| 摘要 |
The present disclosure is related to a signal processing method for a combustion process optimization in industrial burners, which are used for the flame characterization in industrial and central furnace burners that generate energy and are fed with fossil fuels. Particularly, the present invention refers to a method for the combined signal processing from an arrangement of flame detectors and the operating data of the furnace or boiler, for the flame characterization. The method of the present invention processes the data online of an arrangement of the flame detectors combining statistical methods on linear and non-linear characteristics in the time and frequency domain and uses advanced control schemes as heuristic methods based on an artificial bee colony for the combustion optimization, Fuzzy-c Means (FCM) Artificial neural networks for the modeling of the combustion process and an expert system for advanced multivariable control routines. The flame characterization is useful in the burner’s optimization in industrial furnaces and boilers in electricity generation processes, tuning or adjustment of combustion in the burners, monitoring and programming of the burner maintenance conditions and repairs, improvements in the fuel efficiency and the reduction of harmful emissions in the chimney. |
