Abstract: Fuzzy control is a human-imitating control technique, which is independent of mathematical model of plants. It utilizes priori knowledge to carry out approximate reasoning. But it is lack of the ability of self-tuning or self-learning in industrial applications. The temperature control process of industrial Process is a multivariable and nonlinear dynamic system. To cater for the practical requirements, this paper presents a fuzzy network control strategy which is able to enhance the capacity of self-learning of fuzzy control rules, based on the self-leaning ability of neural networks. Simulation research and physical analog experiment prove the feasibility of this control strategy.