The basic components of an industrial robot can be seen in Figures 3 and 4. Figure 3 shows an industrial robot driven by an electric motor, and Figure 4 shows an industrial robot driven by a hydraulic motor. Welding robots basically belong to these two categories. Arc welding robots mostly use electric motor drives because the welding torch generally weighs less than 10 kg. Spot welding robots, however, have welding clamps that weigh over 35 kg. Some use hydraulic drives because they have a greater gripping capacity, but most spot welding robots still use high-power servo motors due to their lower cost and more compact system. An industrial robot consists of four basic parts: a manipulator, a controller, a driver, and a teach pendant. For electric motor driven robots, the controller and driver are generally housed in a control box, while for hydraulically driven robots, the hydraulic drive source is a separate component. These are briefly described below:
(1) Robotic Arm: Also known as a manipulator, the robotic arm is the manipulator part of a robot. It directly drives the end effector (such as a welding torch or spot welding clamp) to achieve various movements and operations. Its structural forms are diverse, entirely determined by the task requirements. The goals pursued are high precision, high speed, high flexibility, large workspace, and modularity. Currently, the main structural forms of industrial robot manipulators are as follows:
The machine tool type robotic arm has a structure similar to a machine tool. Its three spatial movements (x, y, z) are linear motions, while the end effector's posture is determined by rotational motion, as shown in Figure 5. The advantages of this type of robotic arm are its simple kinematic model and easily improved control precision; the disadvantages are its large size, large footprint, and small workspace. This type is commonly used in simple and specialized welding robots.
Figure 3. Electric motor driven industrial robot
Figure 4. Hydraulic press driving industrial robot
The structure of a fully articulated robotic arm resembles the human waist and hand, with its position and posture entirely determined by rotational motion. Figure 6 shows a normally aligned fully articulated robotic arm, and Figure 7 shows an offset fully articulated robotic arm. This is the most common structural form of industrial robot robotic arms. Its advantages include compact structure, high flexibility, small footprint, and large workspace. Its disadvantages are high precision and difficulty in control. The difference between the offset and normally aligned types is that the wrist joint is located on the outside of the forearm or within the forearm's range of motion, but its kinematic model is more complex. Currently, welding robots mainly use fully articulated robotic arms.
Figure 5 Machine tool type robot arm
Planar articulated robotic arms are characterized by linear motion for vertical movement and rotational motion for all other movements. This structure offers high rigidity in the vertical direction and excellent flexibility in the horizontal direction, making it well-suited for assembly operations primarily involving insertion. Therefore, it is widely used in assembly robots and is also known as a SCARA type robotic arm, as shown in Figure 8. While the specific structures of robotic arms vary, they are all composed of commonly used mechanisms. The internal mechanism of the PUMA robotic arm from the United States will be briefly described using Figure 9 as an example. It consists of four parts: a base, an upper arm, a forearm, and a wrist. There are three rotational joints between the base and the upper arm, the upper arm and the forearm, and the forearm and the wrist to ensure arbitrary positioning within the workspace. The wrist also has three rotational joints: wrist rotation, wrist flexion, and wrist swing, to achieve arbitrary spatial posture of the end effector. The end of the wrist is a flange for connecting to the end effector. Each joint is driven by a servo motor. PUMA robotic arms use gear reduction and lever transmission, but different manufacturers use different mechanisms. Common reduction mechanisms include four types: gears, harmonic reducers, ball screws, and worm gears. Transmission methods include lever transmission, chain transmission, and gear transmission. The key technology is to ensure backlash-free bidirectional transmission (i.e., no backlash in both forward and reverse transmission), which is the mechanical guarantee of robot precision. High efficiency and a compact mechanism are also required.
(2) Drivers Since most welding robots use servo motors, this section will only introduce this type of driver. The motor drivers currently used in industrial robots can be divided into four categories:
Stepper motor drivers use stepper motors, especially microstepping stepper motors, as the drive source. Since these systems are generally open-loop controlled, they are mostly used in economical industrial machines with lower cost.
The DC servo motor system utilizes a DC servo motor system, which can achieve three closed-loop control of position, speed, and acceleration. It offers high precision, a wide speed range, and excellent dynamic performance. Therefore, it is currently the main drive method for industrial robots.
The AC motor servo system driver utilizes an AC servo motor system, which possesses all the advantages of a DC servo system. Furthermore, it eliminates the need for commutated carbon brushes, reducing the need for regular brush replacement and significantly extending the robot's maintenance cycle. Therefore, it is being widely adopted in robotics.
