Symbolic regression and evolutionary computation in setting an optimal trajectory for a robot

Abstract

The paper deals with a novelty tool for symbolic regression - Analytic Programming (AP) which is able to solve various problems from the symbolic regression domain. One of tasks for it can be setting an optimal trajectory for artificial ant on Santa Fe trail which is the main application of Analytic Programming in this paper. In this contribution main principles of AP are described and explained In second pan of the article how AP was used for setting an optimal trajectory for artificial ant according the user requirements is in detail described. AP is a superstructure of evolutionary algorithms which are necessary to run AP. In this contribution 3 evolutionary algorithms were used Self Organizing Migrating Algorithm, Differential Evolution and Simulated Annealing. The results show that the first two used algorithms were more successful than not so robust Simulated Annealing.