Electrode elevation control method and system of submerged arc furnace based on neural network decoupling

The invention discloses an electrode elevation control method and system of a submerged arc furnace based on neural network decoupling. According to the method, an electrode current-position decoupling model of the submerged arc furnace is established, a generalized external interference item is identified from the electrode current-position decoupling model on the basis of an RBF neural network algorithm, an electrode elevation displacement amount is derived from the electrode current-position decoupling model of the submerged arc furnace, the generalized external interference item and a tracking error, the tracking error represents a deviation of the real-time electrode current value from an electrode current set value, and electrode elevation of the submerged arc furnace is controlled on the basis of the electrode elevation displacement amount. The technical problems that a severe coupling effect and random interference in the production process of the submerged arc furnace causes instability of electrode