A parallel surrogate-assisted multi-objective evolutionary algorithm for computationally expensive optimization problems

Abstract

A parallel surrogate-assisted multi-objective evolutionary algorithm for computationally expensive optimization problems. Computation, CEC 2008 (pp. 3177-3184 In: 2008 IEEE Congress on Evolutionary Preprint of paper published in Proceedings of IEEE Congress on Evolutionary Computation, 2008Abstract-This paper presents a new efficient multiobjective evolutionary algorithm for solving computationallyintensive optimization problems. To support a high degree of parallelism, the algorithm is based on a steady-state design. For improved efficiency the algorithm utilizes a surrogate to identify promising candidate solutions and filter out poor ones. To handle the uncertainties associated with the approximative surrogate evaluations, a new method for multi-objective optimization is described which is generally applicable to all surrogate techniques. In this method, basically, surrogate objective values assigned to offspring are adjusted to consider the error of the surrogate. The algorithm is evaluated on the ZDT benchmark functions and on a real-world problem of manufacturing optimization. In assessing the performance of the algorithm, a new performance metric is suggested that combines convergence and diversity into one single measure. Results from both the benchmark experiments and the realworld test case indicate the potential of the proposed algorithm.

Parallel surrogate-assisted global optimization with expensive functions – a survey

“…However, Syberfeldt et al (2008) point out that generational evolutionary algorithms have a disadvantage with respect to parallelization compared to steady state evolutionary algorithms. First the number of processors may not be a good match for the ideal population size, and also time may be wasted waiting for individuals with slow simulations.…”

Evolutionary optimisation of noisy multi-objective problems using confidence-based dynamic resampling

“…The proposed CDR technique is integrated in "Multi-Objective Parallel Surrogate-Assisted EA" (MOPSA-EA) (Syberfeldt et al, 2008). This section outlines the fundamentals of MOPSA-EA and presents how its parameters have been configured in the three optimisation problems.…”

“…The surrogate objective values assigned to solutions in O are adjusted to take the imprecision of the surrogate into consideration. This is done by modifying the values based on the calculated error of the surrogate (the details of this procedure are described in Syberfeldt et al, 2008). Based on the adjusted surrogate objective values, the most promising solution in O is selected to be to inserted into P .…”

Multi-objective evolutionary simulation-optimisation of a real-world manufacturing problem

“…4, the general implementation of MOPSA-EA is outlined. A detailed description of the algorithm can be found in [8]. The following parameter settings of the algorithm have been used in the study:…”

“…MOPSA-EA (described in [8]) is designed for complex and timeconsuming real-world SO problems and therefore suits well for the problem under consideration.…”