm optimization is more sensitive to inertia weight and swarm size while the number of iterations has minimal effect. The results also depict that 30% crossover, 2% mutation, and smaller population size yield faster convergence rate and optimise fitness function, which improves genetic algorithms performance. Sensitivity analysis results are verified by comparing the performance of particle swarm optimization with the genetic algorithms which show the former parameter setting outperforms the latter. The relay operational speed is reduced by 15% for particle swarm optimization and system selectivity is maximised. The optimal protection coordination achieved using particle swarm optimization showed superiority of the algorithm, its ability to circumvent premature convergence, consistency, and efficiency.

Research on the optimisation of complex models of large-scale building structures dependent on adaptive grey genetic algorithms

Genetic algorithm (GA) is a bionics algorithm based on the biological evolution theory that has received extensive attention in the field of computer science and optimisation in recent years. This paper analyses and integrates the relevant contents of genetic algorithm and its application in the optimal design of large-scale building structures and analyses and researches briefly several key factors when the genetic algorithm is applied to the optimal design of large-scale building structures, such as mathematical modelling, constraint condition treatment, generation of initial population and selection of control parameters of genetic algorithm. However, because the simple genetic algorithm is only good at global search, and the local search ability is not enough, it will take quite a long time to achieve the real optimal solution. For the shortcomings of simple genetic algorithm, an improved adaptive grey genetic algorithm is proposed in this paper. The example shows that the obtained adaptive genetic algorithm can improve the convergence and calculation speed when the genetic algorithms is applied to structural optimisation design.

Research on the optimisation of complex models of large-scale building structures dependent on adaptive grey genetic algorithms

Genetic algorithm (GA) is a bionics algorithm based on the biological evolution theory that has received extensive attention in the field of computer science and optimisation in recent years. This paper analyses and integrates the relevant contents of genetic algorithm and its application in the optimal design of large-scale building structures and analyses and researches briefly several key factors when the genetic algorithm is applied to the optimal design of large-scale building structures, such as mathematical modelling, constraint condition treatment, generation of initial population and selection of control parameters of genetic algorithm. However, because the simple genetic algorithm is only good at global search, and the local search ability is not enough, it will take quite a long time to achieve the real optimal solution. For the shortcomings of simple genetic algorithm, an improved adaptive grey genetic algorithm is proposed in this paper. The example shows that the obtained adaptive genetic algorithm can improve the convergence and calculation speed when the genetic algorithms is applied to structural optimisation design.

Parameter Tuning Method for Genetic Algorithm using Taguchi Orthogonal Array for Non-linear Multimodal Optimization Problem

Genetic algorithm (GA) is the most widely used meta-heuristic optimization algorithm that can solve complex large scale optimization problems successfully. The only problem lies with the setting of GA control parameters and their levels for the optimal performance of the algorithm. The statistical method can be used for parameter tuning that allows us to collect data properly. Also, it analyzes the collected data accurately and presents the appropriate results. For statistical analysis, Taguchi’s robust factorial design which is highly fractional in nature with a special set of L32 orthogonal array (OA) is used. Taguchi’s robust design is a proficient method to find an optimum solution with a minimal number of designs of experiment (DOE). Analysis of variance (ANOVA) is conducted to check whether GA control parameters are statistically significant or not for non-linear multimodal optimization problems. Taguchi OA design and ANOVA analysis experimental study is conducted using Stat-Ease software and for the Genetic Algorithm control parameter setting GA solver of MATLAB is used.

Software vendors travel management decisions using an elitist nonhomogeneous genetic algorithm

Vendors supplying software need to travel various places for product installation, troubleshooting, maintenance, upgrading, and promoting new products. These problems can be classified under asymmetric traveling salesman problems. Also, they are NP-hard in nature. In this paper, a new class of asymmetric traveling salesman problems is studied in which travel times are stochastic. We consider travel times as double truncated lognormal random variable and obtain the deterministic equivalent of the stochastic problem. An elitist nonhomogeneous genetic algorithm is proposed to solve this problem. The genetic algorithm control parameters, crossover and mutation probabilities vary in each successive generation. The best chromosome having the minimum travel time, together with the optimum crossover and mutation probabilities are selected using an elitist strategy. It is proved that the proposed algorithm follows a nonhomogeneous Markov chain process and almost certainly converges to the population that contains the optimal solution. Two case examples are presented to demonstrate the usefulness of the proposed approach.

