Integrated Qualitativeness in Design by Multi-Objective Optimization and Interactive Evolutionary Computation

Abstract

The concept of qualitativeness in design is an important one, and needs to be incorporated in the optimization process for a number of reasons outlined in this paper. Interactive Evolutionary Computation and Fuzzy Systems are two of the widely used approaches for handling qualitativeness in design optimization. This paper classifies the types of qualitativeness observed in design optimization, makes the case for their necessity, and proposes a novel framework for handling them, combining the two approaches in an evolutionary multi-objective optimization platform. Two components of the framework are tested using the floor-planning problem, and observations are reported. Future work is defined on the development of the framework.

Evolutionary algorithms with user’s preferences for solving hybrid interval multi-objective optimization problems

“…In the method proposed by Brintrup et al an optimization problem with several implicit objectives is focused on, and an evolutionary individual is evaluated in the performance of these implicit objectives combining IGAs with fuzzy systems. Having formulated the MOP together with the explicit objectives, they then solved it by using NSGA-II, and the proposed algorithm, called multi-objective IGA, was applied to an interior layout problem [8]. In the work of Brintrup et al, the performance of sequential IGA and multi-objective IGA is further compared, and verified that the latter is better than the former in solving a floor-planning problem [9].…”

Evolutionary algorithms for optimization problems with uncertainties and hybrid indices

“…We validate the performance of the proposed algorithm in solving an optimization problem with both precise parameters in explicit indices and precise implicit indices by comparing it with multi-objective IGA, a typical and effective algorithm to solve this problem [1]. For multi-objective IGA, a small population is employed when two kinds of indices are optimized, the user evaluates all individuals in each generation, and the comparison among different individuals is based on the Pareto dominance.…”

Evaluation of sequential, multi-objective, and parallel interactive genetic algorithms for multi-objective optimization problems

“…A more detailed description of this kind of problem can be found in ref. 4. Figure 5(b) shows an example of the user interface used in our experiments.…”

“…Our previous survey [4] showed that many problems previously believed to be quantitatively dominated, such as engineering design or system design, had (1) multiple conflicting objectives, and (2) subjective objectives among them. Subjective (qualitative) objectives may act as conflicting or complementary to quantitative objectives and are essentially unpredictable [5].…”

“…The process of solving a multi-objective optimization problem by EC is called Evolutionary Multi-objective Optimization (EMO). The particular suitability of EC for multi-objective optimization tasks together with the subjectivity inherent in many of those tasks led to the introduction of EMO incorporating IEC in order to capture subjective opinions of the designer [4]. Some other researchers followed similar ideas, such as gathering objective preferences from the designer [6], or asking the designer to pick favourable search areas [7], and our interactive multi-objective design optimization framework [8] used subjective opinions of the designer as an additional objective function, with promising results.…”