Pii: S0378-4754(99)00117-2
نویسندگان
چکیده
A novel approach to adaptive direct neurocontrol is discussed in this paper. The objective is to construct an adaptive control scheme for unknown time-dependent nonlinear plants without using a model of the plant. The proposed approach is neural-network based and combines the self-tuning principle with reinforcement learning. The control scheme consists of a controller, a utility estimator, an exploration module, a learning module and a rewarding module. The controller and the utility estimator are implemented together in a single radial basis function network. The learning method involves structural adaptation (growing neural network) and parameter adaptation. No prior knowledge of the plant is assumed, and the controller has to begin with exploration of the state space. The exploration–exploitation dilemma is solved through smooth transitions between the two modes. This enables rapid exploration response to novel plant dynamics and stable operation in the absence of changes in plant dynamics. The controller is capable of asymptotically approaching the desired reference trajectory, which is demonstrated in a simulation study. ©2000 IMACS/Elsevier Science B.V. All rights reserved.
منابع مشابه
Pii: S0378-4754(01)00382-2
Noise-induced absolute instability is demonstrated in computer experiments with a one-dimensional reaction– diffusion system. This phenomenon may occur in spatially extended flow systems of different nature, when the parameters are chosen close to the point of transition from convective to absolute instability. © 2002 IMACS. Published by Elsevier Science B.V. All rights reserved.
متن کاملPii: S0378-4754(00)00296-2
In this paper we produce numerical, genuinely three-dimensional, hexagonal traveling wave solutions of the Euler equations for water waves using a surface integral formulation derived by Craig and Sulem. These calculations are free from the requirements of either long wavelength or two-dimensionality, both of which are crucial to the KdV and KP scaling regimes, and we produce hexagonal travelin...
متن کاملPii: S0378-4754(01)00293-2
We have proved a new rotational sorting algorithm capable of reducing the complexity of data assignment process embedded in the maximum likelihood (ML)-based solution of a multitarget tracking problem from O(N3) of the conventional Hungarian type routines to O(N2) provided that the bearings-only measurements from an array of passive sensors are free from cluttering and missing data. © 2001 IMAC...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
عنوان ژورنال:
دوره شماره
صفحات -
تاریخ انتشار 1999