Apparatus for combustion optimization and method therefor

An apparatus for combustion optimization, comprising:

• a data layer configured to derive learning data necessary for designing a combustion model and a combustion controller from currently measured real-time data for boiler combustion and previously measured past data for the boiler combustion, wherein the data layer further comprises
• a data pre-processor configured to perform pre-processing for data comprising the real-time data and the past data, and
• a data analyzer configured to derive the learning data from the pre-processed data for designing the combustion model and the combustion controller based on analyzing a correlation between the pre-processed data using tags of the pre-processed data, and then selecting input data whose correlation degree with respect to output data of the combustion model is greater than or equal to a predetermined value, wherein the input data for analyzing the correlation comprises fuel input amount, air input amount, water input amount, and air temperature and the output data of the combustion model for analyzing the correlation comprises power output, temperature of steam and exhaust gas, composition of exhaust gas, and residual oxygen amount after combustion;
• a model layer configured to generate the combustion model and the combustion controller through the learning data;
• and
• an optimal layer configured to select any one combustion model among a plurality of combustion models by analyzing the real-time data, to calculate a target value for the combustion optimization by using the selected combustion controller after selecting any one combustion controller among a plurality of combustion controllers based on the selected combustion model, and to output a control signal according to the calculated target value.

A method for optimizing combustion, comprising:

• collecting, by a management layer, currently measured real-time data for boiler combustion, and determining whether to tune a combustion model and a combustion controller by analyzing the collected real-time data;
• deriving, by a data layer, learning data from the real-time data and previously measured past data for the boiler combustion in response to determining that the management layer determines to tune, wherein the deriving of the learning data further comprises
• pre-processing, by a data pre-processor of the data layer, data comprising the real-time data and the previously measured past data, and
• deriving, by a data analyzer of the data layer, the learning data necessary for designing the combustion model and the combustion controller based on analyzing a correlation between the pre-processed data using tags of the pre-processed data, and then selecting input data whose correlation degree with respect to output data of the combustion model is greater than or equal to a predetermined value, wherein the input data for analyzing the correlation comprises fuel input amount, air input amount, water input amount, and air temperature and the output data of the combustion model for analyzing the correlation comprises power output, temperature of steam and exhaust gas, composition of exhaust gas, and residual oxygen amount after combustion;
• generating, by a model layer, the combustion model and the combustion controller by using the learning data;
• calculating, by an optimal layer, a target value for combustion optimization through the selected combustion model and combustion controller among a plurality of combustion models and combustion controllers comprising the generated combustion model and combustion controller;
• and
• outputting, by the optimal layer, a control signal according to the calculated target value.