Premium
Learning optimal control in deterministic systems
Author(s) -
Pareigis Stephan
Publication year - 1998
Publication title -
zamm ‐ journal of applied mathematics and mechanics / zeitschrift für angewandte mathematik und mechanik
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.449
H-Index - 51
eISSN - 1521-4001
pISSN - 0044-2267
DOI - 10.1002/zamm.19980781585
Subject(s) - discretization , dynamic programming , temporal difference learning , computer science , optimal control , mathematical optimization , similarity (geometry) , bellman equation , simple (philosophy) , control (management) , reinforcement learning , state space , state (computer science) , mathematics , artificial intelligence , algorithm , mathematical analysis , philosophy , statistics , epistemology , image (mathematics)
Learning algorithms for optimal control problems have similarity with numerical treatment of the Bellman‐equation of dynamic programming. The main difference is, that in case of learning the value iteration depends on information from the system, which is not necessarily given in the nodes of a state‐space discretization. Two updating‐schemes are presented for evaluating the learned information and their applicability is demonstrated on a simple learning problem.