Abstract: This paper introduces a novel non-contact robot measurement and control system, discussing its design concept, composition, and working principle. The system consists of a non-orthogonal coordinate actuator, a laser displacement sensor, a CCD camera, and a master-slave control system. It can achieve high-precision measurement of the geometric dimensions and surface defects of spherical and rotary workpieces within a limited measurement space. Tests show that the measurement and control system has reliable stability and high measurement accuracy. 1 Introduction With the development of science and technology and modern manufacturing, the manufacturing precision of workpieces is increasing, thus demanding higher precision and functionality from measuring equipment. The need for new, specialized measuring equipment is also growing. Traditional measuring machines are mostly based on a single geometric coordinate system, such as Cartesian or cylindrical coordinates. These measuring machines have relatively intuitive mechanical structures, simple control algorithms, high measurement accuracy, and their error models have been perfected through years of research. However, in some special applications, these measuring machines are unsuitable. Non-orthogonal coordinate measurement systems, due to their high flexibility, have become the development trend of coordinate measuring machines. Through extensive investigation, research, scheme comparison, parameter calculation and optimization, and computer simulation, and fully considering accuracy, efficiency, reliability, operability, and spatial compatibility, this paper develops a novel robot measurement and control system by combining a robot mechanism with laser non-contact measurement sensor technology to achieve automatic non-contact measurement of a hemispherical object in a limited space. This system features high speed and high precision in measuring the geometric dimensions and surface defects of spherical workpieces. 2. System Working Principle This system consists of three parts: a mechanical actuator, a CCD camera and laser displacement sensor, and a measurement and control system. The mechanical actuator is the main part of the non-contact measurement system. As shown in Figure 1 , the mechanical structure mainly consists of a base, a robot arm, a high-precision rotary table, and a two-dimensional leveling worktable. The laser displacement sensor is installed at the end of the robot arm, and the CCD camera is installed below the laser displacement sensor. The robot arm, with three degrees of freedom, serves as the main measuring component. A high-precision rotary spindle is mounted at the center of the base. A two-dimensional leveling table is installed at the top of the spindle to align the rotation axis of the workpiece with that of the rotary table. During measurement, the workpiece is placed on the measuring platform, and the rotary table rotates one revolution. The measuring system measures the position of the hemisphere's center, and the positional error is displayed on the panel. Following computer prompts, the hemisphere is aligned concentrically with the rotation axis. Then, the robotic arm moves to the first measurement position, the table rotates, and the laser sensor performs the measurement. The robotic arm then moves to the next measurement position, repeating the process. Once the entire spherical surface is scanned, the surface features of the sphere are obtained. This measuring system can measure not only hemispherical shells but also various combinations thereof, as well as other rotary workpieces. [align=center] 1-Laser sensor and CCD camera 2-Robot arm 3-Sector wheel 4-Measured part 5-Rotary table and leveling table Figure 1 Mechanical structure of the measuring robot[/align] 3 Control System Design This system has complex motion control, large signal volume, and high requirements for system stability and accuracy. Therefore, the control system adopts a multi-CPU structure and distributed control method. This control system uses a centralized management and decentralized control method, which has high stability, working speed, and control performance. To further improve the real-time performance and stability of the system, a DSP multi-axis motion controller is used in the control system. The main computer is a PC host, and the control computer is an industrial control computer. The entire software system architecture is based on Windows NT and Windows 2000. To improve the overall measurement accuracy of the system, error compensation technology is used in the specific program. Long gratings are used to measure arc length to achieve the high-resolution requirement of angle measurement and realize high-precision positioning in space. The program development language is C++. The reason for choosing C++ is its efficiency, and more importantly, its good compatibility and portability with other programs. 3.1 Control System Structure The control system of this measuring machine is shown in Figure 2. It mainly consists of a main computer, a control computer, a DSP motion controller, a servo system, and a control box. This control method enhances the reliability and real-time performance of motion control. [align=center] Figure 2 Schematic diagram of the control system[/align] (1) The main computer adopts a PC host, which performs tasks such as controlling the control computer, human-machine interface, data communication, data processing, measurement element evaluation, database management, image acquisition and processing. The control computer adopts an industrial control computer, which controls the spatial posture of the mechanism, monitors the operating status of the mechanism in real time, and reads spatial coordinate values ​​and measurement data. (2) The control computer communicates with the main computer through the standard communication port RS232, receives control commands issued by the main computer, completes various motion commands and motion control, and returns the collected data and operating status to the main computer. (3) The motion controller is a 4-axis DSP motion controller from MEI (Motion Engineering, Inc.) in the United States. The DSP communicates bidirectionally with its data storage area, I/O ports, and other peripherals, such as analog inputs and outputs, timers, and position buffers, via its own address and data bus. As an intelligent motion controller for a PC, the main CPU directly accesses these addresses and data buses through its own 3-byte I/O address, exchanging data with the DSP through the external data storage area. This eliminates the need for the main CPU to transmit data character by character via the DSP registers, avoiding the conversion from ASCII to binary. It directly transmits binary numbers, significantly improving the communication speed between the CPU and DSP. Generally, the main CPU transmits only one data structure to the DSP frame by frame in each control cycle. During motion control, the main tasks of the DSP-based motion controller are: ① Servo control function: The