Research On Signal Processing Method Of Ultrasonic Detection Of Heat Transfer Fouling Based On Sparse Decomposition

Heat exchange systems play a huge role in industrial production and daily life in electric power,aerospace,chemical industry,etc,the heat exchange fouling formed inside the heat exchange system equipment due to the accumulation of impurities will not only lead to lower heat transfer efficiency,but also may cause heat exchange system safety accidents,threatening personal life and property safety,therefore the establishment of a set of perfect heat exchange fouling real-time detection system is an important part of overcoming this problem.Based on the accumulated experience of the previous research,it is found that the use of ultrasonic time-domain reflection method to detect heat exchange fouling has the advantages of real-time non-destructive,safe and simple.However,due to the complex structure of heat exchange equipments,heat exchange fouling texture rough and loose,ultrasonic reflection echoes mostly contain a lot of noise,echoes also exist the problem of mutual interference,resulting in the reading of fouling detection information becomes very difficult.To address the problem that it is difficult to accurately read useful information from the heat transfer fouling ultrasonic detection echo,this paper established a finite element model based on the theory of ultrasonic time-domain reflection detection of heat exchange fouling,and studied the sparse decomposition feature extraction method of ultrasonic reflection echo from both atomic dictionary model and sparse decomposition algorithm.Firstly,according to the structural characteristics of the reflected echo of heat transfer fouling ultrasonic detection,the mathematical model with the highest degree of similarity is selected to build overcomplete atomic dictionary,so that the accuracy of signal feature extraction can be guaranteed from the root.Secondly,this paper also conducted an improvement study for the sparse decomposition algorithm of matching pursuit,and proposed to optimize the search process of the best atom searched by matching pursuit method with the imperialist competitive algorithm to improve the search speed and accuracy of the best atom.Finally,in view of the limitations of the imperialist competitive algorithm,this paper proposed to improve the movement and update strategy of colony countries,improve the shortcomings of the original algorithm in global optimization,further improve the search accuracy of the optimal atom,and achieve high accuracy of signal reconstruction.Based on the above studies,the improved sparse decomposition method presented in this paper was applied to the processing of the simulated and measured signals of heat exchange fouling detection,and the signal processing effect was analyzed,the detection information was read.The signal processing results show that the signal processing method has strong noise reduction ability and the ability to separate the mixed signals,and can restore the reflected echo characteristic information with high accuracy,and the fouling thickness reading results have small error,which indicates that the heat exchange fouling ultrasonic signal processing method proposed in this paper has high value.