六足行走機器人設(shè)計,六足行走機器人設(shè)計,行走,機器人,設(shè)計 附錄B Key Techniques of the X2ray Inspection Real-timeImaging Pipeline Robot This paper presents a robotic system for weld-joint inspection of the big-caliber pipeline , which is developed for the purpose of being utilized as automation platform for X-ray real-time imaging inspection technique (RTIIT) . The robot can perform autonomous seeking and locating of weld-seam position in-pipe , and under the control of synchro-follow control technique it can accomplish the technologic task of weld inspection. The robotic system is equipped with a small focal spot and directional beam X-ray tube ,so the higher definition image of weld-seam can be obtained.Several key techniques about the robotic system developed are also explained in detail . Its construction is outlined. Key words : X-ray inspection ； real-time imaging ； robot 0 　Introduction Compared with radiographic examination technique(RET) , X-ray real time imaging inspection technique(RTIIT) has many advantages such as higher efficiency ,lower cost , better feasible automation and weld-defects evaluation on-line. Furthermore , up to date technology allows the X-ray RTIIT to be used in Non-Destructive Testing (NDT) of pipelines , and the inspection quality of this Technique is as good as that of the RET[1 ,2 ] . Therefore ,NDT equipments , which are used commonly in pipeline inspection and basing on the RET , need to be renovated by basing on the X-ray RTIIT. To employ the X-ray RTIIT in NDT of pipeline there must be an automation platform , and X-ray inspection real-time imaging pipeline robot ( IRTIPR) is designed for the purpose. In fact , besides the problems that have been resolved[3 ] and are involved in the X-ray IRTIPR , several key techniques are presented in this paper , in which we address the robot focusing on its intelligent control, i . e.the autonomous motion in-pipe , the synchro-follow controltechnique and the communication of cooperation between in-pipe and out-pipe , and we also outline the construction of the robot . 1 　Composing and Working Principle of the Robot The X-ray IRTIPR consists of the two parts of in-pipe and out-pipe , as illustrated in Figl 1. The out-pipe part is composed of image collecting and processing system (8 ,9 ,10) , out-pipe synchro-rotary mechanism and its driving system (11 ,12) . The image intensifier is driven by the out-pipe rotary mechanism to rotate round the center of pipeline to collect weld image and transmit video signal to image processing computer by image-collecting card. The in-pipe part is composed of in-pipe computer (1) , power and inverters system (2) , walking and driving system (3) , X-ray system (4) , in-pipe synchro-rotary mechanism and its driving system (5 ,6) and weld-seam autonomous seeking and locating system (7) . The X-ray tube in X-ray system is driven by the in-pipe rotary mechanism to rotate round the center of pipeline. Fig.1 The structure of X-ray IRTIPR The main working principle of the robot is explained as follows : Under the control of weld-seam autonomous seeking and locating system the in-pipe crawler finishes the localization of working position , at which the in-pipe crawler is in a state of waiting. When it receives the command signal from out-pipe , which is transmitted by low frequency electromagnetic wave , the in-pipe computer operates immediately the controller of X-ray system to realize its out-pipe control . In sequence the in-pipe and out-piperotary mechanisms are controlled by the synchro-followcontrol technique to rotate with the same center of pipeline and finish weld-seam inspection in the manner of rotating-irradiating-rotating. 2 　The Control System of the Robot According to the technologic process of working principle , the control system of X-ray IRTIPR is proposed and mainly made up of several key techniques such as the synchro-follow control technique based on the X-ray image of benchmark lead wire , the weld-seam autonomous seeking and locating technique based on data fusion and the communication of low frequency electromagnetic wave. 2. 