Using a Double Genetic Algorithm for Correlating Thermal Models

Correlating a thermal model could be a challenging problem in global optimization, which often requires employing a genetic algorithm. This work studies choosing the optimal set of the genetic algorithm control parameters when it is used for this purpose. Two practical thermal models are considered. For the small model the optimal set of the genetic algorithm parameters was found by applying a double genetic algorithm where the outer algorithm tried to minimize the best fitness value achieved by the inner one by varying the inner algorithm control parameters. In addition, it was demonstrated that in a practical calculation increasing the population size can easily result in a worse accuracy. The proper choice of the population size was studied for the large model. The double genetic algorithm was modified by starting each run of the inner algorithm with the best solution obtained until then, instead of using the same initial population in all the runs. Doing all the runs with the same population size and choosing the population size randomly were also tried and it was shown that this way a considerable improvement in accuracy could be achieved.

Genetic Algorithm Parameter Optimization using Taguchi Robust Design for Multi-response Optimization of Experimental and Historical Data

This paper presents a methodology for robust optimization of Genetic Algorithm (GA) involving complex interactions among the control parameters. Finding the Optimum GA parameters to solve an optimization problem for producing best results with least variability is still an open area of research. The proposed research approach primarily covers the robust optimization of Genetic Algorithm control parameters using Taguchi Design of Experiment (DOE) with a special set of L25 orthogonal array (OA). The experimental design and the study is conducted with MATLAB Genetic Algorithm internal control parameters using real-coded Genetic Algorithm fitness functions operates directly on real values of two different case studies. One of them is based on experimental data for the development of a new product and the other one is based on the historical data of an existing product under large scale complex manufacturing system. The performance characteristics (Best fitness value) of GA are evaluated. The main effect for signal-tonoise ratios (SNR), main effect for means, response tables and analysis of variance (ANOVA) are employed and optimum parameter settings are obtained. The confirmation experiments are carried out with the optimum setting parameters and the results are compared with the predicted & default setting values of GA. The experimental results show an increase of signal-to-noise ratio (SNR) by around 54 % and mean by 12 to 25 % from the default setting of GA to the optimum settings arrived during the experimental process, which are extremely significant. Further, cross over was found to be the most influential parameter for both the case studies, followed by other parameters like population size, selection process for case study-1 and mutation rate, population size for case study-2. It is evident that the GA performances are also sensitive to the objective function (transfer function in GA terms). The study clearly shows the robust optimization of GA parameters leading to best level of performance characteristics with least variability for the solution of real life optimization problem, using Taguchi Experimental Design Techniques.

Improved Genetic Algorithm with Two-Level Approximation Method for Laminate Stacking Sequence Optimization by Considering Engineering Requirements

Laminated composites have been widely applied in aerospace structures; thus optimization of the corresponding stacking sequences is indispensable. Genetic algorithms have been popularly adopted to cope with the design of stacking sequences which is a combinatorial optimization problem with complicated manufacturing constraints, but they often exhibit high computational costs with many structural analyses. A genetic algorithm using a two-level approximation (GATLA) method was proposed previously by the authors to obtain the optimal stacking sequences, which requires significantly low computational costs. By considering practical engineering requirements, this method possesses low applicability in complicated structures with multiple laminates. What is more, it has relatively high dependence on some genetic algorithm control parameters. To address these problems, now we propose an improved GA with two-level approximation (IGATLA) method which includes improved random initial design, adaptive penalty fitness function, adaptive crossover probability, and variable mutation probability, as well as enhanced validity check criterion for multiple laminates. The efficiency and feasibility of these improvements are verified with numerical applications, including typical numerical examples and industrial applications. It is shown that this method is also able to handle large, real world, industrial analysis models with high efficiency.

Determining Relative Importance and Effective Settings for Genetic Algorithm Control Parameters.

Setting the control parameters of a genetic algorithm to obtain good results is a long-standing problem. We define an experiment design and analysis method to determine relative importance and effective settings for control parameters of any evolutionary algorithm, and we apply this method to a classic binary-encoded genetic algorithm (GA). Subsequently, as reported elsewhere, we applied the GA, with the control parameter settings determined here, to steer a population of cloud-computing simulators toward behaviors that reveal degraded performance and system collapse. GA-steered simulators could serve as a design tool, empowering system engineers to identify and mitigate low-probability, costly failure scenarios. In the existing GA literature, we uncovered conflicting opinions and evidence regarding key GA control parameters and effective settings to adopt. Consequently, we designed and executed an experiment to determine relative importance and effective settings for seven GA control parameters, when applied across a set of numerical optimization problems drawn from the literature. This paper describes our experiment design, analysis, and results. We found that crossover most significantly influenced GA success, followed by mutation rate and population size and then by rerandomization point and elite selection. Selection method and the precision used within the chromosome to represent numerical values had least influence. Our findings are robust over 60 numerical optimization problems.

