Three-objective Problems Research Articles

An Analysis of Quality Indicators Using Approximated Optimal Distributions in a 3-D Objective Space

Although quality indicators play a crucial role in benchmarking evolutionary multi-objective optimization algorithms, their properties are still unclear. One promising approach for understanding quality indicators is the use of the optimal distribution of objective vectors that optimizes each quality indicator. However, it is difficult to obtain the optimal distribution for each quality indicator, especially when its theoretical property is unknown. Thus, optimal distributions for most quality indicators have not been well investigated. To address these issues, first, we propose a problem formulation of finding the optimal distribution for each quality indicator on an arbitrary Pareto front. Then, we approximate the optimal distributions for nine quality indicators using the proposed problem formulation. We analyze the nine quality indicators using their approximated optimal distributions on eight types of Pareto fronts of three-objective problems. Our analysis demonstrates that uniformly-distributed objective vectors over the entire Pareto front are not optimal in many cases. Each quality indicator has its own optimal distribution for each Pareto front. We also examine the consistency among the nine quality indicators.

Explicit Control of Implicit Parallelism in Decomposition-Based Evolutionary Many-Objective Optimization Algorithms [Research Frontier

Over the past two decades, Evolutionary Multi-objective Optimization (EMO) algorithms have demonstrated their ability to find and maintain multiple trade-off solutions in two and three-objective problems, making EMO as one of the most emergent and exciting fields of research and application within Computational Intelligence (CI) area. The main reason for EMO's success is that the population-based EMO operators are able to establish an implicit parallel search within an evolving population to find multiple Pareto-optimal regions of the search space parallelly. For many-objective optimization problems involving a largedimensional objective space, the extent of implicit parallelism is argued here to be too generic, compared to the same in a lower-dimensional objective space. Decomposition-based EMO algorithms - a recent trend in EMO literature - which divide the overall computing task into a number of sub-tasks of focusing within a region of the search space have found to be successful in solving many-objective problems. In this paper, we study the effect of explicit control of an algorithm's implicit parallelism mechanism for achieving an enhanced performance of decomposition-based EMO algorithms. We consider three decomposition-based many-objective evolutionary algorithms (EAs) - MOEA/D, MOEA/D-M2M, and NSGA-III - for this purpose. We also investigate another explicit control strategy of suitably choosing a normalization method of objectives for improving the performance of MOEA/D and MOEA/ D-M2M methods, and report much improved performance than their original counterparts. The principles of this study are valid for any population-based search and optimization algorithms and can be extended to improve the performance other single-objective EA, EMO, and other relevant CI methods.

Efficient computation of expected hypervolume improvement using box decomposition algorithms

In the field of multi-objective optimization algorithms, multi-objective Bayesian Global Optimization (MOBGO) is an important branch, in addition to evolutionary multi-objective optimization algorithms. MOBGO utilizes Gaussian Process models learned from previous objective function evaluations to decide the next evaluation site by maximizing or minimizing an infill criterion. A commonly used criterion in MOBGO is the Expected Hypervolume Improvement (EHVI), which shows a good performance on a wide range of problems, with respect to exploration and exploitation. However, so far, it has been a challenge to calculate exact EHVI values efficiently. This paper proposes an efficient algorithm for the exact calculation of the EHVI for in a generic case. This efficient algorithm is based on partitioning the integration volume into a set of axis-parallel slices. Theoretically, the upper bound time complexities can be improved from previously O (n^2) and O(n^3), for two- and three-objective problems respectively, to varTheta (nlog n), which is asymptotically optimal. This article generalizes the scheme in higher dimensional cases by utilizing a new hyperbox decomposition technique, which is proposed by Dächert et al. (Eur J Oper Res 260(3):841–855, 2017). It also utilizes a generalization of the multilayered integration scheme that scales linearly in the number of hyperboxes of the decomposition. The speed comparison shows that the proposed algorithm in this paper significantly reduces computation time. Finally, this decomposition technique is applied in the calculation of the Probability of Improvement (PoI).

Two efficient algorithms for constructing almost even approximations of the Pareto front in multi-objective optimization problems

ABSTRACTIn this article, two algorithms are proposed for constructing almost even approximations of the Pareto front of multi-objective optimization problems. The first algorithm is a hybrid of the ε-constraint and Pascoletti–Serafini scalarization methods for solving bi-objective problems. The second is a modification of the successive Pareto optimization (SPO) algorithm for solving three-objective problems. In these algorithms, the MATLAB fmincon solver is used to solve single-objective optimization problems, which returns a local optimal solution. Some metrics are considered to evaluate the quality of approximations obtained by the suggested algorithms on six test problems, and their results are compared with other algorithms (normal constraint, weighted constraint, SPO, differential evolution, multi-objective evolutionary algorithm/decomposition–differential evolution, non-dominated sorting genetic algorithm-II and S-metric selection evolutionary multi-objective algorithm). Experimental results show that the proposed algorithms provide almost even approximations of the whole Pareto front, and better quality of approximation and CPU time compared with established algorithms.

