Research On Automatic Assembling Positioning Method Of Shield Segment Based On Photoelectric Detection

A shield machine is a type of tunnel boring equipment by shield method.It is widely used in infrastructure projects such as traffic tunnels,municipal pipelines,and energy pipelines because of its advantages of safe operation,fast boring speed and no impact on surrounding facilities.As an important subsystem of the shield machine,the erector is used for the assembly of tunnel lining segments,and its work performance is directly related to construction safety and tunnel quality.Manual assembly of segments has problems such as dangerous operation and low assembly quality.Therefore,it is urgent to conduct research on high-performance automatic segment assembly technology..In this thesis,to achieve automatic assembly and improve its performance,we carried out the research on the automatic assembly and positioning method of shield segment based on photoelectric detection from three aspects:positioning trajectory planning,segment pose detection and temperature detection of hydraulic positioning system.The main contents of this thesis are as follows:1.In order to realize the safe and accurate automatic assembly of shield machine segments,a motion model for automatic assembly and positioning of shield machine segments based on photoelectric detection is constructed.A positioning trajectory planning method based on the combination of joint space and Cartesian space based on photoelectric detection of segment poses is proposed,and a stable automatic assembly positioning trajectory is obtained,which meets the needs of automatic assembly positioning of shield machine segments based on photoelectric pose detection.2.The automatic assembly positioning method of segments based on vision-laser pose detection is studied.The automatic assembly positioning model of segments based on vision-laser poses detection is established.Aiming at the problems of low illuminance,low image contrast,and easy contamination of segment positioning targets,a feature extraction algorithm for visual positioning targets based on a two-stage deep neural network is proposed.Experiments show that compared with the single-stage deep neural network extraction algorithm,the average position detection distance error of this algorithm is reduced by 40.3%,and the recognition rate is increased by 28.3%.3.In response to the need for automated assembly for temperature detection of positioning hydraulic actuators,a temperature detection method based on semiconductor core fiber microcavity is proposed,and a temperature detection model for semiconductor core fiber microcavity is established.The thermo-optical characteristics near the absorption edge were analysed.Experiments show that the temperature sensor has a temperature sensitivity of ～221 pm/°C,and the detection error and standard deviation are ±0.184°C and 0.016°C,respectively,which can meet the temperature detection requirements of automatic assembly positioning hydraulic actuator.4.The experimental and applied research on automatic assembly positioning method of shield machine is carried out.The design and implementation of the segment automated assembly system based on photoelectric detection are completed..Using the vision-laser detection system composed of 4 smart cameras and 7 laser displacement sensors,the automatic assembly and positioning test of shield machine segments was carried out on the construction site,which realized the automatic assembly and positioning of shield machine segments.The test results show that:compared with the 10 mm positioning error of manual assembly,the automatic assembly positioning error of the segment is less than 3 mm,which verifies the effectiveness and reliability of the automatic assembly positioning method based on photoelectric detection.