Abstract
This article shows how Machine learning techniques are tested to predict the performance of different exploration algorithms: Random Walk, Random Walk WSB and Q Learning, for robots moving on a bi-dimensional grid. The overall objective is to create a tool to help select the best performing exploration algorithm according to a configurable testing scenario, without the need to perform new experiments, either physical or simulated. The work presented here focuses on optimizing the topology of an Artificial Neural Network (ANN) to improve prediction results versus a previously proposed approach. The Hill Climbing algorithm is tested as optimization method, compared with manual trial and error optimization. The ANN was selected because it has the best performance indicators in terms of Relative Absolute Error and Pearson Correlation Coefficient compared with Random Forest and Decision Trees. The metric used to measure the performance of the exploration algorithms is Maximum Number of Steps to target.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Zhang, Y., Gong, D., Zhang, J.: Robot path planning in uncertain environment using multi-objective particle swarm optimization. Neurocomputing 103, 172–185 (2013)
Wu, H., Tian, G., Huang, B.: Multi-robot collaboration exploration based on immune network model. In: IEEE/ASME International Conference on Advanced Intelligent Mechatronics, AIM 2008, pp. 1207–1212 (2008)
Andreychuk, A., Bokovoy, A., Yakovlev, K.: An empirical evaluation of grid-based path planning algorithms on widely used in robotics raspberry pi platform. In: The 2018 International Conference on Artificial Life and Robotics (ICAROB 2018), pp. 1–4 (2018)
Akutsu, T., Yaoi, S., Sato, K., Enomoto, S.: Development and comparison of search algorithms for robot motion planning in the configuration space. In: Proceedings IROS 1991: IEEE/RSJ International Workshop on Intelligent Robots and Systems ’91, pp. 429–434 (1991)
Faigl, J., Simonin, O., Charpillet, F.: Comparison of task-allocation algorithms in frontier-based multi-robot exploration. In: Bulling, N. (ed.) EUMAS 2014. LNCS (LNAI), vol. 8953, pp. 101–110. Springer, Cham (2015). https://doi.org/10.1007/978-3-319-17130-2_7
Kulich, M., Juchelka, T., Přeučil, L.: Comparison of exploration strategies for multi-robot search. Acta Polytech. 55(3), 162 (2015)
Holz, F., Behnke, S., Basilico, N., Amigoni, F.: Evaluating the efficiency of frontier-based exploration strategies. In: ISR/Robotik 2010, vol. 1, no. June, p. 8 (2010)
Juliá, M., Gil, A., Reinoso, O.: A comparison of path planning strategies for autonomous exploration and mapping of unknown environments. Auton. Robots 33(4), 427–444 (2012)
Amigoni, F.: Experimental evaluation of some exploration strategies for mobile robots. In: IEEE International Conference on Robotics and Automation (2008)
Martínez Puerta, J.J., Vallejo Jiménez, M.M.: Comparación de estrategias de navegación colaborativa para robótica móvil. Universidad Autónoma de Manizales (2016)
Caballero, L., Benavides, C., Percybrooks, W.: Machine learning-based system to estimate the performance of exploration algorithms for robots in a bi-dimensional grid. In: 2017 IEEE 3rd Colombian Conference on Automatic Control (CCAC), pp. 1–6 (2018)
Pitarque, A., Ruiz, J.C., Roy, J.F.: Las redes neuronales como herramientas estadísticas no paramétricas de clasificación. Psicothema 12(SUPPL. 2), 459–463 (2000)
Tabachnick, B., Fidell, L.: Using Multivariate Statistics. Harper & Row, New York (1996)
Cohen, J., Cohen, P.: Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences, 545 p. Taylor & Francis, Milton Park (1983)
James, G., Witten, D., Hastie, T., Tibshirani, R.: An Introduction to Statistical Learning, 8th edn. Springer, New York (2017). https://doi.org/10.1007/978-1-4614-7138-7
Lau, B., Sprunk, C., Burgard, W.: Efficient grid-based spatial representations for robot navigation in dynamic environments. Rob. Auton. Syst. 61(10), 1116–1130 (2013)
Flach, P.: Data, Machine Learning: The Art and Science of Algorithms that Make Sense of Data. Cambridge University Press, New York (2012)
machine learning - How to interpret error measures? - Cross Validated (2017). https://stats.stackexchange.com/questions/131267/how-to-interpret-error-measures. Accessed 28 Feb 2018
Cross-Validation - MATLAB & Simulink (2018). https://la.mathworks.com/discovery/cross-validation.html. Accessed 20 Apr 2018
Izquierdo, S.K., Rodó, D.M., Bakx, G.E., Iglésias, R.B.: Inteligencia artificial avanzada. Editorial UOC - Editorial de la Universitat Oberta de Catalunya (2013)
Koller, D., Friedman, N.: Probabilistic Graphical Models: Principles and Techniques. MIT Press, Cambridge (2009)
Author information
Authors and Affiliations
Corresponding authors
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2018 Springer Nature Switzerland AG
About this paper
Cite this paper
Caballero, L., Jojoa, M., Percybrooks, W. (2018). Optimized Artificial Neural Network System to Select an Exploration Algorithm for Robots on Bi-dimensional Grids. In: Serrano C., J., Martínez-Santos, J. (eds) Advances in Computing. CCC 2018. Communications in Computer and Information Science, vol 885. Springer, Cham. https://doi.org/10.1007/978-3-319-98998-3_2
Download citation
DOI: https://doi.org/10.1007/978-3-319-98998-3_2
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-98997-6
Online ISBN: 978-3-319-98998-3
eBook Packages: Computer ScienceComputer Science (R0)