Abstract: This paper introduces the software and hardware design of an intelligent robot cutting system based on a motion control card. The robot has four degrees of freedom and adopts a three-axis linkage working mode. The third axis can be tangent to the motion trajectory of the selected two axes. Experimentation shows that the robot cutting system works stably and meets the processing requirements of industries such as glass cutting. Keywords: robot , motion control card, glass cutting 1 Introduction Industrial robots are mechatronic automated production equipment that mimics human operation, is automatically controlled, reprogrammable, and can complete various tasks in three-dimensional space. They are particularly suitable for flexible production equipment with multiple varieties and variable batches. Based on their motion patterns, robots can generally be divided into five categories: cylindrical coordinate robots, spherical coordinate robots, Cartesian coordinate robots, articulated robots, and parallel robots. Cutting robots belong to the Cartesian coordinate type of industrial robots. Each type of cutting robot corresponds to specific processing materials and processes, such as domestically developed high-pressure water jet cutting robots and steel section cutting robot systems. For brittle materials such as glass, diamond-tipped tools are required, necessitating that the angle of the cutting head always remain tangent to the trajectory of the cutting pattern to achieve the best blade mark and slicing effect. Based on this application, we developed this intelligent robot cutting system. A schematic diagram of the robot model is shown in Figure 1: [align=center] Figure 1: Cutting Robot Model[/align] 2. Robot Hardware System Composition The control system uses three sets of AC servo systems and one set of cylinder equipment to realize the robot's four degrees of freedom. The robot system is driven by a GALIL DMC-1842 motion control card and an industrial PC. The X and Y directions are converted into linear motion through the transmission device to realize positioning and trajectory interpolation in the Cartesian coordinate system; the cutting head has two degrees of freedom, one is the Z direction of the cutting head rotation (as shown in Figure 1), the cutting head direction is always tangent to the trajectory of the XY plane during cutting, and the other degree of freedom is the cylinder linkage device on the cutting head that controls the lifting and lowering of the cutting head. In addition, an ISA bus IPC 5372-2 digital input card is used to detect the robot servo alarm status and improve the robot's self-diagnosis. 2.1 Motion control system and structural block diagram: The robot motion control adopts the DMC-1842 motion control card of the PCI bus of American GALIL [1]. This control card uses a 32-bit microprocessor, can control 4 axes, and has multi-axis linear interpolation, circular interpolation and other functions. In order to obtain the best control effect, the GALIL control card provides PID filter compensation function with speed, acceleration feedforward, Notch and low-pass filtering. All filter parameters can be adjusted to obtain the best performance of the servo system. The robot's AC servo drive system is set to speed mode to achieve optimal speed response performance. Among all servo parameters, the speed command voltage amplitude (VCMS) and speed loop proportional gain (Kvp) are connected in series in the main system loop and have the greatest impact on system performance, requiring joint tuning with the control card parameters. The robot motion control system block diagram is shown in Figure 2. [align=center] Figure 2: Robot Control System Block Diagram[/align] 2.2 Robot Input/Output Signals: Robot input signals include: servo alarm status for each of the X, Y, and Z axes (four alarm input lines per axis, requiring a total of 12 signal acquisitions); X and Y axis forward and reverse limit switch input signals; origin input signals for each of the X, Y, and Z axes; and two TTL input signals (cutting head probing protection signal IN1, laser detection of cutting object signal IN2). The robot output signals are six TTL output signals (three alarm reset output signals, and three output signals for the lubricant switch, positioning rod switch, and laser searcher switch). Among them, the X and Y axis forward and reverse limit switch input signals FLSX, RLSX, FLSY, RLSY and the cutting head probing protection signal IN1 are configured as interrupt signals and implemented using the control card command EI[3]. The cutting head pressure is generated by the open-loop output command OF of the control card W axis, which can accurately generate an analog voltage signal of +/-10V. 3 Robot software design 3.1 Robot cutting program flowchart and human-machine interface: The program flowchart of the robot cutting system is shown in Figure 3, and the human-machine interface is shown in Figure 4: [align=center] Figure 3: Robot cutting system program flowchart Figure 4: Robot human-machine interface[/align] 3.2 Thread implementation: The servo status monitoring thread obtains the servo alarm status of each axis by scanning the board address of the digital acquisition card through the CPU. If a servo alarm occurs, the thread will communicate with the robot cutting main thread to take corresponding protection measures. When the robot cutting main thread is idle, it is blocked and waits for the "cutting start" signal from the upper interface. The cutting function encapsulates the commands of the motion control card and can automatically select cutting parameters such as cutting speed, acceleration, and vector velocity smoothing coefficient based on the input parameters. The cutting graphic can be created by the user in the editing command area or by directly loading a CAD file. The cutting function and the human-machine interface were developed using VC++ 6.0 in Windows 98 Second Edition, with OpenGL as the drawing language, and the imported CAD file type is AutoCAD R14 DXF file. The main motion control functions of the robot are defined as follows: int CutLine(double x0,double y0,double x1,double y1) // Linear cutting function int CutArc(double radius,double startangle,double anglewidth) // Arc cutting function int CutEllipse(double MajorLength,double MinorLength,double startAngle, double angleWidth, double rotation) // Elliptical arc cutting function int ContinuousCut(char *filename,int cut_counts) // Polyline cutting function int CutLines(char *filename) // Arbitrary curve cutting function int SearchObject(double coordinate[3][2]) // Function for searching and cutting objects. For interrupt monitoring threads, the interrupt enable characteristics of the motion control card must first be configured. The limit interrupt enable and input 1 interrupt enable are configured using the command EI1024,1. The application can handle the interrupts issued by the control card by processing the WM_DMCINTERRUPT message, which is defined in DMCCOM.H[6]. 