A Novel Sequential Multi-Objective Optimization Using Anchor Points in the Design Space of Global Variables

Multi-objective problems are encountered in many engineering applications and multi-objective optimization (MOO) approaches have been proposed to search for Pareto solutions. Due to the nature of searching for multiple optimal solutions, the computational efforts of MOO can be a serious concern. To improve the computational efficiency, a novel efficient sequential MOO (S-MOO) approach is proposed in this work, in which anchor points in the design space for global variables are fully utilized and a data set for global solutions is generated to guide the search for Pareto solutions. Global variables refer to those shared by more than one objective or constraint, while local variables appear only in one objective and corresponding constraints. As a matter of fact, it is the existence of global variables that leads to couplings among the multiple objectives. The proposed S-MOO breaks the couplings among multiple objectives (and constraints) by distinguishing the global variables, and thus all objectives are optimized in a sequential manner within each iteration while all iterations can be processed in parallel. The computational cost per produced Pareto point is reduced and a well-spread Pareto front is obtained. Six numerical and engineering examples including two three-objective problems are tested to demonstrate the applicability and efficiency of the proposed approach.

Graphical exploration of the weight space in three-objective mixed integer linear programs

In this paper we address the computation of indifference regions in the weight space for multiobjective integer and mixed-integer linear programming problems and the graphical exploration of this type of information for three-objective problems. We present a procedure to compute a subset of the indifference region associated with a supported nondominated solution obtained by the weighted-sum scalarization. Based on the properties of these regions and their graphical representation for problems with up to three objective functions, we propose an algorithm to compute all extreme supported nondominated solutions adjacent to a given solution and another one to compute all extreme supported nondominated solutions to a three-objective problem. The latter is suitable to characterize solutions in delimited nondominated areas or to be used as a final exploration phase. A computer implementation is also presented.

Corrections on the Box Plots of the Coverage Metric in “Multiobjective Immune Algorithm with Nondominated Neighbor-based Selection”

In this study, we analyze the experimental results in Gong et al. (2008), and find that the box plots of the coverage metric in solving the five three-objective problems are wrong. The corresponding corrected box plots are presented. The corrected results are still consistent with the original conclusions.

Multi-objective optimization of an industrial fluidized-bed catalytic cracking unit (FCCU) using two jumping gene adaptations of simulated annealing

Multi-objective optimization of industrial chemical engineering operations using simulated annealing (SA), genetic algorithm (GA) and their adaptations often requires large amounts of computational (CPU) time. Any adaptation to speed up the solution procedure is, thus, desirable. The performance of the evolutionary technique, multi-objective simulated annealing (MOSA), is found to improve by the use of two adaptations, MOSA-JG and MOSA-aJG, inspired by the concept of jumping genes (JG) in biology. This is because of the increase in diversity. These adaptations are similar to those used with NSGA-II [Kasat, R. B., & Gupta, S. K. (2003). Multiobjective optimization of an industrial fluidized-bed catalytic cracking unit (FCCU) using genetic algorithm with the jumping genes operator. Computers & Chemical Engineering, 27, 1785]. Optimization of three benchmark problems and two computationally intensive, real-life, two- and three-objective problems [Kasat, R. B., Kunzru, D., Saraf, D. N., & Gupta, S. K. (2002). Multiobjective optimization of industrial FCC units using elitist nondominated sorting genetic algorithm. Industrial & Engineering Chemistry Research, 41, 4765] for an industrial FCCU, are studied. The aJG adaptation of MOSA leads to non-dominated optimal solutions having a good spacing and reasonable maximum spread of the non-dominated solutions. These adaptations can prove to be of considerable value for solving other computationally intensive problems in chemical engineering.

角型構造の大規模多目的ファジィ線形計画問題に対する分解手法の有効性

In this paper, we focus on large-scale multiobjective fuzzy linear programming problems with the block angular structure and examine the efficiency of the Dantzig-Wolfe decomposition method in the interactive fuzzy satisficing method proposed by M. Sakawa et al. After overviewing the Dantzig-Wolfe decomposition method (1961) and the interactive fuzzy satisficing method, three-objective linear programming problems with 15 coupling constraints are considered in order to show the efficiency of the Dantzig-Wolfe decomposition method over the revised simplex method. Through a lot of computational experiments on workstation for numerical examples with both 50 and 200 variables, the advantages of the Dantzig-Wolfe decomposition method are discussed with respect to processing time and required memory storage.

THE NONINFERIOR SET ESTIMATION (NISE) METHOD FOR THREE OBJECTIVE PROBLEMS

The noninferior set estimation (NISE) method is a powerful technique for the generation of the Pareto optimal set for multicriteria optimization problems with two dimensions. This paper extends the NISE method to three-objective problems. The concepts which allow this extension are presented and an algorithm described. The method is applied to a problem drawn from the literature to produce numerical results.