Multi-Objective Optimization with Controlled Model Assisted Evolution Strategies

Abstract

Evolutionary algorithms perform robust search in complex and high dimensional search spaces, but require a large number of fitness evaluations to approximate optimal solutions. These characteristics limit their potential for hardware in the loop optimization and problems that require extensive simulations and calculations. Evolutionary algorithms do not maintain their knowledge about the fitness function as they only store solutions of the current generation. In contrast, model assisted evolutionary algorithms utilize the information contained in previously evaluated solutions in terms of a data based model. The convergence of the evolutionary algorithm is improved as some selection decisions rely on the model rather than to invoke expensive evaluations of the true fitness function. The novelty of our scheme stems from the preselection of solutions based on an instance based fitness model, in which the selection pressure is adjusted to the quality of model. This so-called λ-control adapts the number of true fitness evaluations to the monitored model quality. Our method extends the previous approaches for model assisted scalar optimization to multi-objective problems by a proper redefinition of model quality and preselection pressure control. The analysis on multi-objective benchmark optimization problems not only confirms the superior convergence of the model assisted evolution strategy in comparison with a multi-objective evolution strategy but also the positive effect of regulated preselection in contrast to merely static preselection.

Developing and Applying the Information Construction Analysis Visualization System

“…WhereY i 0 is measured value of surrounding rock deformation, Y i is calculated value of surrounding rock deformation,m is the number of measured value, x i is i-th parameter, n is number of the parameter, x i a and x i b is Upper and lower limits of x i. Y i is a complicated nonlinear implicit function which selects parameter of rock x i (x 1 ,x 2 ,….,x n )as variables,introduced global optimization method-Difference Evolution (DE) Algorithm to global optimize parameter x i .Calculation process of back analysis of DE -FEM Coupling which have been formed by imbeding the FEM in DE [6] is as follows:…”

“…VTK(Visualization Toolkit) is the world famous open source graphics application function library [5] ,it provides a effective way for developing high quality visualization software efficiently. In the field of intelligent optimization, Difference Evolutionary(DE) is put forward by Kenneth Price and Rainer Storm, University of California, Berkeley [6] ..Appearance of above new technology method provides the opportunity to the development of tunnel monitoring analysis software system.…”