Reduced human fatigue interactive evolutionary computation for micromachine design

Abstract

This paper presents a novel method of using Interactive Evolutionary Computation (IEC) for the design of Microelectromechanical Systems (MEMS). A key limitation of IEC is human fatigue. Based on the results of a study of a previous IEC MEMS tool, an alternate form that requires less human interaction is presented. The method is applied on top of a conventional multi-objective genetic algorithm, with the human in a supervisory role, providing evaluation only every n th -generation. Human interaction is applied to the evolution process by means of Pareto-rank shifting, which is used for the fitness calculation used in selection.Results of a test of 13 users shows that this IEC method can produce statistically significant better MEMS resonators than non-interactive evolutionary synthesis.

Interactive Evolutionary Computation with Evaluation Characteristics of Multi-IEC Users

“…Once the user's evaluation behavior is learned and can be modeled, there are several methods that use this information to reduce IEC user fatigue. For example, displaying phenotypes of individuals to a user in the order of predicted evaluation value helps to reduce the user fatigue because the user need only compare neighboring individuals only [4]; displaying individuals with predicted evaluation values reduces the number of user evaluation inputs required because there is no need for the user to input an evaluation value if it is same as the displayed predicted evaluation values [12]. A particularly interesting method to reduce fatigue is displaying to the user only the top 10 or 20 individuals obtained during an evolution of a large population while using a model to evaluate the remaining individuals.…”

“…The parameters of 3 lights irradiated to a CG model are adjusted by IEC, and then the user can create a desired lighting effect. 10 subjects evaluated 15 lightings models for their effectiveness [12]. An interface of the system is as shown in Fig.…”

“…We compare the distance of the evaluation between real users in three applications, i.e., Logo design system [17], CG lighting system [18] and MEMS optimization system [12], with the result of the simulation using the pseudo-IEC user mentioned in sections III. If the distance is comparable with the pseudo-user, we can expect the similar convergence as the simulation.…”

Interactive Genetic Algorithms for User Interface Design

“…Kamalian et al make the user evaluate a subset of the population every t th generation [7]. We adopt a similar approach putting the user in a supervisory role and thus reduce the amount of user feedback, but focusing on how the frequency of user input affects IGA performance and fatigue.…”

Evolutionary algorithms for optimization problems with uncertainties and hybrid indices

“…After evolving the population for some generations by employing traditional multi-objective GAs, the user then evaluates the optimal solutions according to implicit indices and their offspring are then generated based on the evaluation of the user [19]. Kamalian et al also launched a proposal for a number of individuals to be evaluated by the user in every some generations and to have their Pareto ranks changed based on these evaluations [20]. Recently, we solved an optimization problem with multi-dimensional hybrid indices using IGAs.…”