|
|
 |
| |
|
|
|
A Self-Organizing Recurrent Wavelet Neural Network for Nonlinear Dynamic System Identification |
|
|
|
PP: 125-132 |
|
|
|
Author(s) |
|
|
|
Cheng-Jian Lin,
Chun-Cheng Peng,
Cheng-Hung Chen,
Hsueh-Yi Lin,
|
|
|
|
Abstract |
|
|
| To solve identification of nonlinear dynamic systems, a recurrent wavelet neural network (RWNN) model is proposed in
this paper. The proposed RWNN model has four-layer structure. Temporal relations embedded in the network by adding some feedback
connections representing the memory units in the second layer. An online learning algorithm, which consists of structure learning and
parameter learning, is proposed and is able to construct the wavelet neural network dynamically. The structure learning is based on
the input partitions to determine the number of wavelet bases, and the parameter learning is based on the supervised gradient descent
method to adjust the shape of wavelet functions, feedback weights, and the connection weights. Computer simulations were conducted
to illustrate the performance and applicability of the proposed model. |
|
|
|
|
|
|
|