Optimization Research of Process Parameters for Laser Cladding Valve Parts Based on Genetic Algorithm

The optimization research of process parameters for big power Laser cladding valve parts, is a research focus of modern surface hardening technology. The article discussed in detail for solving the optimum process parameters of Laser cladding for the selection approach of strategy of genetic algorithm, the quantitative relationship model was established between process parameters and the valve parts property using neural network method , which process parameters are laser power (P), scanning speed (V), powder feeding rate (G), scan spacing (D) and thickness ( ) etc., the best configuration program of Genetic Algorithm control parameters has been obtain by means of the parameters encoding、initial group setting、fitness function design,genetic operation design and algorithm control parameters setting. The optimization of process parameters is obtained to fit the Laser cladding technology by using genetic algorithm toolbox in the MATLAB environment, and the optimization goal of the valve parts property has also been achieved. Practice has proved that the optimal process parameters are correct by the genetic algorithm , and has a very good production practice guide.

Genetic algorithm convergence analysis using a unified representation of genes and the hamming distance

This paper first introduces a unified representation of both real-coded genes and binary-coded genes, in order to efficiently analyse of a genetic algorithm. In literature, syntagm the entire population premature convergence is used with meaning of closing evolution before reaching optimal point. It can be emphasised only on a test function with known landscape. If function landscape is unknown, one can only notice population convergence. This paper aims to answer to question: how can we influence control parameters of genetic algorithm so that exploration period of parameter space be longer and risk of premature be reduced? The answer to above question implies selection of a crossover operator with good performance in landscape exploration and use of two indicators for detection of population convergence. In order to choose appropriate control parameters of genetic algorithm, fitness function landscape must be taken into consideration.

Integrated GA–Fuzzy Multi-Objective Model for Centralized Volt/Var Control in Distribution Systems

This article presents a fuzzy multi-objective optimization model (FMOOM) for Volt/Var control within the frame of distribution management system (DMS) functions. Optimal control actions are of centralized character for the whole distribution system, based on historical operation database and available decentralized set of real-time measurements. Within the set of optimization goals/constraints, real power losses, overall voltage quality index and power factor are considered, with the possibility for the inclusion of other factors, without losing features of the FMOOM and calculation speed. To achieve an optimal defuzzified Volt/Var control solution with flexible input data, the robust technique of genetic algorithm (GA), is used. The application possibility of GA for the solution of real-time problems (like Volt/Var control) is explored. The aim is to give clear recommendations to decision-makers in view of setting GA control parameters that should ensure good quality of the near global optimal solution. The effectiveness of the proposed integrated GA–FMOOM for Volt/Var control, applied to various operation conditions, is verified on the example of a real radial distribution system from the Electric Power Industry of Serbia.

Process synthesis of laser forming by genetic algorithm

Significant progress has been made in analyzing and predicting laser forming (LF) processes of sheet metal. Process synthesis in laser forming, on the other hand, is concerned with determination of laser scanning paths and heat condition given a target shape to form. This paper reports the development of a process synthesis methodology for laser forming of a class of shapes based on genetic algorithm (GA). The effects of GA control parameters on the synthesis process are discussed. The effects of fitness function types on achieving multiple objectives are also investigated. The synthesis process is experimentally validated through several cases under diverse conditions including one that involves close to thirty decision variables.

Solving the unit commitment problem with a genetic algorithm through a constraint satisfaction technique

This paper proposes a genetic algorithm (GA) in conjunction with constraint handling techniques to solve the thermal unit commitment problem. To deal effectively with the constraints of the problem and prune the search space of the GA in advance, the difficult minimum up- and down-time constraints are embedded in the binary strings that are coded to represent the on-off states of the generating units. The other constraints are handled by integrating penalty factors into the cost function within an enhanced economic dispatch program. The proposed GA approach has been tested on a practical Taiwan Power (Taipower) thermal system over a 24-hour period for different utility factors and GA control parameters. Test results reveal that the features of easy implementation, fast convergence, and a highly near-optimal solution in solving the UC problem can be achieved by the proposed GA approach.