I. Functions to be achieved by the industrial robot control system
The robot control system is an important component of a robot, used to control the manipulator to complete specific work tasks. Its basic functions are as follows:
1. Memory function: Stores work sequence, movement path, movement mode, movement speed and information related to production process.
2. Teaching Functions: Offline programming, online teaching, and indirect teaching. Online teaching includes two types: teach pendant and guided teaching.
3. Communication with peripheral devices: Input and output interfaces, communication interface, network interface, synchronization interface.
4. Coordinate setting function: There are four coordinate systems: joint, absolute, tool, and user-defined.
5. Human-machine interface: teach pendant, operation panel, display screen.
6. Sensor interfaces: position detection, vision, touch, force, etc.
7. Position servo function: multi-axis linkage of the robot, motion control, speed and acceleration control, dynamic compensation, etc.
8. Fault Diagnosis and Safety Protection Functions: System status monitoring during operation, safety protection under fault conditions, and fault self-diagnosis.
II. Composition of Industrial Robot Control System
1. Control Computer: The dispatching and command mechanism of the control system. It is generally a microcomputer or microprocessor, such as a 32-bit or 64-bit CPU, like the Pentium series CPU and other types of CPUs.
2. Teach pendant: It is responsible for setting the working trajectory and parameters of the teach pendant robot, as well as all human-computer interaction operations. It has its own independent CPU and storage unit, and communicates with the host computer via serial communication.
3. Control panel: Consists of various operation buttons and status indicator lights, and only performs basic functions.
4. Hard disk and floppy disk storage: peripheral storage for storing the robot's working program.
5. Digital and analog inputs and outputs: Inputs or outputs of various status and control commands.
6. Printer interface: Records various information that needs to be output.
7. Sensor interface: Used for automatic information detection to achieve compliant robot control, generally force, touch and vision sensors.
8. Axis Controller: Controls the position, speed, and acceleration of each joint of the robot.
9. Auxiliary equipment control: Control of auxiliary equipment used in conjunction with the robot, such as gripper positioners.
10. Communication interface: Enables information exchange between the robot and other devices, generally including serial interface, parallel interface, etc.
11. Network Interface
1) Ethernet interface: Enables direct communication between multiple robots or a single robot to a PC via Ethernet, with a data transmission rate of up to 10 Mbit/s. After application programming using Windows library functions on the PC, it supports TCP/IP communication protocol and loads data and programs into each robot controller through the Ethernet interface.
2) Fieldbus interface: Supports a variety of popular fieldbus specifications, such as Devicenet, ABRemoteI/O, Interbus-S, profibus-DP, M-NET, etc.
III. Classification of Industrial Robot Control Systems
1. Program control system: By applying a certain control action to each degree of freedom, the robot can achieve the required spatial trajectory.
2. Adaptive Control System: When external conditions change, to ensure the required quality or to improve control quality automatically with the accumulation of experience, the process is based on observing the state of the operator and servo errors, and then adjusting the parameters of the nonlinear model until the error disappears. The structure and parameters of this system can automatically change with time and conditions.
3. Artificial intelligence system: It is impossible to program the movement in advance. Instead, it requires the system to determine the control action in real time based on the surrounding state information obtained during the movement.
4. Point-to-point: This requires the robot to accurately control the pose of the end effector, regardless of the path.
5. Trajectory-based: This requires the robot to move along the taught trajectory and speed.
6. Control Bus: International Standard Bus Control System. The control system uses international standard buses, such as VME, MULTI-bus, STD-bus, and PC-bus.
7. Custom bus control system: The bus used by the manufacturer is defined as the control system bus.
8. Programming method: Physical setting programming system. The operator sets fixed limit switches to realize the start and stop program operation, which can only be used for simple picking and placing operations.
9. Online programming: A programming method that completes the memorization process of operational information through human instruction, including direct instruction, simulated instruction, and teach pendant instruction.
