TY - CHAP
T1 - Performance Analysis of a Distributed Steady-State Genetic Algorithm Using Low-Power Computers
AU - Martínez-Vargas, Anabel
AU - Cosío-León, M. A.
AU - García-Pérez, Andrés J.
AU - Montiel, Oscar
N1 - Publisher Copyright:
© 2021, The Author(s), under exclusive license to Springer Nature Switzerland AG.
PY - 2021
Y1 - 2021
N2 - In this chapter, we describe the implementation and evaluation of a distributed steady-state genetic algorithm on low-power computers. Specifically, we integrate the NVIDIA Jetson card and the ESP32 cards to form a master-slave model. NVIDIA Jetson card is the master, whereas the ESP32 cards are the slaves. We evaluated the distributed steady-state genetic algorithm on two challenging combinatorial problems: the n-Queens problem and the travelling salesman problem. To compare the performance of the distributed steady-state genetic algorithm on those well-known problems, we implemented a sequential steady-state genetic algorithm. The simulation results indicate that the distributed steady-state genetic algorithm can escape local minima in the travelling salesman problem; hence, the solutions have a better quality of fitness than the sequential genetic algorithm ones. In contrast, for the n-Queens problem, both genetic algorithms’ performance is remarkably similar.
AB - In this chapter, we describe the implementation and evaluation of a distributed steady-state genetic algorithm on low-power computers. Specifically, we integrate the NVIDIA Jetson card and the ESP32 cards to form a master-slave model. NVIDIA Jetson card is the master, whereas the ESP32 cards are the slaves. We evaluated the distributed steady-state genetic algorithm on two challenging combinatorial problems: the n-Queens problem and the travelling salesman problem. To compare the performance of the distributed steady-state genetic algorithm on those well-known problems, we implemented a sequential steady-state genetic algorithm. The simulation results indicate that the distributed steady-state genetic algorithm can escape local minima in the travelling salesman problem; hence, the solutions have a better quality of fitness than the sequential genetic algorithm ones. In contrast, for the n-Queens problem, both genetic algorithms’ performance is remarkably similar.
KW - Distributed steady-state genetic algorithm
KW - ESP32
KW - Master-slave model
KW - NVIDIA jetson
KW - Travelling salesman problem
KW - n-Queens problem
UR - http://www.scopus.com/inward/record.url?scp=85103730143&partnerID=8YFLogxK
U2 - 10.1007/978-3-030-68776-2_3
DO - 10.1007/978-3-030-68776-2_3
M3 - Capítulo
AN - SCOPUS:85103730143
T3 - Studies in Computational Intelligence
SP - 41
EP - 70
BT - Studies in Computational Intelligence
PB - Springer Science and Business Media Deutschland GmbH
ER -