Aimed at improving the insufficient search ability of constraint differential evolution with single constraint handling technique when solving complex optimization problem, this paper proposes a constraint differentia...Aimed at improving the insufficient search ability of constraint differential evolution with single constraint handling technique when solving complex optimization problem, this paper proposes a constraint differential evolution algorithm?based on ensemble of constraint handling techniques and multi-population?framework, called ECMPDE. First, handling three improved variants of differential evolution algorithms are dynamically matched with two constraint handling techniques through the constraint allocation mechanism. Each combination includes three variants with corresponding constraint handling technique?and these combinations are in the set. Second, the population is divided into three smaller subpopulations and one larger reward subpopulation. Then a combination with three constraint algorithms is randomly selected from the set, and the three constraint algorithms are run in three sub-populations respectively. According to the improvement of fitness value, the optimal constraint?algorithm is selected to run on the reward sub-population, which can share?information and close cooperation among populations. In order to verify the effectiveness of the proposed algorithm, 12 standard constraint optimization problems?and 10 engineering constraint optimization problems are tested. The experimental results show that ECMPDE is an effective algorithm for solving constraint optimization problems.展开更多
Robust design (RD) has received much attention from researchers and practitioners for years, and a number of methodologies have been studied in the research community. The majority of existing RD models focus on the m...Robust design (RD) has received much attention from researchers and practitioners for years, and a number of methodologies have been studied in the research community. The majority of existing RD models focus on the minimum variability with a zero bias. However, it is often the case that the customer may specify upper bounds on one of the two process parameters (i.e., the process mean and variance). In this situation, the existing RD models may not work efficiently in incorporating the customer’s needs. To this end, we propose two simple RD models using the ε?constraint feasible region method - one with an upper bound of process bias specified and the other with an upper bound on process variability specified. We then conduct a case study to analyze the effects of upper bounds on each of the process parameters in terms of optimal operating conditions and mean squared error.展开更多
Over the years, a number of methods have been proposed for the generation of uniform and globally optimal Pareto frontiers in multi-objective optimization problems. This has been the case irrespective of the problem d...Over the years, a number of methods have been proposed for the generation of uniform and globally optimal Pareto frontiers in multi-objective optimization problems. This has been the case irrespective of the problem definition. The most commonly applied methods are the normal constraint method and the normal boundary intersection method. The former suffers from the deficiency of an uneven Pareto set distribution in the case of vertical （or horizontal） sections in the Pareto frontier, whereas the latter suffers from a sparsely populated Pareto frontier when the optimization problem is numerically demanding （ill-conditioned）. The method proposed in this paper, coupled with a simple Pareto filter, addresses these two deficiencies to generate a uniform, globally optimal, well-populated Pareto frontier for any feasible bi-objective optimization problem. A number of examples are provided to demonstrate the performance of the algorithm.展开更多
文摘Aimed at improving the insufficient search ability of constraint differential evolution with single constraint handling technique when solving complex optimization problem, this paper proposes a constraint differential evolution algorithm?based on ensemble of constraint handling techniques and multi-population?framework, called ECMPDE. First, handling three improved variants of differential evolution algorithms are dynamically matched with two constraint handling techniques through the constraint allocation mechanism. Each combination includes three variants with corresponding constraint handling technique?and these combinations are in the set. Second, the population is divided into three smaller subpopulations and one larger reward subpopulation. Then a combination with three constraint algorithms is randomly selected from the set, and the three constraint algorithms are run in three sub-populations respectively. According to the improvement of fitness value, the optimal constraint?algorithm is selected to run on the reward sub-population, which can share?information and close cooperation among populations. In order to verify the effectiveness of the proposed algorithm, 12 standard constraint optimization problems?and 10 engineering constraint optimization problems are tested. The experimental results show that ECMPDE is an effective algorithm for solving constraint optimization problems.
基金This work was supported partly by the 2005 Inje University research grant.
文摘Robust design (RD) has received much attention from researchers and practitioners for years, and a number of methodologies have been studied in the research community. The majority of existing RD models focus on the minimum variability with a zero bias. However, it is often the case that the customer may specify upper bounds on one of the two process parameters (i.e., the process mean and variance). In this situation, the existing RD models may not work efficiently in incorporating the customer’s needs. To this end, we propose two simple RD models using the ε?constraint feasible region method - one with an upper bound of process bias specified and the other with an upper bound on process variability specified. We then conduct a case study to analyze the effects of upper bounds on each of the process parameters in terms of optimal operating conditions and mean squared error.
文摘Over the years, a number of methods have been proposed for the generation of uniform and globally optimal Pareto frontiers in multi-objective optimization problems. This has been the case irrespective of the problem definition. The most commonly applied methods are the normal constraint method and the normal boundary intersection method. The former suffers from the deficiency of an uneven Pareto set distribution in the case of vertical （or horizontal） sections in the Pareto frontier, whereas the latter suffers from a sparsely populated Pareto frontier when the optimization problem is numerically demanding （ill-conditioned）. The method proposed in this paper, coupled with a simple Pareto filter, addresses these two deficiencies to generate a uniform, globally optimal, well-populated Pareto frontier for any feasible bi-objective optimization problem. A number of examples are provided to demonstrate the performance of the algorithm.
