Keywords: Dongguan; industrial robot; serial communication; controller
1. Introduction <br />With the intersection, penetration, and integration of new technologies such as information technology, materials technology, and new energy technology with manufacturing technology, the manufacturing industry today has undergone many significant and profound changes compared to the past. Increasing complexity is a key characteristic of modern manufacturing, manifested not only in the manufacturing system but also in the products manufactured, the manufacturing processes, and the structure of enterprises. This complexity presents a good opportunity for the development of industrial robots, but also a challenge! Dongguan University of Technology is located in the Songshan Lake High-tech Development Zone of Dongguan. Dongguan, as a renowned modern manufacturing city, is at a critical stage of industrial transformation. If industrial robots can be applied to the transformation of production lines in various enterprises in a timely manner, there will undoubtedly be enormous potential and prospects.
This industrial robot is a precision assembly robot, driven and controlled by four-axis servo motors to achieve four-axis spatial linkage. Equipped with different toolkits, it can perform tasks such as handling, palletizing, and assembly. It features high speed, high precision, and good flexibility. Driven by AC servo motors, it consists of a manipulator (mechanical body), controller, servo drive system, and sensing devices.
(1) Manipulator: Through the application of modern design methods such as finite element analysis, modal analysis and simulation design, the robot manipulator has achieved optimized design.
(2) Controller: From initially using the GT series motion controller from Hong Kong Googol Technology Co., Ltd. to independently developing an ARM9-based controller, and realizing software servo and full digital control.
(3) Servo drive system: Fuji servo is used.
(4) Sensing devices: We are trying to use laser sensors, vision sensors and force sensors in the robot system to achieve automatic positioning of objects on the automated production line and precision assembly operations, which greatly improves the robot's performance and adaptability to the environment.
(5) Network communication: The robot controller has realized the connection with CANbus, Profibus bus and some networks, which has made the robot take a big step from the past independent application to networked application, and also made the robot develop from the past dedicated equipment to standardized equipment.
2. Data Communication Method <br />Since serial communication uses fewer transmission lines and can transmit over longer distances, which is beneficial for real-time control and management, this system adopts serial communication. In serial communication, data is transmitted bit by bit on a single data line, with each bit occupying a fixed time period. However, since the computer CPU and the interface transmit data in parallel, while the interface and peripherals transmit data serially, the serial interface must have a "receive shift register" and a "transmit shift register" to handle the conversion between serial and parallel data. The circuit capable of performing the "serial <-> parallel" conversion function is called a "universal asynchronous transceiver." While serial communication is significantly slower than parallel communication, its advantages of lower cost and longer communication distance are obvious.
Information transmission occupies only one communication line in one direction. This line serves as both a data line and a communication line. The protocol between the host computer and the robot is organized according to the following three levels.
(1) Physical layer: refers to the hardware used to perform this communication, including data lines and control lines. This system uses an RS232 serial port. One end of the robot has a 9-pin D-shaped connector, and the other end has an RJ-45 connector connected to the serial interface of the Terminal Server. Then, it is connected to the FMS LAN through the network interface of the Terminal Server, and finally realizes the connection with the robot control computer. The hardware platform is shown in Figure 1.
(2) Data Link Layer: This layer is a low-level communication protocol that uses special control characters for sending and receiving to ensure reliable information transmission. The information exchange between the industrial robot controller and the host computer consists of strings, which include the actual information to be exchanged and additional control characters. These additional characters are mandatory so that the receiving device can determine whether a complete message has been received. Sending a robot control command involves the following six stages: sending open phase, packet transmission, sending closed phase, receiving open phase, receiving unpacking, and receiving closed phase.
(3) Application Layer: This layer is a high-level communication protocol that defines the content and response of each type of information. After receiving a message, the robot controller must decode it and take corresponding actions. Software that can respond to information commands from the host computer is called the application-level protocol. In this system, the host computer mainly sends four types of commands: L0ADV, SAVEV, JwAIT, and START, to enable the microcomputer to start the robot and perform real-time monitoring.
