Uncertainty and Constraint Handling in Evolutionary Algorithms

Abstract

This paper proposes two evolutionary algorithms. Firstly, a dynamic evolutionary algorithm is proposed that uses variable relocation vectors to adapt the current population to the new environment. The relocation vectors introduce a certain radius of uncertainty to be applied specifically to each individual and in effect restoring diversity and accelerating exploration. Furthermore, the algorithm provides higher re-usage, faster convergence and better adaptation. As a technique to be used at transient periods, the proposed algorithm provides the next evolutionary cycle with better initial population than any other randomly generated population. The algorithm can be easily integrated into standard evolutionary algorithms and other uncertainty handling techniques. Secondly, this paper proposes a new constraint handling technique for multi-objective evolutionary algorithms based on adaptive penalty functions and distance measures. Through this design, the objective space is modified to account for the performance and constraint violation of each individual. The modified objective functions are used in the non-dominance sorting to facilitate in evolution of optimal solutions not only in the feasible space but also in the infeasible space. The number of feasible individuals in the population is used to guide the search process either toward finding more feasible solutions or toward locating optimal solutions. The proposed method is simple to implement and does not need any parameter tuning.