Multi-input system identification and its applications using wavelet constructive method

In this paper, algorithms for the identification of a nonlinear multi-input dynamical system are developed. These algorithms are then used for developing a decision supporting tool for local short term rain storm forecasting. A wavelet network structure is applied to track the dynamics of a nonlinear system and used for identifying a state space model of the system using radar images as input and the rain gauge measurement as output. When comparing with traditional artificial neural network approaches, we have shown that this wavelet approach offers significant improvement in terms of speed, reliability and accuracy