The invention relates to a
circulating fluidized bed boiler combustion optimization and diagnosis method for acquiring and sorting out operation data such as
bed temperature,
bed pressure, load, efficiency,
coal supply amount, limestone supply amount, primary
air volume, primary wind pressure, upper secondary
air volume and wind pressure, lower secondary
air volume and wind pressure,
oxygen content, induced air volume, slagging amount,
NOx, Sox, etc., to build a neural
network model as the basis for optimization adjustment and diagnosis.
Field data of design value, measured value during operation, fuel
chemistry assay value and etc. of the
circulating fluidized bed boiler are utilized as the basis for optimization adjustment and diagnosis to calculate
boiler efficiency as the basis for optimization adjustment and diagnosis; a judgment criterion for the optimization adjustment of
combustion work condition is established and a corresponding solution is proposed, and therefore, optimization adjustment can be carried out when the
circulating fluidized bed boiler deviates from the optimum
combustion work condition; a judgment criterion for abnormal operation is established and a solution thereof is proposed; and therefore, problems which can not be solved after optimization adjustment can be diagnosed and analyzed. Common accident phenomena are summarized and
processing strategies can be proposed, and therefore, guidance can be provided to accident handling.