In recent years, by utilizing soft computing and wavelet theory, a number of efficient techniques are represented, among which are wavelet networks and fuzzy wavelet networks. In wavelet networks, the “Universal Approximation” property is guaranteed and an explicit link between the network coefficients and the wavelet transform is fulfilled and accordingly an initial guess for network parameters can be derived. Also, potential achievement of the same extent of approximation is provided with a network of reduced size. On the other hand, the wavelet networks are optimal approximators, because they require the smallest number of bits to obtain an arbitrary precision. The localization property of wavelet decomposition is reflected in the important properties of wavelet networks. The wavelet neural networks can approximate any function to an arbitrary precision with a finite sum of wavelets and can capture different behaviours of (global or local) approximated function. Also, the wavelet network provides an adaptive discretization of the wavelet transform by choosing influential wavelets based on a given data set and it is possible to handle problems of large dimension. In this thesis, we have presented a new adaptive fuzzy wavelet network controller (A-FWNC) for control of nonlinear affine systems, iired by the theory of multiresolution analysis (MRA) of wavelet transforms and fuzzy concepts. The proposed adaptive gain controller, which results from the direct adaptive approach, has the ability to tune the adaptation parameter in the THEN-part of each fuzzy rule during real-time operation. Each fuzzy rule corresponds to a sub-wavelet neural network (sub-WNN) and one adaptation parameter. Each sub-WNN consists of wavelets with a specified dilation value. The degree of contribution of each sub-WNN can be controlled flexibly. Orthogonal Least Square (OLS) method is used to determine the number of fuzzy rules and to purify the wavelets for each sub-WNN. Since the efficient procedure of selecting wavelets used in the OLS method is not very sensitive to the input dimension, the dimension of the approximated function does not cause the bottleneck for constructing FWN. Fuzzy wavelet network is constructed based on the training data set of the nominal system and the constructed fuzzy rules can be adjusted by learning the translation parameters of the selected wavelets and also determining the shape of membership functions. Then, the constructed adaptive FWN controller is employed, such that the feedback linearization control input