Research On Micro Bubble Detection Based On 51 Single Chip Microcomputer

In recent years,with the rapid development of social economy,China has paid more and more attention to marine safety and marine resources development,and marine engineering technology has also been developed.As an indispensable part of ocean engineering technology,submarine pipeline has been shouldering an important energy transport task.However,because the submarine pipeline has been in the seawater environment for a long time,it is extremely vulnerable to corrosion and leakage.In recent years,pipeline leakage accidents have occurred frequently at home and abroad.In order to prevent greater harm,it is necessary to find and confirm the location of submarine pipeline leakage in time.At the initial stage of leakage of underwater pipelines such as submarine pipelines,a bubble flow composed of tiny bubbles will be formed around the pipelines.However,the current detection technology is not only costly,but also limited in detection range and low in sensitivity,making it difficult to detect initial leakage.Therefore,a microbubble flow detection system consisting of PSO-SVM model and bubble flow detection circuit is designed in this paper.The system can detect the micro bubbles around the pipeline more intelligently,quickly and accurately,give timely warning for the risk of underwater pipeline leakage,and avoid economic losses to the greatest extent.In this paper,the propagation theory of underwater ultrasonic is introduced briefly,the influence of microbubble flow on ultrasonic transmission is clarified,and the feasibility of different detection methods is discussed.An experimental system was designed and built to simulate the formation of bubble flow in underwater pipelines,and the influence of bubble flow on ultrasonic signals was studied under controllable experimental conditions.Secondly,a preprocessing algorithm for microbubble flow detection is proposed,including signal denoising using wavelet threshold denoising,extracting the characteristics of waveform signals through Hilbert transform,and unifying the data of signal magnitude.Then,three machine learning models are used to predict the distance from the bubble flow to the emission probe.According to the evaluation index and other factors,the appropriate support vector machine model is selected for particle swarm optimization,and the PSO-SVM model is obtained.Thirdly,in order to improve the efficiency of bubble flow detection,the bubble flow detection circuit is designed with the 51 single-chip microcomputer STC12C5A60S2 as the core.The hardware of the circuit mainly includes the minimum system of single chip microcomputer,signal processing circuit and driving circuit,which can meet the requirements of data transmission,signal filtering and audible and visual alarm.The software of the circuit mainly includes the preparation of the MCU control program and the upper computer program,which can control the operation of the MCU and enable the upper computer to receive data and run the PSO-SVM model.Finally,in order to verify the effectiveness of the system functions,the micro bubble flow detection system composed of PSO-SVM model and bubble flow detection circuit is tested experimentally.Through Proteus software simulation,hardware test,communication test and prediction error test,it is confirmed that the functions of the microbubble flow detection system can operate normally,and can detect the gas tightness and leakage position of the pipeline.Aiming at the application background of ocean engineering,this research proposes the ultrasonic detection method of microbubble flow,and uses the PSO-SVM model based on machine learning algorithm and 51 single chip microcomputer to verify it in the laboratory under controllable conditions.The experiment shows that this method can quickly,effectively and accurately predict the position of bubble flow,and provide early warning for underwater pipeline leakage.It has broad application prospects.