Research On Transverse Vibration Monitoring Method Of Ship Propulsion Shaft Based On Machine Vision

As the main component of the intelligent engine room power unit system,the ship propulsion shaft system undertakes an important mission and plays a very important role in the efficient and safe navigation of ships.The lateral vibration signal is an important signal that needs to be perceived to promote the safe operation of the shafting system,and has high research significance in the perception and analysis of the working state of the shaft system.Most of the commonly used vibration sensors in shafting systems are electrical measurement type,installed and fixed position,but the vibration signal at a single point may not show the overall operating status for the shaft system with a large span,there are certain limitations,in addition,such sensors are expensive,the installation route is complicated,and the maintenance process is complicated and cumbersome.In order to solve the above problems,this thesis starts from the newly developed machine vision vibration perception technology,and proposes a vision-based transverse vibration monitoring method for propulsion shafting to realize the perception and analysis of the vibration displacement of the shafting at multiple points.The main research contents of the thesis are as follows:(1)Construction of machine vision vibration recognition model.Firstly,rectangular stripe labels are selected according to the rotation and vibration characteristics of the propulsion shafting.Secondly,the image processing algorithm is used to preprocess the video picture to reduce the influence of external environmental factors on the machine vision processing results.Then,in order to reduce the influence of camera vibration,the physical displacement is processed and calculated by selecting the pixel coordinate value of the relative center of gravity,and then the axial visual vibration displacement data is obtained.Finally,the dynamic kernel principal element analysis algorithm is used to analyze the obtained multi-point vibration signal of the shafting system,so as to realize the overall perception and comprehensive analysis of the operating state of the shafting system.(2)Build a test bench to verify the feasibility of the machine vision vibration recognition model.Based on the factors such as low speed of the propulsion shaft system,complex components and the characteristics of force and lateral vibration,this thesis independently designs and constructs the shafting visual perception test platform by selecting the shafting size of the "Yukun" ship,which mainly includes the rotary shafting unit and the visual perception unit.According to the simulation results of shafting rotation,the test scheme is further improved,and three types of test schemes are planned to obtain the monitoring video of the shafting system in different motion states,and then obtain vibration displacement data through calculation.(3)Model verification and analysis.By comparing and analyzing the fixed point visual vibration data obtained by the test shaft system under different test schemes with the laser displacement sensor data,it can be obtained that the data trend and peak value obtained by the algorithm proposed in this thesis at low speed are similar to those of the displacement sensor,and the reliability and practicability of the information extraction algorithm are verified.The signal analysis model is used to analyze the vibration data of 60r/min and 180r/min,and the final results show that the proposed analysis model is accurate for data anomaly localization and the overall perception effect meets the analysis requirements.According to the experimental verification and analysis results,the proposed method can realize the perception and analysis of rotating shaft vibration signals,and has certain application prospects in the auxiliary perception of intelligent ship shafting vibration displacement state.