1 　The Synchro-follow Control Technique of In-pipe and Out-pipe Rotary Mechanism In the light of the technologic requirement of X-ray RTIIT , the X-ray tube and the image intensifier must be required to rotate synchronously with the same center. Because the X-ray IRTIPR adopts wireless working manner , i . e. there is no tether cables linking in-pipe with out-pipe parts of the robot . How to realize the synchro-message communication between in-pipe and out-pipe control systems of rotary mechanism , or how to realize synchro-control , then becomes a key technique that must be solved. The synchro-follow rotating can be described as : when the in-pipe rotary mechanism drives X-ray tube to rotate an angle of α, the out-pipe rotary mechanism drives image intensifier to rotate the same angle synchronously with the same center too (Fig12) . Because of the shielding function of metal pipeline and wireless feature , the means of communication existed is difficult to accomplish control-message communication between in-pipe and out-pipe parts[4 ,5 ] . According to the particularity of X-ray IRTIPR , we put forward the synchro-control scheme as follows : a benchmark lead wire perpendicular to weld-seam is placed on the irradiation window of X-ray tube ; when the weld-seam is irradiated by X-ray , the benchmark lead wire is also imaged in out-pipe computer. As long as the in-pipe and out-pipe rotary mechanisms are in a synchronous position , namely the axis of irradiation window of X-ray tube is coincident with that of image intensifier (α= 0) (Fig12) , the image of benchmark lead wire is in the middle position of computer’s screen , i . e. the image of benchmark lead wire is coincident with the position of benchmark center-line ( H = 0) , see Fig13. When in-pipe rotary mechanism rotates an angle of α, the image of benchmark lead wire will deviate from benchmark center line on the screen , the distance is H. Then the distance H is used as an error input of control system of out-pipe rotary mechanism to regulate its rotating motion. Until the distance H is zero or smaller than appointed value , the synchro-follow motion of out-pipe rotary mechanism can be realized. The test and simulation prove that the above-mentioned synchro-follow control technique is correct . The synchro-motion satisfies the technologic requirement of X-ray RTIIT. The method utilizes X-ray as vision source , and the synchro-motion message of in-pipe and out-pipe rotary mechanisms is transmitted by the screen’s distance that the X-ray image of benchmark lead wire deviates from the benchmark center line , thus the synchro-motion is performed. The method has been applied for invention patent . Fig12 　The synchro2rotary mechanism Fig13 　The X2ray image of benchmark lead wire 2. 2 　Weld-seam Autonomous Seeking and Locating Technique Autonomous seeking and locating mean that the robot determines automatically where is the working position in-pipe with the help of sensors but without any one’s inter-meddling. This control-manner is actually“intelligent”. The precision and reliability of seeking and locating a system have direct relation with if a robot can realize autonomous motion in-pipe. If this system is disabled , the robot will take the place of the accident of“death”or“l(fā)ose the way”in-pipe[6 ] . Generally , methods for detecting the position of weld-seam are as follows : (1) Utilize encoder or cyclometer ; (2) Utilize the displacement caused by the protrusion-concave changing of weld-seam surface ; (3) Utilize if the zone of weld-seam conducts electricity ; (4) Utilize radioactive isotope ( such as γ ray source) ; (5) Utilize vision ; (6) Utilize low frequency electromagnetic wave. Because these methods are influenced by many factors such as walking wheel’s skid , the in-pipe environment , manmade factors , radioactive injury , locating precision and efficiency , satisfactory result can’t be obtained when one of the methods is used alone. Considering weld-seam regular array , i . e. the space between each weld-seam is about 12m , and advantages and disadvantages of each position-detection method , one system of weld-seam autonomous seeking and locating based on multi-sensors is put forward to improve and enhance the precision , efficiency and reliability of localization. Multi-sensors consist of the cyclometer , CCD camera and the receiver and emitter of low frequency electromagnetic wave. Systematic block diagram is depicted as Fig14. Fig14 　Weld-seam autonomous seeking and locating system The system adopts position feedback for enhancing the efficiency of localization. Vision servo is structured