Product Configuration Of Interactive Genetic Algorithm Based On Hesitate Degrees

Interactive Genetic Algorithm（IGA） is an evolutionary computation methodwhich takes the subjective evaluation value as an individual fitness. The most notablefeature of IGA is the integration of human intelligence. Individual fitness is specifiedby the user rather than function, so it is suitable to solve the fuzzy optimizationproblem of implicit performance indicators or evaluation indicators, and oneimportant application is in the field of product configuration. However, there are somelimitations of IGA, because the user via the HMI designating individual fitness.Frequent interaction can easily lead to user fatigue, and in the evaluation process, dueto the uncertainty of the target individual and the effects of fatigue, the sameindividuals at different evaluation stages may have different fitness value, which iscalled the evaluation of noise. In addition, using interactive genetic algorithm to solvethe problem of product configuration, it will produce an individual that does not meetthe constraints due to crossover and mutation randomness. Therefore, this paper doesthe following research on the above problem.(1) For the problem of noise and fatigue, a concept of hesitation was proposed,and the quantitative formula of hesitation degree was given. And the hesitationadjustment mechanism was established to reduce the evaluation process noise atdifferent stages of the evaluation process, ultimately to achieve acceleratedconvergence, reduce fatigue, and improve satisfaction purposes.(2) For the problem that crossover and mutation process does not meet theconstraints, the constraint is divided into structural and non-structural constraints. Andtwo kinds of constraint handling strategies-removal policy and modify strategies wasproposed to solve two kinds of constraints problems on the product configuration.(3) Taking into account the traditional product configuration model is not wellsuited for interactive genetic algorithm, the PIC model based on instance is proposed,to integrate the resources of one enterprise as a whole, and clarify the constraintbetween the various components of the module conditions for the subsequentoperation of constraint handling strategy. Finally, this paper establishes hesitate adjustment mechanism based on aninteractive genetic algorithm, and establish the car console concept design systembased on HAM-IGA. Comparing with the traditional interactive genetic algorithm, theresults show that the method can effectively reduce the system evaluation process, inparticular the evaluation of the evaluation of pre-noise, thus reducing the user’sevaluation of algebraic and evaluation time, to avoid the long-time rating and fatigue.And this method can significantly improve the results of user satisfaction with thesystem, and it has good application prospects in the field of product configuration.