A nonlinear adaptive wavelet controller (AWC) using constructive wavelet networks is proposed for a class of general nonlinear dynamic systems. Wavelet networks are employed as a universal approximator for the highly nonlinear and non-affine dynamics. By virtue of the novel orthonormal property and multi-resolution property of wavelet networks, the structure of the nonlinear adaptive wavelet controller can be online adjusted constructively, according to the tracking performance of the dynamic system. The orthonormal property ensures that adding a new resolution (new wavelets) does not affect the existing wavelet network that may have been well tuned. The multiresolution property assures the improvement of the approximation precision when a new resolution is added. A coarse adaptive wavelet controller can be first constructed with a simple structure. An adaption period is introduced to specify the admissible waiting period for the tracking error to converge under the current wavelet structure. If the system fails to converge after the elapse of the admissible waiting period, a new wavelet resolution is considered to be necessary and added directly. In this manner, the adaptive wavelet controller can be easily constructed and tuned from a coarse to a fine level while retaining the closed-loop stability. It also provides an extra degree of freedom for users, in order to balance the network complexity and convergence rate.
[1]
S. Mallat.
Multiresolution approximations and wavelet orthonormal bases of L^2(R)
,
1989
.
[2]
R. Sanner,et al.
Gaussian Networks for Direct Adaptive Control
,
1991
.
[3]
Qinghua Zhang,et al.
Wavelet networks
,
1992,
IEEE Trans. Neural Networks.
[4]
Jun Zhang,et al.
Wavelet neural networks for function learning
,
1995,
IEEE Trans. Signal Process..
[5]
Jay A. Farrell.
On performance evaluation in online approximation for control
,
1998,
IEEE Trans. Neural Networks.
[6]
Fuchun Sun,et al.
Stable neural-network-based adaptive control for sampled-data nonlinear systems
,
1998,
IEEE Trans. Neural Networks.
[7]
M. Polycarpou,et al.
Stable adaptive tracking of uncertain systems using nonlinearly parametrized on-line approximators
,
1998
.
[8]
Alexander S. Poznyak,et al.
Nonlinear adaptive trajectory tracking using dynamic neural networks
,
1999,
IEEE Trans. Neural Networks.