10. Offline Programming: Instead of directly teaching the robot in the actual operation environment, it demonstrates programming outside of the actual operation environment.
IV. Robot Control System Structure
Robot control systems can be classified into three categories according to their control methods.
1) Centralized Control System: All control functions are implemented by a single computer. It has a simple structure and low cost, but poor real-time performance and is difficult to expand. This structure was often used in early robots. Its block diagram is shown in Figure 2.
PC-based centralized control systems fully utilize the openness of PC resources, achieving excellent openness: various control cards and sensor devices can be integrated into the control system through standard PCI slots or standard serial and parallel ports. The advantages of centralized control systems are: lower hardware costs, easier information acquisition and analysis, easier achievement of optimal system control, better overall integrity and coordination, and convenient hardware expansion. However, their disadvantages are also obvious: lack of system control flexibility, a high risk of concentrated control failures, and a wide-ranging and severe impact from any malfunction; due to the high real-time requirements of industrial robots, large-scale data calculations can reduce system real-time performance, and the system's responsiveness to multiple tasks may conflict with real-time requirements; furthermore, complex system wiring can reduce system reliability.
2) Master-Slave Control System: This system employs a two-tiered master and slave processor architecture to implement all control functions. The master CPU handles management, coordinate transformation, trajectory generation, and system self-diagnosis; the slave CPU controls the motion of all joints. Its block diagram is shown in Figure 3. The master-slave control system offers good real-time performance and is suitable for high-precision, high-speed control, but its system scalability is poor, and maintenance is difficult.
3) Distributed Control System: This system divides control into several modules based on its nature and method. Each module has different control tasks and strategies. The modules can have a master-slave relationship or an equal relationship. This approach offers good real-time performance, facilitates high-speed and high-precision control, is easily expandable, and enables intelligent control. It is currently a popular method, and its control block diagram is shown in Figure 4.
The main idea behind DCS is "distributed control, centralized management," meaning the system can comprehensively coordinate and allocate its overall goals and tasks, and complete control tasks through the coordinated work of subsystems. The entire system is distributed in terms of function, logic, and physical aspects; therefore, DCS systems are also called distributed control systems or decentralized control systems. In this structure, subsystems consist of controllers and different controlled objects or devices, and the subsystems communicate with each other through networks. Distributed control architecture provides an open, real-time, and precise robot control system. Two-level control is often used in distributed systems.
A two-level distributed control system typically consists of a host computer, slave computers, and a network. The host computer can perform different trajectory planning and control algorithms, while the slave computers perform interpolation subdivision, control optimization, and other research and implementation. The host computer and slave computers coordinate with each other through a communication bus, which can be in the form of RS-232, RS-485, EEE-488, or USB bus, etc.
The development of Ethernet and fieldbus technologies has provided robots with faster, more stable, and more efficient communication services. Fieldbus, in particular, is applied in production environments to enable bidirectional, multi-node digital communication between microcomputer-based measurement and control devices, thus forming a new type of network-integrated, fully distributed control system—the Fieldbus Control System (FCS). In factory production networks, devices that can be connected via fieldbus are collectively referred to as "field devices/instruments." From a systems theory perspective, industrial robots, as one of the production devices in a factory, can also be categorized as field devices. Introducing fieldbus technology into robot systems further facilitates the integration of robots into industrial production environments.
The advantages of distributed control systems are: good system flexibility, reduced risk of the control system, and the use of multi-processor distributed control, which is conducive to the parallel execution of system functions, improves system processing efficiency, and shortens response time.
For industrial robots with multiple degrees of freedom, centralized control handles the coupling relationships between individual control axes very well and can be easily compensated for. However, when the number of axes increases to the point that the control algorithm becomes very complex, its control performance deteriorates. Moreover, when the number of axes or the control algorithm in the system becomes very complex, it may lead to a redesign of the system. In contrast, in a distributed architecture, each motion axis is handled by a controller, which means that the system has fewer inter-axis couplings and higher system reconfigurability.