Direct drive motors are the latest development in robot actuators. Direct drive motors have speed ratios greater than 10,000, providing stable power and high dynamic performance even at low speeds. They can directly drive joints in robotic arms, eliminating reduction gears, simplifying the mechanism, and improving efficiency. This represents the future direction of robot actuator development. The Adapt robot from the United States is a direct drive robot. Industrial robot actuators are typically arranged with one actuator per joint. The basic components of an actuator include: a power supply, a power amplifier board, a servo control board, a motor, a protractor, a tachometer, and a brake. Its function is not only to provide sufficient power to drive the joints of the robotic arm but also to achieve rapid and frequent starts and stops, precise positioning, and movement. Therefore, position closed-loop, velocity closed-loop, and acceleration closed-loop mechanisms are necessary. A current closed-loop mechanism is also required to protect the motor and circuitry. To meet the frequent starts and stops and high dynamic performance requirements of robots, low-inertia motors are generally used. Therefore, robot actuators are high-performance drive systems.
To achieve the three motion closed loops mentioned above, high-precision angle and speed sensors are installed in the robotic arm actuator. Speed sensors typically use tachogenerators, while angle sensors generally use precision potentiometers or photoelectric encoders, especially photoelectric encoders. Figure 10 shows its schematic diagram. The photoelectric encoder is mounted coaxially with the motor. When the motor rotates, the encoder with finely etched grooves rotates at the same speed, and the light beam from the fixed light source directed at the phototubes intermittently, thus outputting electrical pulses. The actual encoder outputs two pulses. Because two pairs of phototubes are arranged within the encoder, there is a certain angular difference between them, resulting in a fixed phase difference between the two pulses. When the motor rotates in both directions, the phase difference of the output pulses is different, thus allowing the determination of the motor's rotation direction. The machine outputs more than one pulse.
(3) Controller The robot controller is the core component of the robot. It performs all information processing and motion control of the robot arm.
Figure 11 is a schematic diagram of the controller's working principle.
Most industrial robot controllers adopt a two-level computer structure. The area within the dashed box represents the first-level computer, whose task is planning and management. When the robot is in teach mode, it receives the position and posture information, motion parameters, and process parameters of each teach point from the teach system, and converts the teach (joint) coordinate values of each point into Cartesian coordinate values through calculation and stores them in the computer memory.
Figure 10 Schematic diagram of photoelectric encoder
Figure 11 Controller Working Principle Diagram
When the robot reproduces its state, it retrieves its position and orientation coordinates point by point from memory and performs circular or linear interpolation calculations at certain time intervals (also known as sampling periods) to calculate the position and orientation coordinates of each interpolation point. This is path planning generation. Then, it converts the position and orientation coordinates of each interpolation point into joint coordinates point by point and distributes them to each joint. This is the entire planning process of the first-level computer. The second-level computer is the execution computer, and its task is to perform closed-loop control of the servo motors. After receiving the expected position and orientation of each joint for the next step from the first-level computer, it performs another uniform subdivision to make the motion trajectory smoother. Then, it sends the expected value of the next step for each joint point by point to the drive motor, while detecting the photoelectric encoder signal, until it accurately reaches its position.
All of the above are real-time processes, and a large number of calculations must be completed during the control process. Taking the PUMA robot controller as an example, the sampling period of the first-level computer is 28ms, that is, every 28ms, it sends the joint coordinates of each joint to the second-level computer for the next position and orientation. The second-level computer then divides each joint value into 30 equal steps and sends the joint coordinate value to each joint every 0.875ms.
(4) Teach pendant: The teach pendant is the human-computer interaction interface for teaching robots. Currently, there are three ways for humans to teach robots:
Hands-on teaching, also known as full-process teaching, involves a person holding the end effector of the robot's robotic arm and guiding the robot to perform the actual task. During this process, the robot controller's computer records the position and orientation values of each joint point by point without performing coordinate transformations. When reproducing the result, these values are retrieved point by point. This teaching method requires a large amount of computer memory, and due to the resistance of the mechanism, the teaching accuracy cannot be very high. Currently, it is only used in painting and spraying robots.
Teach pendant teaching is the most common method of robot teaching, where a human teaches the robot by operating it through a teach pendant. Currently, welding robots all use this method.
Offline programming teaching means that the robot can be programmed on a computer based on drawings without human intervention, and the programming data can then be transmitted to the robot controller. It has advantages such as not taking up robot time, facilitating optimization, and being safer, making it the future direction of development.
Figure 12 shows the teach pendant of the ESAB welding robot. It connects to the control box via a cable, allowing the operator to hold the teach pendant at the most intuitive position near the workpiece for teaching. The teach pendant itself is a dedicated computer that continuously scans the functions, numeric keys, and joysticks on the pendant and sends information and commands to the controller. While teach pendants vary between manufacturers, their ultimate goal is to facilitate the operator's use.
Figure 12. Teach pendant for welding robot
The buttons on the teach pendant mainly fall into three categories:
Teach function keys, such as teach/playback, save, delete, modify, check, return to zero, linear interpolation, circular interpolation, etc., are used for teach programming.
Motion function keys such as blade direction movement, y-axis movement, z-axis movement, forward/reverse movement, and joint rotation from 1 to 6 are used for teaching and manipulating the robot.
Parameter setting keys include settings for each axis speed, welding parameters, and oscillation parameters.
