Research On Blockage Prediction Of Coal Slurry Transportation Pipeline In Gangue Power Plant

The coal slurry as one waste is transported through pipelines to circulating fluidized bed boiler for mixed burning in gangue power plant. The blockage problem is always one of the important factors which affect the production efficiency and safety of coal slurry transportation pipelines. Blockage fault prediction and localization will significantly improve the performance of the coal slurry transportation system, which is also the important topic to guarantee the normal and safe running of circulating fluidized bed boiler. Based on wavelet transform and extreme learning machine, the methods for coal slurry transportation pipeline blockage prediction and localization are studied, in which the main research contents are presented as follows:Aimed at the problem that the monitoring data of coal slurry transportation system are susceptible to interference, the characteristics of field monitoring data are analyzed and missing data filling and data denoising are preprocessed. Three exponential smoothing method is used for filling the data processing, wavelet transform method is used for data denoising. The denoising effects of four different threshold methods are compared, and wavelet spatial correlation method is used for signal denoising, the better denoising effect is obtained. The monitoring data preprocessing is realized for the coal slurry pipeline, and the foundation is made for pressure distribution, blockage prediction and localization.For resistance loss problem of complex pipelines when the coal slurry is conveyed, coal slurry pipeline resistance loss mechanism and influence factors are analyzed, the resistance loss is the result of the combination of shear stress and frictional resistance. Based on mechanism analysis, complicated pipelines(horizontal straight pipe, tilted pipe, vertical pipe) pressure distribution model is established. The stress analysis is made for one segment of coal slurry in pipeline, force equilibrium equations are established, and to be solved in the nonlinear constraint conditions. The complex mathematical model of the pipeline pressure distribution is obtained, which is subject to complex index relationship, and improved calculation method is used in determining the frictional resistance coefficient. The multivariate factors of pressure distribution are analyzed, pressure distribution model based on QGA-BP which combines quantum genetic(QGA) and BP neural network is established.For coal slurry pipeline blockage predict problem, blockage predict method based on particle swarm optimization kernel function extreme learning machine(PSOKELM) is put forward. Here the blockage prediction mechanism is analyzed in coal slurry pipeline, the characteristic variable for blockage prediction is determined, and kernel function of support vector machine is introduced into extreme learning machine, the parameters are optimized by particle swarm algorithm, and applied to HuangLing coal gangue power plant, the actual test data are used for simulation experiments. Compared with support vector machine predict model optimized by particle swarm algorithm(PSOSVM) and kernel function extreme learning machine predict model(KELM), the results prove that the prediction model based on PSOKELM is superior to prediction model based on PSOSVM in speed and accuracy and superior to KELM prediction model in accuracy.For coal slurry pipeline blockage location problem, the positioning principle based on transient positive and negative pressure wave method is analyzed, wavelet transform spatial correlation method is used for positive and negative pressure wave signal denoising, singularity detection method is put forward, which is based on WPEMD-WTM, that is combination of empirical mode decomposition with wavelet packet preprocessing and wavelet transform modulus maxima. Then positive and negative pressure wave time difference is obtained, combined with the pressure wave velocity, the blockage point location is determined and the effectiveness and accuracy of the proposed block locating method is verified by simulation.For coal slurry pipeline blockage fault alarm problem, the blockage fault abnormal analysis method is studied. Pressure prediction value and prediction intervals are calculated, and according to the statistical characteristics of the pressure of sample data, abnormal pressure situation is judged. The warning threshold is determined, the warning level is divided, and then according to the different warning grades, different safety control measures are made, the rationality of the blockage fault safety control methods is verified.The slurry pipeline pressure distribution model that is put forward in this paper can provide a basis for pipeline design and paste pump selection. The blockage prediction model based on kernel function extreme learning machine and blockage localization method based on wavelet analysis can provide new security control decisions and means for blockage fault problem of coal slurry transportation system.