作者：
He, ZhaoLiu, HuiCent South Univ
Inst Artificial Intelligence & Robot IAIR Sch Traff & Transportat Engn Key Lab Traff Safety TrackMinist Educ Changsha 410075 Hunan Peoples R China
The key to solving constrained multiobjective optimization problems (CMOPs) lies in maintaining the feasibility, convergence, and diversity of the population. In recent years, various constraint handling techniques (C...
详细信息
The key to solving constrained multiobjective optimization problems (CMOPs) lies in maintaining the feasibility, convergence, and diversity of the population. In recent years, various constraint handling techniques (CHTs) and strategies have been proposed to enhance the performance of constrainedmultiobjective evolutionary algorithms (CMOEAs). However, most of these algorithms face difficulties in dealing with problems that have large infeasible regions and discontinuous small feasible regions, as they have trouble crossing large infeasible regions while simultaneously maintaining the convergence and diversity of the population. To tackle this issue, this paper proposes a dual-population auxiliary coevolutionary algorithm with an enhanced operator, denoted as DAEAEO. Auxiliary population 1 employs an improved epsilon-constraint handling technique to provide high-quality feasible solutions for the main population. Auxiliary population 2 adopts the non-dominated sorting method to provide favorable objective information for the main population to help it cross the infeasible region. In addition, to further improve diversity, each population adopts an enhanced operator and a genetic operator to generate offspring, respectively. Finally, knowledge transfer between offspring is realized. Compared to six state-of-the-art CMOEAs on DASCMOPs, LIR-CMOPs, DOC test suites, and two real-world problems, the proposed DAEAEO achieved superior performance, especially for CMOPs with large infeasible regions and discontinuous small feasible regions.
The use of evolutionary algorithms (EAs) to solve problems with multiple objectives (known as multiobjectiveoptimizationproblems (MOPs)) has attracted much attention recently. Population based approaches,such as EAs...
详细信息
ISBN:
(纸本)0780395387
The use of evolutionary algorithms (EAs) to solve problems with multiple objectives (known as multiobjectiveoptimizationproblems (MOPs)) has attracted much attention recently. Population based approaches,such as EAs, offer a means to rind a group of pareto-optimal solutions in a single run. However, most studies are undertaken on unconstrained MOPs. Recently, we developed the co-evolutionary algorithms for unconstrained *** objective of this paper is to introduce a modification to coevolutionary algorithms for handling constraints. The solutions, provided by the proposed algorithm for one test problem, are promising when compared with an existing well-known algorithm.
In this paper, we propose a spectral projected subgradient method with a 1-memory momentum term for solving constrained convex multiobjectiveoptimization problem. This method combines the subgradient-type algorithm f...
详细信息
In this paper, we propose a spectral projected subgradient method with a 1-memory momentum term for solving constrained convex multiobjectiveoptimization problem. This method combines the subgradient-type algorithm for multiobjectiveoptimizationproblems with the idea of the spectral projected algorithm to accelerate the convergence process. Additionally, a 1-memory momentum term is added to the subgradient direction in the early iterations. The 1-memory momentum term incorporates, in the present iteration, some of the influence of the past iterations, and this can help to improve the search direction. Under suitable assumptions, we show that the sequence generated by the method converges to a weakly Pareto efficient solution and derive the sublinear convergence rates for the proposed method. Finally, computational experiments are given to demonstrate the effectiveness of the proposed method.
In this paper, we propose a multi-stage evolutionary framework with adaptive selection (MSEFAS) for efficiently handling constrained multi-objective optimizationproblems (CMOPs). MSEFAS has two stages of optimization...
详细信息
In this paper, we propose a multi-stage evolutionary framework with adaptive selection (MSEFAS) for efficiently handling constrained multi-objective optimizationproblems (CMOPs). MSEFAS has two stages of optimization in its early phase of evolutionary search: one stage that encourages promising infeasible solutions to approach the feasible region and increases diversity, and the other stage that enables the population to span large infeasible regions and accelerates convergence. To adaptively determine the execution order of these two stages in the early process, MSEFAS treats the optimization stage with higher validity of selected solutions as the first stage and the other as the second one. In addition, at the late phase of evolutionary search, MSEFAS introduces a third stage to efficiently handle the various characteristics of CMOPs by considering the relationship between the constrained Pareto fronts (CPF) and unconstrained Pareto fronts. We compare the proposed framework with eleven state-of-the-art constrained multi-objective evolutionary algorithms on 56 benchmark CMOPs. Our results demonstrate the effectiveness of the proposed framework in handling a wide range of CMOPs, showcasing its potential for solving complex optimizationproblems.
Solving constrained multiobjective optimization problems is one of the most challenging areas in the evolutionary computation research community. To solve a constrainedmultiobjectiveoptimization problem, an algorith...
详细信息
ISBN:
(纸本)9781509006229
Solving constrained multiobjective optimization problems is one of the most challenging areas in the evolutionary computation research community. To solve a constrainedmultiobjectiveoptimization problem, an algorithm should tackle the objective functions and the constraints simultaneously. As a result, many constraint-handling techniques have been proposed. However, most of the existing constraint-handling techniques are developed to solve test instances (e.g., CTPs) with low dimension and large feasible region. On the other hand, experimental comparisons on different constraint-handling techniques remain scarce. In view of these two issues, in this paper we first construct 18 test instances, each of which exhibits different properties. Afterward, we choose three representative constraint-handling techniques and combine them with nondominated sorting genetic algorithm 11 to study the performance difference on various conditions. By the experimental studies, we point out the advantages and disadvantages of different constraint-handling techniques.
For constrained multiobjective optimization problems (CMOPs), one of the fundamental issues faced by researchers is how to make comparisons between individuals within the population that results in a balanced selectio...
详细信息
For constrained multiobjective optimization problems (CMOPs), one of the fundamental issues faced by researchers is how to make comparisons between individuals within the population that results in a balanced selection of better individuals. The selection of better individuals need to be balanced between elitism, diversity and feasibility. In this paper, we propose a hybrid constraint handling technique of population trimming strategy and adaptive penalty function for multiobjective evolutionary algorithm NSGA-II to solve CMOPs. In our approach, a method of objectivization of constraint violations and proportional reduction is used to compare two individuals and trim the population, and as a result the new parent population consisted of the optimal feasible individuals and good infeasible individuals is obtained. To our knowledge, the distant matrix in proportional reduction procedure is firstly proposed for comparison and maintaining diversity in infeasible individuals. Furthermore, an adaptive penalty function method is utilized to give the fitness of individuals in parent population. Numerical simulations indicate that the proposed algorithm outperforms the current state-of-the-art algorithms, e.g. NSGA-II-CD,NAGA-II-WTY, in both convergence and diversity.
暂无评论