Research On Optimization Method Of Plasma Spectral Stability Of Fly Ash Particle Flow

As an important economic and technical indicator for the operation of coal-fired boilers,the carbon content in fly ash directly reflects the combustion efficiency of the boilers.Real-time monitoring of the carbon content in fly ash helps to balance the combustion efficiency and the NOx concentration in the furnace by adjusting the ratio of air to pulverized coal and the fineness of pulverized coal in time,thereby ensuring the efficient and clean combustion of the boiler.However,the analysis results of traditional off-line fly ash carbon content detection method are seriously lagging behind,which makes it difficult to provide guidance for real-time adjustment of boiler operating parameters.And the current on-line detection methods have problems such as measuring cavity ash plugging and poor coal adaptability.In the early 21 th century,laser induced breakdown spectroscopy(LIBS)was proposed as a promising online detection method for carbon content in fly ash.When LIBS is used to detect powdery samples such as fly ash,the samples are usually pressed into pellets or stacked and tiled to be measured,or directly measured as a particle flow.Compared with the former two,direct measurement of LIBS in the form of particle flow does not require an additional pretreatment process,and can also realize online detection of carbon content in fly ash.However,due to the instability of fly ash particle distribution and particle size,the direct detection of particle samples by LIBS will lead to unstable plasma positions and incomplete ablative excitation,which will cause fluctuations in spectral intensity and even the appearance of invalid spectra,thus affecting subsequent quantitative analysis results.Therefore,this paper is devoted to the research on optimization method of particle flow fly ash plasma spectral data.An emission spectroscopy experimental device was set up for particle flow measurement,and quartz sand and fly ash particle samples were used to analyze the influence degree and causes of particle flow detection on spectral data instability.To reduce the influence of invalid spectra,a new conditional data processing scheme named standard deviation(SD)method was presented and evaluated for identifying spectral data of fly ash particle flow.The SD method was compared with the other two conditional data processing methods called signal-to-noise ratio(SNR)method and absolute peak intensity method.The results showedthat the SD method has great advantages in the accuracy and stability of valid data identifying.And the improvement of the quantitative analysis accuracy of fly ash samples further validates the feasibility of the SD method for optimizing the spectral stability.What's more,a spectral data optimization method named particle flow-spark induced breakdown spectroscopy(PF-SIBS)was proposed to solve the problem of excitation instability in LIBS particle flow detection.And the advantages of PF-SIBS technology compared with LIBS technology in the stability of spectral data were analyzed from qualitative and quantitative perspectives.Finally,the SD method and PF-SIBS technology were comprehensively applied to optimize the stability of spectra data obtained from fly ash particle samples.By combining with the multiple regression method,the quantitative analysis accuracy for prediction of carbon content in fly ash has been greatly improved,which verifies the feasibility of using the two together to optimize the stability of particle flow fly ash spectral data.