Recognition of multi-scroll chaotic attractors using wavelet-based neural network and performance comparison of wavelet families

In this comparative study, the implementation of feature extraction and classification algorithm based on wavelet based neural network (WBNN) is presented for recognition of multi-scroll chaotic attractors using only one of the state variables of Chua’s circuit with a multi-segment resistor. Sixteen different feature extraction methods (Db1, Db2, Db6, Db10, Sym2, Sym3, Sym5, Bior1.1, Bior1.3, Bior2.2, Bior2.4, Bior2.6, Bior4.4, Coif1, Coif2, and Coif5) are generated by separately using Daubechies, Biorthogonal, Coiflets, and Symlets wavelet filters. WBNN model is used, which consists of two layers: adaptive wavelet entropy and multi layer perceptron (MLP) neural networks for expert multi-scroll chaotic attractor classification. The performance of this comparison system is evaluated by using total 600 different chaotic signals that have different initial values and resistors values for each of these feature extraction methods. The performance comparison of these features extraction methods and the advantages and disadvantages of the methods are examined.