Method for monitoring running state of thermal power generating unit equipment

CN120909233APending Publication Date: 2025-11-07HUADIAN ELECTRIC POWER SCI INST CO LTD

Patent Information

Application Number
CN202511025861.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-07-24
Publication Date
2025-11-07

Smart Images

  • Figure CN120909233A_ABST
    Figure CN120909233A_ABST
Patent Text Reader

Abstract

The invention discloses an operation state monitoring method of thermal power generating unit equipment in the technical field of state monitoring, which comprises the following steps: acquiring real-time DCS data of thermal power generating unit operation, inputting the real-time DCS data into a pre-trained BSAT model, and outputting prediction parameters reflecting unit faults, monitoring the operation state of the thermal power generating unit based on the prediction parameter reflecting the fault of the unit and a pre-determined early warning threshold value; the training process of the BSAT model comprises the following steps: acquiring historical DCS data of thermal power generating unit operation; performing pretreatment; determining output parameters and input parameters according to the feature parameters obtained through preprocessing; performing standardization processing to obtain a training set and a verification set; constructing a BSAT model comprising a prediction base value output model and a residual prediction base value generation model; and training and verifying the BSAT model by using the training set and the verification set to obtain a trained BSAT model. According to the invention, a worker can make a judgment as early as possible, and unnecessary loss is reduced.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Thermal power generating unit adjusting device state monitoring method based on recurrent neural network

    CN118730499A

Cited By

  • Photovoltaic time sequence dynamic prediction method and system based on attention mechanism

    CN122118688A