motion controller provides PID and position servo loop filters, and also provides speed and acceleration feedforward control to reduce trajectory errors in the servo system; ② Motion control function: It can perform linear and circular interpolation, and automatically complete trapezoidal or S-curve acceleration and deceleration control; ③ Zero-position and limit detection; ④ Real-time motion status monitoring. (4) The servo system is used to realize the position servo control and spindle speed servo control of the system. This system adopts a dual-loop control mode (full closed loop) of speed inner loop and position outer loop. The structure of the single-axis servo control system is shown in Figure 3. Its working principle is as follows: the position signal (encoder signal) is subdivided and shaped and then sent to the counter to obtain the actual spatial coordinate value. The DSP compares the actual coordinate value with the command coordinate value (the set coordinate value is calculated by interpolation) to obtain the position error. The DSP motion controller substitutes the position error into the PID regulator to obtain the control voltage, and sends the control voltage to the servo driver through the analog channel. The servo driver controls the motor to run, thus forming the external position loop. The speed regulation loop receives the motor encoder signal from the servo driver to control the speed. This forms the dual-loop control mode of this system. [align=center] Figure 3 Single-axis servo system[/align] (5) The control box provides the operator with a convenient on-site operation front panel, allowing the user to operate the measuring machine at close range and realize certain specific functions. The control box adopts a smart front-end with a single-chip microcomputer as the core, and communicates with the control computer through a standard interface to realize the on-site control of the main body of the measuring machine and display the information sent by the control computer in real time. (6) The readings of the four axes of X, Y, Z and W of the measurement system are all read by the control computer and transmitted to the main computer. The camera probe has relative independence. In order to facilitate its development and connection with the entire measurement system software, motion control and image acquisition are separated: motion control is realized by the control computer, and image acquisition is directly realized by the main computer. The motion control and data acquisition of the laser probe are both completed by the control computer. 3.2 Control System Software 3.2.1 Trajectory Planning and Measurement Control Software System [align=center] Figure 4 Motion Control Software Flow [/align] The trajectory planning and measurement control software is the core of the system. It is responsible for human-computer interaction, controlling the measuring machine to perform measurement according to the measurement path, safety control, data acquisition and management. The flow of the trajectory planning and measurement control software is shown in Figure 4, which mainly includes the following functional modules: (1) Communication module. Responsible for managing the communication between the control computer and the main computer and the control box. Among them, the main computer uses serial port 1 and the control box uses serial port 2. When an instruction is received, an interrupt is generated and the instruction flag is set. The main program determines whether there is an instruction by detecting the flag. (2) Main measurement module. It mainly completes the measurement path planning and implements special measurement functions, including the measurement control of the inner sphere, the measurement control of the outer sphere, the measurement control of the cylinder, the measurement control of the plane, the measurement control of surface defects, etc., the calibration measurement control of system parameters, and is also responsible for the point-to-point movement of the control mechanism. (3) Initialization module. It mainly initializes the system parameters, flag bits, communication ports and measurement system. (4) Auxiliary function module. Measuring machine zeroing, reset, workpiece deviation adjustment, single-axis movement, three-axis linkage and follow-up, etc. (5) Motion status monitoring module. It mainly monitors the measuring machine hardware and software limits, DSP motion controller initialization, system parameters and motor running status, etc. Once an error is detected, the current measurement is stopped, and an alarm message and error reason are sent to the main computer for user adjustment, ensuring the safety of the measuring machine operation. (6) Data acquisition module. It is mainly responsible for the initialization of the data acquisition card and the real-time acquisition and processing of data from the probe and joint encoder. 3.2.2 System Monitoring Software In order to ensure the safety and reliability of the measurement system, motion monitoring is necessary. The motion monitoring process is shown in Figure 5, which mainly includes the monitoring of the main computer, the control box stop command, and the motion status monitoring of the measuring machine. The function of the measuring machine motion status monitoring is mainly to monitor the measuring machine hardware and software limits, DSP motion controller initialization, system parameters, and motor running status. Once an error is detected, the current measurement is stopped, and an alarm message and the cause of the error are sent to the main computer for user adjustment. Motion monitoring ensures the safety of the measuring machine operation and is a very important module in motion control. 4 System Calibration 4.1 System Calibration [align=center] Figure 5 Motion Monitoring Process[/align] In order to achieve the coordinates of the end effector pose required by the control trajectory, this measurement system uses a dedicated physical reference as a standard part and adopts the relative reference self-calibration method to calibrate the system parameters. 4.2 System Testing The robot measurement and control system was successfully applied to the non-destructive measurement of the geometric dimensions and surface morphology of a spherical shell workpiece. Table 1 shows the measurement results for a certain hemisphere. The data in the table demonstrates that the measurement and control system has high measurement accuracy and reliable stability. [align=center]Table 1[/align] 5 Conclusion This paper introduces a novel measurement and control system that combines robotics, non-contact measurement technology, and master-slave control to achieve geometric measurement and defect identification. The system itself possesses certain advanced features and broad application prospects. Testing has shown that the measurement and control system exhibits high measurement accuracy and reliable stability. Further research and experiments are underway to improve the system's stability and measurement accuracy.