Font Size: a A A

Research On Air-tightness Test Method Of Pressure Vessel Based On Multi-sensors Fusion

Posted on:2021-04-06Degree:MasterType:Thesis
Country:ChinaCandidate:J H LinFull Text:PDF
GTID:2392330614470094Subject:Computer technology
Abstract/Summary:PDF Full Text Request
Air-tightness test is one of the important means to ensure the quality of pressure vessels.It has become a development trend to use science and technology to improve the accuracy and efficiency of air-tightness tests.However,the air-tightness test methods currently proposed are still difficult to apply to real scenes,and their detection efficiency and accuracy cannot meet industrial requirements.This paper takes the air-tightness test method based on machine vision and pressure detection as the research goal,and aims to use the time series image and pressure changes to calculate the leakage source and leakage amount.The main research work and results are as follows:First,this paper proposes a multi-sensor fusion air-tightness test device.This device can effectively improve the efficiency of the air-tightness test.At the same time,the device is equipped with industrial cameras,pressure sensors,serial servers and industrial control machines.Among them,industrial cameras provide video frame data sources for leak location methods,and pressure sensors provide pressure data sources for leak volume calculation methods.Second,this paper explores solutions to the precise positioning of pressure vessels in multiple scenarios.Firstly,a pressure vessel initial positioning method based on convolutional neural network was proposed to roughly locate the vessel.Then,for the accuracy of the prediction frame and the test scenario,a precise positioning method for pressure vessels based on edge detection and a precise positioning method for pressure vessels based on seal plug positioning are proposed respectively.Third,this article discusses the solution to the leakage source location andleakage volume calculation.A multi-region leak location method based on background modeling is proposed.This method utilizes the distribution of front sight spots and frequency information of time series images to realize the location of multiple leak sources.A calculation method of local leakage is proposed.This method combines the visual data with pressure data to realize the calculation of leakage in the local leakage area.Fourth,the air-tightness test system and the air-tightness test cloud platform for pressure vessels have been developed.Among them,the pressure vessel air-tightness test system can effectively help production enterprises to improve the efficiency and reliability of air-tightness tests.The air-tightness test cloud platform can assist the special equipment supervision and inspection unit to manage the air-tightness test situation of various production enterprises.
Keywords/Search Tags:air-tightness test, computer vision, pressure detection, leakage calculation, bubble detection
PDF Full Text Request
Related items