Robotcontrol can be categorized into control in joint space and control in Cartesian space. For tandem multi-joint robots, joint space control is control for variables at each joint of the robot, and Cartesian space control is control for variables at the end of the robot. According to the different control quantities, robot control can be categorized into: position control, velocity control, acceleration control, force control, force-position hybrid control and vibration control.
According to the different operational tasks, the robot control can be divided into four control methods: point control, continuous trajectory control, force (torque) control and intelligent control. In this paper, the four control methods are introduced from the division of operational tasks.
1,Point position control mode (PTP)
Point control in the field of mechatronics and robotics industry and its wide range of applications, machinery manufacturing in the CNC machine tools for parts contour tracking, industrial robot finger end trajectory control and walking robot path tracking and so on are typical applications of point control systems.
In the control, the industrial robot is required to be able to move quickly and accurately between neighboring points, and there is no stipulation on the movement trajectory to reach the target point.
Positioning accuracy and the time required for movement are the two main technical indicators of this control method. Since this control method is easy to realize and does not require high positioning accuracy, it is often used in operations such as loading and unloading, handling, spot welding, and insertion of components on circuit boards, which only require that the position of the end-effector be maintained accurately at the target point. This approach is relatively simple, but it is quite difficult to achieve a positioning accuracy of 2 to 3 um.
Point control system is actually a position servo system, their basic structure and composition is basically the same, only focusing on different things, their control complexity is also different; according to the feedback method, can be divided into closed-loop system, semi-closed-loop system and open-loop system.
2,Continuous trajectory control mode (CP)
PTP point control, the beginning and end speed is 0, during which there can be a variety of speed planning methods.
CP control is a continuous control of the industrial robot end-effector position in the operating space, the speed of the middle point is not 0, coherent movement, through the speed look ahead to get the speed size of each point. Generally continuous trajectory control mainly uses the method of speed look ahead: forward speed limit, corner speed limit, backward speed limit, maximum speed limit, contour error speed limit.
This control method requires it to move strictly according to the predetermined trajectory and speed within a certain accuracy range, and the speed is controllable, the trajectory is smooth, and the movement is smooth in order to complete the operational tasks.
The joints of the industrial robot continuously and synchronously carry out the corresponding movement, and its end-effector can form a continuous trajectory. The main technical indicators of this control method is the trajectory tracking accuracy and smoothness of the end-effector posture of industrial robots, usually arc welding, painting, deburring and testing operations robots use this control method.
3,Force (torque) control method
With the continuous broadening of the application boundary of the robot, vision alone can not meet the complexity of the actual application, it is necessary to introduce force / torque control output, or force / torque as a closed-loop feedback into the control.
In assembly, gripping and placing objects, etc., in addition to the requirements of accurate positioning, but also requires the use of force or torque must be appropriate, it is necessary to use (torque) servo mode. The principle of this type of control is basically the same as that of position servo control, except that the input and feedback are not position signals but force (torque) signals, so there must be a force (torque) sensor in the system. Sometimes also use the proximity, sliding and other sensing functions for adaptive control.
Because the contact between the robotic arm and the work surface is often an unknown complex surface, so this force/torque sensing should also have multi-dimensional capabilities.
4, intelligent control method
Intelligent control of the robot is a control mode with intelligent information processing and intelligent information feedback as well as intelligent control decision-making, obtaining knowledge of the surrounding environment through sensors (such as cameras, image sensors, ultrasonic transmitters, lasers, conductive rubber, piezoelectric components, pneumatic components, travel switches, and other electro-mechanical components) and making corresponding decisions based on its own internal knowledge base.
The development of intelligent control technology depends on the rapid development of artificial intelligence such as artificial neural networks, genetic algorithms, genetic algorithms, expert systems and so on. In recent years, intelligent control technology has progressed significantly, and the fuzzy control theory and artificial neural network theory, as well as the fusion of the two, have greatly improved the speed and accuracy of the robot. The main applications such as multi-joint robot tracking control, lunar robot control, weeding robot control, cooking robot control and so on.
Robot intelligent control can be further subdivided into: fuzzy control, adaptive control, optimal control, neural network control, fuzzy neural network control, expert control and so on.
With the addition of intelligent control technology, industrial robots are truly intelligent, but it is also the most difficult to realize, on the algorithm, components since the serious.
At present, industrial robots, in most cases, are still at the more bottom of the spatial localization control stage, there is not much intelligent content, there is still a long way to go from intelligence. Therefore, China's robotics experts from the application environment, the robot is divided into two categories, namely, industrial robots and intelligent robots.