with image given feedback for realizing accurate localization. The data fusion based on three kinds of measure-data , which are the data of cyclometer , low frequency electromagnetic wave and vision , adopts the estimate-algorithm with priority to process data. In terms of the characteristics of three localization methods , the above data have different effective function region respectively. If X1 represents the measure-data of cyclometer , X2 of low frequency electromagnetic wave , X3 of vision. X represents the actual position in-pipe of the robot , the space between each weld-seam is 12m. Then , the effective function regions of three kinds of measure-data are as follows respectively : X1 ∈[ 1m ,12m] ; X2 ∈[ 0. 1m ,1m] ; X3 ∈[ -10cm ,10cm ] , the final localization goal is X3 = 0. The data fusion’s rule of three kinds of measure-data is described as : when the distance X1 away from weld-seam position is greater than 100cm , the cyclometer is employed for localization in order to enhance the efficiency , and let the in-pipe crawler move at a high speed ; when the data X2 is smaller than 100cm , the“attention”of the controller changes into the method of low frequency electromagnetic wave , and let the in-pipe crawler move at alow speed ; when the weld-seam enters the vision range , the vision servo is adopted for accurate localization. 　The data fusion’s rules are expressed as : X = X1 　　 if ( X3 > - 10) and ( X3 < 10) , then X = X3 ; The above-mentioned method that is realized with fuzzy control and datafusion has perfectly solved the contradiction between the precision and the efficiency of localization. The test result of localization precision is within ≤±3 mm , which can meet the design requirement . 2.3 The Communication of Low Frequency Electromagnetic Wave Besides the function of localization , low frequency electromagnetic wave is still utilized to transmit the off-on signal between in-pipe and out-pipe parts. Considering its dangers , the X-ray system is often operated with remote control from out-pipe. Because the robot is wireless and in view of the shielding function of metal pipeline , other methods cannot accomplish the mission that transmits the off-on signal between in-pipe and out-pipe parts. So the low frequency electromagnetic wave is adopted to transmit operation command for in-pipe computer to control the X-ray system. 3 　Conclusion Key techniques of the X-ray IRTIPR are assurances for X-ray RTIIT to realize automation. If a robot adopts the working means of having no cable and the synchro-follow control technique of in-pipe and out-pipe rotary mechanisms being not solved , it will be impossible for the X-ray RTIIT to realize automation at all . The weld-seam autonomous seeking and locating technique is a concrete embodiment of“intelligence”for the robot , and is also an assurance for the robot to work with high reliability. Low frequency electromagnetic wave realizes communication between a control system’s in-pipe and out-pipe parts under the condition of metal pipe’s shielding , and plays the role of closed loop of control system. The X-ray IRTIPR based on these key techniques can be used in the inspection of the big-caliber pipeline ( at ? 660 ～? 1400mm) , whose working distance is about 2km without charging and working speed is at 18m/ min. Because the robot is equipped with a small focal spot and directional beam X-ray tube , an image of weld-seam with higher definition can be obtained compared with other kinds of X-ray tube.These key techniques are proved in test and meet perfectly the technologic requirements of X-ray RTIIT. References [1 ] Zeng X Z, Sun Z C. Nondestructive Testing , 2001 , 12(12) : 530 [2 ] Zheng S C. Nondestructive Testing , 2000 , 22(7) : 328 [3 ] Jiang S Y. Research on in2pipe X2ray inspection robot technol2 ogy and theory : [ dissertation ] . Harbin : Harbin Institute of Technology , 2001 [4 ] Blettner A , Chauveau D , Becker. Robot for computerized real time radiographic inspection of on shore pipe welds. In : 6th European Conference on NonDestructive Testing. Nice , France : 1994. 225 [5 ] Anon. Welding in the world , 1990 , 28(5) :77 [6 ] Bjorkholm P J , Parker R , Johnson M. Design and application of a digital radiographic weld inspection system. In : Span An2 tonio , eds. 1990 ASNT Spring Conference. Texas , United States : 1990. 187 ７