Sujin Bureerat

Population-Based Incremental Learning for Multiobjective Optimisation

The work in this paper presents the use of population-based incremental learning (PBIL), one of the classic single-objective population-based optimisation methods, as a tool for multiobjective optimisation. The PBIL method with two different updating schemes of its probability vectors is presented. The performance of the two proposed multiobjective optimisers are measured and compared with four other established multiobjective evolutionary algorithms i.e. niched Pareto genetic algorithm, version 2 of non-dominated sorting genetic algorithm, version 2 of strength Pareto evolutionary algorithm, and Pareto archived evolution strategy. The optimisation methods are implemented to solve 8 bi-objective test problems where design variables are encoded as a binary string. The Pareto optimal solutions obtained from the various methods are compared and discussed. It can be concluded that, with the assigned test problems, the multiobjective PBIL methods are comparable to the previously developed algorithms in terms of convergence rate. The clear advantage in using PBILs is that they can provide considerably better population diversity.

Multi-objective topology optimization using evolutionary algorithms

This article deals with the comparative performance of some established multi-objective evolutionary algorithms (MOEAs) for structural topology optimization. Four multi-objective problems, having design objectives like structural compliance, natural frequency and mass, and subjected to constraints on stress, etc., are posed for performance testing. The MOEAs include Pareto archive evolution strategy (PAES), population-based incremental learning (PBIL), non-dominated sorting genetic algorithm (NSGA), strength Pareto evolutionary algorithm (SPEA), and multi-objective particle swarm optimization (MPSO). The various MOEAs are implemented to solve the problems. The ground element filtering (GEF) technique is used to suppress checkerboard patterns on topologies. The results obtained from the various optimizers are illustrated and compared. It is shown that PBIL is far superior to the others. The optimal topologies from using PBIL can be compared with those obtained by employing the classical gradient-based approach. It can be considered as a powerful tool for structural topological design.

Comparative performance of meta-heuristic algorithms for mass minimisation of trusses with dynamic constraints

This paper investigates the search performances of various meta-heuristics (MHs) for solving truss mass minimisation with dynamic constraints. Several established MHs were used to solve five truss optimisation problems. The results obtained from using the various MHs were statistically compared based upon convergence rate and consistency. It was found that the best optimisers for this design task are evolution strategy with covariance matrix adaptation (CMAES) and differential evolution (DE). Furthermore, the best penalty function technique was discovered while four penalty function techniques assigned with several parameter settings were used in combination with the five best optimisers to solve the truss optimisation problems.

Geometrical Design of Plate-Fin Heat Sinks Using Hybridization of MOEA and RSM

The work in this paper is aimed at demonstrating the practical multiobjective optimization of plate-fin heat sinks and the superiority of using a combined response surface method and multiobjective evolutionary optimizer over solely using the evolutionary optimizer. The design problem assigned is to minimize a heat sink junction temperature and fan pumping power. Design variables determine a heat sink geometry and inlet air velocity. Design constraints are given in such a way that the maximum and minimum fin heights are properly limited. Function evaluation is carried out by using finite volume analysis software. Two multiobjective evolutionary optimization strategies, real-code strength Pareto evolutionary algorithm with and without the use of a response surface technique, are implemented to explore the Pareto optimal front. The optimum results obtained from both design approaches are compared and discussed. It is illustrated that the multiobjective evolutionary technique is a powerful tool for the multiobjective design of electronic air-cooled heat sinks. With the same design conditions and an equal number of function evaluations, the multiobjective optimizer in association with the response surface technique totally outperforms the other. The design parameters affecting the diversity of the Pareto front include fin thickness, fin height distribution, and inlet air velocity while the plate base thickness and the total number of fins of the non-dominated solutions tend to approach certain values.

Hybridisation of real-code population-based incremental learning and differential evolution for multiobjective design of trusses

This paper proposes a hybrid evolutionary algorithm for multiobjective optimisation of trusses using real-code population-based incremental learning (RPBIL) to solve multiobjective design problems. Differential evolution (DE) operators are integrated into the main procedure of RPBIL leading to a hybrid algorithm. The newly developed optimiser, along with some established multiobjective evolutionary algorithms (MOEAs) is implemented to solve a number of multiobjective design problems of trusses. Comparative performance based upon a hypervolume indicator shows that the new hybrid multiobjective evolutionary algorithm is superior to the other MOEAs particularly in cases involving large-scale truss design problems.

