Research On On-Line Monitoring Of Characteristics Parameters Of Gas-solid Two-Phase Flow Based On Acoustic Emission

In the operation of the power plant boiler,Monitoring the operating status helps to keep abreast of the internal combustion of the furnace to adjust the operating parameters at the right time.It plays a positive role in reducing pollutant emissions and improving operational economy and safety.In coal-fired power plants,due to the harsh internal environment of the furnace and opaque measurement field,passive acoustic emission technology as a non-contact acoustic measurement method which has advantages of non-invasion and non-interference of the flow field,has a broad application prospect.At the same time,acoustic emission acquisition system based on acoustic technology has also been successively introduced to provide reliable hardware guarantee for the acquisitions of acoustic emission signals.Therefore,this paper takes the application of power plant boiler as the background,acoustic emission non-destructive dynamic detection technology as the means,and the monitoring of the mass flow of pulverized coal particles in the primary air transport process as the goal,and carries out the research of the subject.First of all,the acoustic emission monitoring test platform of pneumatic conveying pipelines is built.Monitoring experiments are carried out after the completion of the experimental table.The research is mainly divided into two parts,one of which is the calibration experiment of the feeder to determine the relationship between the feed voltage and the feed amount.The second is the signal acquisition experiment.Acoustic emission acquisition is performed under 8 different mass flow rates by arranging the same type of sensors in the straight pipe section and the curved pipe section respectively.Secondly,since the collected acoustic emission signals are inevitably mixed with different degrees of noise signals,and the acoustic emission signal intensity itself is relatively weak.It is necessary to perform denoising processing on the collected signals.Therefore,this paper focuses on the denoising algorithm of acoustic emission signals.Based on the traditional wavelet packet analysis and empirical mode decomposition denoising,an improved joint denoising method is proposed.The denoising effect is verified by standard signals.The result shows that the improved algorithm has good denoising effect,and it can be used to complete the process of collecting acoustic emission signals.Finally,in order to realize the on-line measurement of pulverized coal mass flow,a joint training model was established by EEMD decomposition and BP neural network training for the effective acoustic emission signal after denoising.Finally,an optimal network model is found by changing the number of input layers,the number of layers in the hidden layer,and the number of nodes in the neurons.The model accuracy is verified by the actual acquired acoustic emission signals.The calculation results show that the measurement error in the straight pipe is within 15%,the measurement error of the elbow segment is within 12%and the error is within the acceptable range.It can provide an effective means for achieving particle phase mass flow measurement in the primary pneumatic conveying pipeline.