Skip to main content
main-content

Über dieses Buch

The first book of the new, textbook series, entitled Applied Dynamics of Manipulation Robots: Modelling, Analysis and Examples, by M. Vukobratovic, published by Springer-Verlag (1989) was devoted to the problems of dynamic models and dynamic analysis of robots. The present book, the second in the series, is concerned with the problems of the robot control. In conceiving this textbook, several dillemas arouse. The main issue was the question on what should be incorporated in a textbook on such a complex subject. Namely, the robot control comprises a wide range of topics related to various aspects of robotics, starting from the syn­ thesis of the lowest, executive, control level, through the synthesis of trajectories (which is mainly related to kinematic models of robots) and various algorithms for solving the problem of task and robot moti­ on planning (including the solving of the problems by the methods of artificial intelligence) to the aspects of processing the data obtai­ ned from sensors. The robot control is closely related to the robot pro­ gramming (i. e. the development of highly-specialized programming lan­ guages for robot programming). Besides, numerous aspects of the con­ trol realization should be included here. It is obvious that all these aspects of control cannot be treated in detail in the frame of a text­ book.

Inhaltsverzeichnis

Frontmatter

Chapter 1. Concepts of Manipulation Robot Control

Abstract
The tasks that are nowdays assigned to robots are becoming more varied and more complex. More and more often robotic units are becoming parts of flexible technological cells, lines and intelligent technological systems. In view of these facts, the organization of robot control should be based upon the principle of control hierarchy.
Miomir Vukobratović, Dragan Stokić

Chapter 2. Kinematic Control Level

Abstract
As pointed out in Section 1.2, the tasks for which robots are applied in industry and other fields are of very diverse complexity. The more complex the task a robot has to perform and the more strict the requirements for its performing, the more complex should be the control system of the robot. The complexity of the robot control also depends, as it will be shown in Chapter 4, on the robot’s mechanical structure, i.e. on the extent and mode the motion of one joint (mechanical degree of freedom) of the robot influences the other joint. Because of that, different “types” of control systems appear in practice, which in different ways solve the problems at both tactical and executive level and enable accomplishment of tasks of different class. As will be shown below, the “types” of control are most often related to different classes of tasks in robotics, which, on the other hand, have different requirements toward the executive control level. This chapter deals with the problems concerning the tactical control level, while the subject of the coming chapters will be the synthesis of control at the executive level.
Miomir Vukobratović, Dragan Stokić

Chapter 3. Synthesis of Servo Systems for Robot Control

Abstract
In the previous chapter we have considered problems concerning the synthesis of tactical control level. In the text to follow we shall consider the synthesis of the executive control level. As we have already explained the executive control level has to ensure implementation of trajectories (or, only positions) of joint coordinates of a robot. These trajectories are computed at the tactical control level. The implementation of the trajectories (positions) directly involves dynamic behaviour of the robot. Due to this, first we shall briefly present dynamic model of the robot system. The dynamic model of the robot “consists” of dynamic model of the mechanism and models of actuators which drive the joints of the mechanism. Next, we shall consider synthesis of servo system around each joint (actuator) of the robot. In practice, the executive control level is often implemented in the form of independent controllers for each joint of robot. As we shall explain, robot is a complex system with strong interactions between the motions of its joints. However, the independent control of joints is the simplest control law, and so it is the most appropriate approach from the standpoint of control implementation. In this chapter we shall present various methods for synthesis of a local controller around each joint, and in the next chapters we shall discuss certain disadvantages of this approach to control of robots. We shall also consider the methods for improvement of this control in order to meet the requirements which are imposed before the robots in industry.
Miomir Vukobratović, Dragan Stokić

Chapter 4. Control of Simultaneous Motions of Robot Joints

Abstract
In the previous chapter we have considered the synthesis of local servo system for one single joint of the robot. We have assumed that only the considered joint can move, while all the other joints of the robot are kept locked. The robots of the first generation often have implemented such a solution of the control system. In this case the joints of the robot are moving successively one by one until the goal position of the robot hand is reached. During the movement of one joint (until it reaches its desired position) the rest of the joints are kept locked (i.e. they are fixed). The local servos, synthesized in Chapter 3, satisfy such robotic tasks which require the successive movements of the robot joints.
Miomir Vukobratović, Dragan Stokić

Chapter 5. Synthesis of Robot Dynamic Control

Abstract
As we have explained in previous chapter, during simultaneous motions of robot joints there appear dynamic forces (torques) which act around the joints axes and affect the performances of local servo systems. To ensure accurate tracking of nominal trajectories (imposed by the higher tactical control level) we have to compensate for the effects of these dynamic forces. If the executive control level has to ensure only positioning of the robot hand in various positions in work space, or to ensure tracking of “slow” trajectories of the robot hand, then the effects of these dynamic forces might be relatively weak and they might be ignored. Therefore, the local servos might accomplish such control tasks. In the previous chapter we have shown how we can examine whether or not the local servos could ensure accurate tracking of desired trajectories. If local servo systems can not “overcome” the effects of dynamic forces, we must introduce additional control loops which take into account these dynamic forces. Such control law, which accounts for dynamic characteristics of robotic systems is called dynamic control of robots.
Miomir Vukobratović, Dragan Stokić

Chapter 6. Variable Parameters and Concept of Adaptive Robot Control

Abstract
In the proceeding considerations we assumed the robot models and their parameters are being known and determined in advance. Such an assumption holds for the majority of robot parameters, such as masses and inertia moments, link lengths, positions of joints axes, inertia moments of motors, etc. However, some robot parameters change during the work, and they are not always known in advance: these parameters are for example some of the motors parameters, coefficients of viscous and static friction, etc. A common feature of all these parameters is that they change very slowly, so that they can be considered to be quasi-stationary. Very often, their determination is rather difficult. The values of these parameters can be identified when the robot starts to work, and they should be checked from time to time and updated appropriately. Most often, these parameters do not affect significantly the functioning of the control system, and they may be considered as being known. A special group form the parameters which can undergo fast and drastic changes during the robot’s work, and which cannot be always known in advance. To this group belong, primarily, the parameters of the working object.
Miomir Vukobratović, Dragan Stokić

Chapter 7. Control of Constrained Motion of Robot

Abstract
In all the tasks we have considered up to now, the robot does not come into contact with the objects in the workspace, apart from those that are being transferred. However, this does not hold for the one of the most important industrial application of robots, namely, for the assembly of machine parts. In this case, the robot comes into contact with the objects in its environment and experiences actions of the external environmental forces. Similarly, in the processes like cutting, grinding, polishing and forging, the robot gripper has to act upon the given object by certain forces. These external forces acting on the robot gripper make the robot control much more complex. Hence, this chapter will be devoted to the synthesis of control for the robots involved in the realization of the tasks of this type. First our attention will be focused on the assembly process, as one of the most important and most delicate tasks in which the action of external forces is encountered. However, some general approaches to control of constrained motion of robots will be also presented.
Miomir Vukobratović, Dragan Stokić

Backmatter

Weitere Informationen