Hybrid real-code population-based incremental learning and differential evolution for many-objective optimisation of an automotive floor-frame

In this paper, a many-objective hybrid real-code population-based incremental learning and differential evolution algorithm (MnRPBILDE) is proposed based on the concept of objective function space reduction. The method is then implemented on real engineering design problems. The topology, shape and sizing design of a simplified automotive floor-frame structure are formulated and used as test problems. A variety of well-established multi-objective evolutionary algorithms (MOEAs) including the original version of MnRPBILDE are employed to solve the test problems while the results are compared based on hypervolume and C indicators. The results indicate that our proposed algorithm outperforms the other MOEAs. The proposed algorithm is effective and efficient for many-objective optimisations of a car floor-frame structure.

Hybrid population-based incremental learning using real codes

This paper proposes a hybrid evolutionary algorithm (EA) dealing with population-based incremental learning (PBIL) and some efficient local search strategies. A simple PBIL using real codes is developed. The evolutionary direction and approximate gradient operators are integrated to the main procedure of PBIL. The method is proposed for single objective global optimization. The search performance of the developed hybrid algorithm for box-constrained optimization is compared with a number of well-established and newly developed evolutionary algorithms and meta-heuristics. It is found that, with the given optimization settings, the proposed hybrid optimizer outperforms the other EAs. The new derivative-free algorithm can maintain outstanding abilities of EAs.

Improved Population-Based Incremental Learning in Continuous Spaces

Population-based incremental learning (PBIL) is one of the well-established evolutionary algorithms (EAs). This method, although having outstanding search performance, has been somewhat overlooked compared to other popular EAs. Since the first version of PBIL, which is based on binary search space, several real code versions of PBIL have been introduced; nevertheless, they have been less popular than their binary code counterpart. In this paper, a population-based incremental learning algorithm dealing with real design variables is proposed. The method achieves optimization search with the use of a probability matrix, which is an extension of the probability vector used in binary PBIL. Three variants of the new real code PBIL are proposed while a comparative performance is conducted. The benchmark results show that the present PBIL algorithm outperforms both its binary versions and the previously developed continuous PBIL. The new methods are also compared with well-established and newly developed EAs and it is shown that the proposed real-code PBIL can rank among the high performance EAs.

Optimum plate-fin heat sinks by using a multi-objective evolutionary algorithm

This article demonstrates the practical applications of a multi-objective evolutionary algorithm (MOEA) namely population-based incremental learning (PBIL) for an automated shape optimization of plate-fin heat sinks. The computational procedure of multi-objective PBIL is detailed. The design problem is posed to find heat sink shapes which minimize the junction temperature and fan pumping power while meeting predefined constraints. Three sets of shape design variables used in this study are defined as: vertical straight fins with fin height variation, oblique straight fins with steady fin heights, and oblique straight fins with fin height variation. The optimum results obtained from using the various sets of design variables are illustrated and compared. It can be said that, with this sophisticated design system, efficient and effective design of plate-fin heat sinks is achievable and the best design variables set is the oblique straight fins with fin height variation.

Simultaneous topology and sizing optimization of a water distribution network using a hybrid multiobjective evolutionary algorithm

Abstract This paper proposes a new direction for design optimization of a water distribution network (WDN). The new approach introduces an optimization process to the conceptual design stage of a WDN. The use of multiobjective evolutionary algorithms (MOEAs) for simultaneous topology and sizing design of piping networks is presented. The design problem includes both topological and sizing design variables while the objective functions are network cost and total head loss in pipes. The numerical technique, called a network repairing technique (NRT), is proposed to overcome difficulties in operating MOEAs for network topological design. The problem is then solved by using a number of established and newly developed MOEAs. Also, two new MOEAs namely multiobjective real code population-based incremental learning (RPBIL) and a hybrid algorithm of RPBIL with differential evolution (termed RPBIL–DE) are proposed to tackle the design problems. The optimum results obtained are illustrated and compared. It is shown that the proposed network repairing technique is an efficient and effective tool for topological design of WDNs. Based on the hypervolume indicator, the proposed RPBIL–DE is among the best MOEA performers.