3. Controller <br /> Effectively applying research results from other fields (such as image processing, sound recognition, optimal control, artificial intelligence, etc.) to the real-time operation of robot control systems is a challenging research task. Research on modular and standardized robot controllers with open structures is undoubtedly of great significance for improving robot performance and autonomy, and promoting the development of robot technology. A robot controller is a device that controls a robot to complete certain actions or tasks based on instructions and sensor information. It is the heart of the robot and determines its performance.
This system employs a serial control algorithm, with the robot's control algorithm processed by a serial machine. A two-tier distributed architecture (upper and lower level) is used. The upper level is responsible for overall system management, kinematic calculations, trajectory planning, etc. The lower level consists of multiple CPUs, each controlling one joint movement. These CPUs are tightly coupled to the main controller via a bus. This architecture significantly improves the controller's operating speed and control performance. However, these multi-CPU systems share the characteristic of using a functionally distributed architecture tailored to specific problems; each processor undertakes a fixed task, and the controller computer controls the position control portion of the system, employing digital position control. Initially, the hardware platform used a GT series motion controller from Googol Technology, capable of synchronously controlling four motion axes to achieve multi-axis coordinated motion. Its core consists of an ADSP2I81 digital signal processor and an FPGA, enabling high-performance control calculations. During the research process, the following limitations were encountered:
(1) Poor openness: It is limited to a closed structure of "dedicated computer, dedicated robot language, and dedicated microprocessor". The closed controller structure makes it have specific functions and adapt to specific environments, which makes it difficult to expand and improve the system;
(2) Poor software independence: The software structure and its logical structure depend on the processor hardware, making it difficult to port between different systems;
(3) Poor fault tolerance: Due to the inherent characteristics of data correlation, communication and synchronization in parallel computing, the fault tolerance of the controller deteriorates. A failure of one processor may lead to the paralysis of the entire system.
(4) Poor scalability: At present, the research on robot controllers focuses on the joint level, which is more common. For example, for air-to-ground missile guidance, locking and pre-deflection are two frequently used functions, which put forward different requirements on the form of servo system and the selection of rate sensor.
First, adding a velocity sensor to the seeker head could solve the problem, but this would increase system complexity and hinder engineering and miniaturization in product development. Second, calculations revealed that while the guidance optical axis needs to reproduce the relative locking, search, and radar motion signals to the missile body during its sail-mounted flight, the missile's angular velocity during sail-mounted flight is not very high, and the duration of the oscillation is not long. If the seeker head's locking quality meets the requirements, the drawbacks of using a rate gyroscope for feedback can be significantly reduced. Therefore, after comprehensive consideration, it was decided to continue using only a rate gyroscope as the internal loop velocity sensor to better achieve stable tracking of the target by the optical axis.
4. Simulation Verification
Taking the reproduction of a relatively inertial space-stable input by a servo system as an example, to further compare the two schemes, we used ADAMs and MATLAB software to perform joint simulations of the kinematics, dynamics, and control system of the target. The simulation conditions assumed that the target was stationary in space. In the middle loop channel, gyro velocity feedback and tachometer velocity feedback were used respectively to servo the target indication information under the condition of projectile oscillation. The magnitude of the pointing error (i.e., the angle between the optical axis and the line of sight) was recorded. The simulation results are shown in Figure 3.
The solid line represents the error when using gyroscope signals, and the dashed line represents the error when using tachometer signals. The results show that using gyroscope signals as feedback signals can effectively reduce the homing error of the system during projectile oscillation, a result that is largely consistent with the previous theoretical analysis.
5 Conclusion <br />This paper analyzes in detail the different types of input signals for the moving base servo system, and derives the sensor selection principle of using velocimeter feedback for inputs that are relatively stable relative to the base and rate gyroscope feedback for inputs that are relatively stable relative to inertial space. Taking the air-to-air missile seeker as an example, ADAMs and MATLAB joint digital simulation were carried out, and the simulation results verified the correctness of the analysis.
References:
[l] Zuo Zhe, Li Donghai, Dai Yaping, Song Yuejin. State compensation control of gyro-stabilized platform [J]. Acta Aeronautica Sinica, 2008, 29(1): 14147.
[2] Yang Pu, Li Qi. Design and implementation of control system for three-axis gyroscope stabilized platform [J]. Journal of Chinese Inertial Technology, 2007, 15(2): 176.
Click to download: Research on Industrial Robots for Production Lines
