Abstract:
By combining T-S fuzzy model with RBF neural network, a new method for optimizing the coefficient of the consequence of T-S fuzzy RBF neural network is proposed via the adaptive DNA immune algorithm. In this method, the adjusting mechanism is based on antibody concentration and clone selection updating strategy so as to keep the antibody diversity and avoid the premature convergence. The proposed method is utilized to the soft-sensing modeling of the dry-point of gasoline delayed cooking. The experimental simulation results show that the proposed method is effective in the optimizing design of T-S fuzzy neural network system, and is of higher accuracy.