Solving artificial ant problem using two artificial bee colony programming versions
- PDF / 2,478,520 Bytes
- 23 Pages / 595.224 x 790.955 pts Page_size
- 70 Downloads / 175 Views
Solving artificial ant problem using two artificial bee colony programming versions Fateh Boudardara1 · Beyza Gorkemli1
© Springer Science+Business Media, LLC, part of Springer Nature 2020
Abstract Artificial ant problem is considered as a sub-problem of robotic path planning. In this study, it is solved using two different methods: artificial bee colony programming and a new version of it called shrinking artificial bee colony programming. The former is a novel evolutionary computation based automatic programming method based on artificial bee colony algorithm and it was previously applied to this problem by the researchers. However, in this study, more comprehensive analyses and comparison study are provided. The shrinking artificial bee colony programming was developed in this study and its basic idea is to reduce the number of food sources, periodically, instead of a constant number used in the artificial bee colony programming. First, some parameter tuning studies were carried out for the shrinking artificial bee colony programming. Then, performances of the artificial bee colony programming, shrinking artificial bee colony programming and some other evolutionary computation based automatic programming methods were compared on Santa Fe and Los Altos Hills trails. Simulation results and the comparison study show that both of the algorithms can be used to solve the artificial ant problem effectively. Furthermore, the periodically decreasing population size property added to the artificial bee colony programming improves the performance of the algorithm on the artificial ant problem. While the proposed approach shows one of the superior performances among the considered methods, the results of the artificial bee colony programming are competitive to the methods in the comparison study. Keywords Robotic path planning · Artificial ant problem · Evolutionary computation based automatic programming · Artificial bee colony programming · Shrinking artificial bee colony programming · Dynamically decreasing population size
1 Introduction Robots are used in many different fields like robotic manufacturing, unmanned systems and vehicles (land, air, underwater), and they have important roles especially in the dangerous, small or unreachable areas where people do not have access to. Thus, there is an extensive research in the domain of autonomous robots (unmanned systems) [49].
This work was supported by Research Fund of the Erciyes University. Project Number: FYL-2018-7937. Beyza Gorkemli
[email protected] Fateh Boudardara [email protected] 1
Computer Engineering Department, Erciyes University, Kayseri, Turkey
One of the most fundamental issues for unmanned systems and robots is path planning [37]. Robotic path planning and autonomous vehicles are two popular application fields of artificial intelligence that have been extensively studied during the last decade. The robotic path planning is defined as follows: in an environment, which contains some obstacles, a mobile robot has to find an optimal (or
Data Loading...
