Visual Method Of Flame's Development And Application

China is one of the largest energy consumers in the world. A large portion of the fuel comes from coal. As a major power generation facility, power station boilers face a double mission of increasing combustion efficiency and reducing environmental pollution, which can be achieved only when the combustion process is optimized. Meanwhile, appropriate combustion will reduce boiler accidents. As a consequence, optimized combustion control is becoming essential for the improvement of overall boiler performance. However, currently the overwhelming strategy of combustion control is based on the signals (such as the pressure of the steam and the compositions of the flue gas) mainly collected outside combustion chambers, which inevitably causing delays between the collected data and the spontaneous conditions inside the combustion chamber.In fact, there exists abundant information in a flame that directly reflects the combustion conditions. Unfortunately these information have not been fully exploited. This thesis presents my research work in flame information acquisition and facilitation, for the purpose of combustion optimization.The study includes the principle of visual measuring system, the radioactive models for flame imaging, the light splitting and transmitting system, and the temperature reconstruction algorithm novel optical imaging system is developed with complete imaging algorithms.The system is able to perform real-time and field measurement of multiple flame parameters. It consists of an optical probe, a CCD camera, an optical assembly with optical filters, a flame gabber and a microcomputer with application software. The optical assembly, comprising beam splitters/prisms and band-pass filters, splits the flame light into different beams with distinct wavelengths. The selections of the band-widths of the filters are confirmed by the characteristics of flame measurement and the selection of the CCD camera by current-illumination characteristics. Throughtemperature calibration, a data bank was established providing the relationship between temperature and the gray level of the images.The system was applied on a 300MW boiler and valuable data are collected confirming the ability of the system to following the variations of the work conditions and to reflect online combustion characteristics.By the method of Least Square Support Vector Machines, we surveyed the boiler load from flame information. The flames parameters can assist optimize the combustion control technology.The novel instrumentation system for the measurement of the combustion flames has many advantages: flexibility in wavelengths selections, high accuracy of measurement, safety and stability in operation. It can be used extensively in measurement of sub transparent and non-transparent flames.