GMDH application for navigation of autonomous cranberry harvester on basis of cranberry distribution forecasting. Alexander Tyryshkin, Anatoliy Andrakhanov, Andrey Orlov

Abstract. In previous research works the authors showed possibility in principle to solve all the intellectual problems concerning the autonomous mobile robot (AMR) control with the help of group method of data handling (GMDH). In present paper an approach on optimal path planning was presented. One of the mentioned solutions is based on prediction of objective function which consists of extreme and restrictive components. Distribution prediction of objective function components’ partial derivatives of path is carried out with help of modified polynomial neural network. The results of prognoses for distributions of different difficulty and different methods of data sample generation (regular and chaotic grids) are provided. Optimal path planning for autonomous cranberry harvester developed by authors is carried out on basis of obtained prognoses.

Keywords. GMDH, autonomous mobile robot, prediction, objective function, optimal path.

References.

1. Andrakhanov A.A. Autonomous mobile robots: modern developments and control methods // IX International conference “Cybernetics and high technologies of XXI century”, Voronezh: VSU publishing, 2008, V.2. – P. 1141-1152, ISBN 978-5-9900659-3-2 (in Russian).

2. A.G. Ivakhnenko, G.A. Ivakhnenko The Review of Problems Solvable by Algorithms of the Group Method of Data Handling // Pattern Recognition and Image Analysis, Vol. 5, No. 4, 1995. – P. 527-535.

3. A.V. Tyryshkin, A.A. Andrakhanov, A.A. Orlov. GMDH Application for Autonomous Mobile Robot’s Control System Construction // Control systems and computers, No.1, 2009. – P. 62-68.

4. A.V. Tyryshkin, A.A. Andrakhanov, A.A. Orlov. GMDH Application for Autonomous Mobile Robot’s Control System Construction // In Proceedings of 2nd International Conference on Inductive Modelling (ICIM-2008), Kiev, 2008. – P. 306-309.

5. A.I. Samarin. Mobile robot adaptive behaviour model implemented using self-organising neural structures // MEPI Scientific Session - 2003. - Moskow, 2003. - NeuroInformatics-2002. - P. 106-120 (in Russian).

6. G. Booch, R. A. Maksimchuk, M. W. Engel, B. J. Young, J. Conallen, K. A. Houston. Object-Oriented Analysis and Design with Applications // Addison-Wesley Professional, 3/E, 2007.-720p.

7. J.-P. Laumond (ed.) Robot Motion Planning and Control // Lecture Notes in Control and Information Sciences, Vol. 229, Springer, 1998. – 343 p.