3.3 Matching of displacement resolution: Since the X-axis and Y-axis may have mismatched displacement resolution, that is, when the robot runs the same number of pulses on the two axes, the displacement is different. At this time, the ES[3] command is needed to correct the interpolation sequence. If the number of encoder lines of the X and Y axes are 2000 and 5000 respectively, and the walking distance of the motor per revolution is 27.494mm and 27.489mm respectively, then: Let m=1000×2000×27.489/5000=1099.56≈10996 n=1000×27.494=27494 At this time, the ES m, n command of the control card can ensure that the displacement resolution of the two axes is consistent during vector interpolation. 3.4 Clearance of transmission components: There is clearance in the transmission mechanism, also called backlash. Since the DMC-1842 motion control card has canceled the auxiliary encoder interface, it is not possible to simultaneously feed back the shaft end encoder signal and the position signal. One method to eliminate the clearance of transmission components is to upgrade the control card firmware to Rev s63b version, which provides a BK command to eliminate backlash. Other hardware clearance elimination methods include: clearance elimination by clearance-eliminating gears, clearance elimination by flexible gears, clearance elimination by symmetrical transmission, clearance elimination by eccentric mechanism, and clearance elimination by tooth profile elastic cover layer [4]. 3.5 Coordinate transformation: The robot cutting system adopts two cutting methods: positioning cutting and search cutting. It involves two rectangular coordinate systems: machine tool coordinate system and cutting object coordinate system. All graphic drawing is based on the cutting object coordinate system, and then transformed to the machine tool coordinate system during cutting. There are three cylinders on each of the X and Y directions on the robot cutting table. When positioning and cutting, the cylinders are raised and the object to be cut rests against the positioning rod. At this time, the coordinate transformation is equivalent to a coordinate translation. Search cutting involves roughly placing the object to be cut in the cutting area and letting the robot search for the object itself (the robot cutting head is equipped with a laser searcher, which generates an input signal when there is a reflective object below). At this time, it is only necessary to search for one point in the y direction and two points in the x direction of the object to determine the plane of the object to be cut (the object to be cut is usually rectangular). The coordinate transformation at this time is equivalent to a translation + rotation transformation. Let the coordinate system of the object being cut be [O';i', j'], and the coordinate system of the machine tool be [O;i, j]. The coordinates of the origin O' of the coordinate system of the object being cut in [O;i, j] are [x0, y0]T. If the coordinates of point M in [O;i, j] and [O';i', j'] are [x, y]T and [x', y']T respectively, and S is an orthogonal matrix, assuming the object being cut is placed at an angle, it can be understood that the coordinate system of the machine tool [O;i, j] is obtained by rotating the coordinate system of the object being cut by an angle θ counterclockwise. At this time, the transformation formula from the coordinate system of the object being cut to the coordinate system of the machine tool is [5]: 4 Experimental results: The robot servo uses the product of SANYO [2]. The specific parameters of the servo system of each axis are: X axis: Servo model: PY2A050A2 Motor model: P60B13150B Kp: 45 Kvp: 160 Tvi: 20 VCMS: 200 Y-axis: Servo model: PY2A030A2 Motor model: P50B08075H Kp: 37 Kvp: 240 Tvi: 15 VCMS: 200 Z-axis: Servo model: PY2A015A2 Motor model: P50B05020D Kp: 37 Kvp: 200 Tvi: 15 VCMS: 200 Control card parameters: KP 12,8,8 KI 8,2,2 KD 30,10,10 BK 5,5 MT,,-1 CE,,2 OE0,0 Figure 5 shows the XY-axis running trajectory collected when the robot is moving in three axes: [align=center] Figure 5: Robot running trajectory[/align] 5 Conclusion: This article introduces the implementation process of a cutting robot based on the DMC-1842 motion control card, from hardware to software. From the actual operation in the industrial field, the robot cutting system demonstrates stable performance, smooth cutting marks, high cutting accuracy, and fast cutting speed. The robot has reserved 6 TTL input interfaces, which can be used to improve the robot's safety features, such as object intrusion detection in the safe zone, thereby enhancing its safety performance. Furthermore, there are some areas for improvement, such as designing a better transmission mechanism to achieve higher positioning and cutting accuracy, fine-tuning the resolution of each axis according to temperature to overcome the impact of thermal expansion and contraction on accuracy, and designing an automatic tool changer to improve cutting efficiency. Collaborative Project: This project was jointly developed by Wujin Jixin Machinery Co., Ltd. and Jiangnan University. Jiangnan University completed the electrical system and automatic control components. This robot cutting system has already generated more than 2 million yuan in economic benefits for the company. References: 1. DMC-18x2 USER MANUAL Manual Rev.1.0e, By Galil Motion Control, Inc. 2. AC SERVO SYSTEM BL Super P Series PY2, INSTRUCTION MANUAL, SANYO DENKI. 3. DMC-18x2 COMMAND REFERENCE Manual Rev.1.0c, By Galil Motion Control, Inc. 4. "Robotics Technology and Its Applications", edited by Zhu Shiqiang and Wang Xuanyin, Hangzhou: Zhejiang University Press, July 2001. 5. "Linear Algebra and Analytic Geometry", edited by Yang Qitian, Dai Jun, and Han Weixin, Tianjin: Tianjin University Press, October 2002. 6. C/C++ Programmers Tool Kit Reference Manual, By Galil Motion Control, Inc.