(2) Secondary CPU structure, master-slave control mode
The first-level CPU is the host, which functions as system management, robot language compilation and human-machine interface. It also uses its computing power to complete coordinate transformation and track interpolation, and periodically sends the operation result to the common memory as an increment of joint motion. For secondary CPU reading; secondary CPU completes all joint position digital control. There is basically no connection between the two CPU buses of this type of system, and only the exchange of data through the common memory is a loosely coupled relationship. It is very difficult to further disperse functions with more CPUs. The computer system of the Motoman robot (5-joint, DC motor drive) produced in Japan in the 1970s belongs to this master-slave structure.
(3) Multi-CPU structure, distributed control mode
At present, the secondary distributed structure of the upper and lower machines is generally adopted, and the upper computer is responsible for the entire system management, kinematics calculation, and trajectory planning. The lower computer consists of multiple CPUs, each of which controls an articulation. These CPUs are connected to the master by tight coupling in the form of a bus. The controller's operating speed and control performance of this structure are significantly improved. However, the features shared by these multiple CPU systems are functional distributed structures that are specific to the problem, that is, each processor assumes a fixed task. Most commercial robot controllers in the world currently have this structure.
The position control portion of the controller computer control system employs digital position control almost without exception.
The above types of controllers all use a serial machine to calculate the robot control algorithm. They have a common weakness: heavy computing burden and poor real-time performance. Therefore, offline planning and feedforward compensation decoupling are mostly used to reduce the computational burden in real-time control. When the robot is disturbed during operation, its performance will be affected, and it is more difficult to guarantee the accuracy index required in high-speed motion.
Due to the complexity of robot control algorithms and the improvement of robot control performance, many scholars have made efforts to reduce the amount of computation from modeling, algorithms, etc., but it is still difficult to meet the requirements of real-time computing on serial structure controllers. Therefore, a solution must be sought from the controller itself. One of the methods is to use high-end microcomputer or minicomputer; the other method is to use multi-processor for parallel computing to improve the computing power of the controller.
2.2 Parallel processing structure
Parallel processing technology is an important and effective means to improve the speed of computing, and can meet the real-time requirements of robot control. From the literature point of view, regarding the parallel processing technology of robot controllers, people have studied more parallel algorithms and implementations of robot kinematics and dynamics. In 1982, JYSLuh [3] first proposed the parallel processing of robot dynamics. This is because the dynamic equation of the articulated robot is a set of nonlinearly coupled second-order differential equations. The calculation is very complicated. Improving the computational speed of the robot dynamics algorithm also lays a foundation for implementing complex control algorithms such as computational moment method, nonlinear feedforward method, and adaptive control method. One way to develop parallel algorithms is to transform the serial algorithm, parallelize it, and then map the algorithm to a parallel structure. Generally, there are two ways. First, considering the given parallel processor structure, the parallelism of the algorithm is developed according to the computational model supported by the processor structure. Second, the parallelism of the algorithm is first developed, and then the parallel processing of the algorithm is designed. Structure to achieve optimal parallel efficiency.
The computer system of the robot controller constructing the parallel processing structure generally adopts the following methods:
(1) Development of VLSI for robot control [4, 5]
The design-specific VLSI can make full use of the parallelism of the robot control algorithm, and rely on the parallel architecture in the chip to easily solve the large number of calculations in the robot control algorithm, which can greatly improve the calculation speed of the kinematics and dynamic equations. However, since the chip is designed according to a specific algorithm, when the algorithm changes, the chip cannot be used, so the controller constructed in this way is not universal, and is not conducive to system maintenance and development.
(2) Using a chip computer with parallel processing capability (such as Transputer, DSP, etc.) to form a parallel processing network
Transputer is a chip-type computer for parallel processing developed and produced by Inmos, UK. Using the link pair of the 4-bit serial communication of the Transputer chip, it is easy to construct different topologies, and the Transputer has extremely strong computing power. With the Transputer parallel processor, various robot parallel processors, such as pipeline type and tree type, are constructed. In [6], the Transputer network is used to realize the inverse kinematics calculation, while the literature [7] aims to realize the feedforward compensation and the calculation of the torque based on the fixed model.
With the increasing speed of digital signal chips, high-speed digital signal processors (DSPs) are widely used in various aspects of information processing. DSP is known for its extremely fast digital computing speed and is easy to form a parallel processing network [8]. Literature [9] introduced a DSP-based robot controller that uses a parallel/pipeline design to improve controller performance.
Previous Next
Boat Engine Machine,Outboard Motor Machine,Outboard Engine Machine,Outboard Engine
Shaoxing Haotuo Machinery co.,ltd. , https://www.hotochainsaws.com