Real-time online fan running state estimation method based on data mining

The invention discloses a fan running state real-time online estimation method based on data mining. The fan running state real-time online estimation method comprises the steps that historical running state parameters of a unit and a fan are obtained; extracting data under a plurality of normal and stable operation conditions, and constructing a sample set; obtaining a prediction model of a change relation between fan inlet flow and group load or main steam flow, and a prediction model of a change relation between system resistance from a hearth to an inlet section of the induced draft fan and system resistance from an outlet of the induced draft fan to an outlet section of a chimney along with flue gas amount; state operation parameters of the fan are obtained based on the two prediction models; calculating a theoretical stall safety coefficient, a pressure margin coefficient and a flow margin coefficient of the fan; and whether the operation state of the fan is safe after the state parameters of the flue g