Data-based blast-furnace gas dynamic predication method for metallurgical enterprises

The invention discloses a data-based blast-furnace gas dynamic predication method for metallurgical enterprises, which comprises the following steps of: firstly, carrying out noise removal treatment on the known historical data with an empirical mode decomposition noise removal method, and then, dynamically predicting production and consumption of a blast-furnace gas system with a time sequence predication method based on an echo state network; and secondly, introducing a Bayesian reasoning method into the echo state network, and establishing a relation model between the gas flow of all main gas users and the position of each gas chamber for predicting the changes of the position of a blast-furnace gas chamber in an appointed time span after the current time. The data-based blast-furnace gas dynamic predication method disclosed by the invention can be used for accurately predicting the production and consumption of the blast-furnace gas system and the changes of gas chambers so as to provide an online decisio