• Không có kết quả nào được tìm thấy

and Control in Industrial and Service Settings

N/A
N/A
Nguyễn Gia Hào

Academic year: 2023

Chia sẻ "and Control in Industrial and Service Settings"

Copied!
359
0
0

Loading.... (view fulltext now)

Văn bản

The views expressed in this series are those of the authors, but not necessarily those of IGI Global. The target group consists of multidisciplinary participants from society, business, academia and ... the government. The book would be suitable as a good reference for college students and professors. Today there is an increasing number of robots in different environments. Intelligent robots have emerged .. in agriculture, industry and services.It involves fields such as food production, material processing and intelligent navigation. This book mainly focuses on robotics, automation and control.

This book would provide a forum for innovative research and development findings in advanced robotics, automation and control. It aims to promote an international knowledgeexchangecommunity with.. amultidisciplinaryparticipationofresearchers,practitionersandacademicswithinsight. real-life problems towards smarter production and people-centred services. By spreading. latest developments in robotics, automation, control of innovation and transformation based on current and/or .. emerging technology and the requirements of the market, this book covers both theoretical perspectives s. and practical approaches to research and development in smart manufacturing and people-centred services. Chapter 3 presents a new sensor-based Path Planner, which provides a fast local or global moving plan.

This is because the end effector of each robot must maintain contact with the object. Consequently, it. To achieve this, they are constrained to improve their way of industrial management, both at the strategic level, and adapt technologically. market trends. At the strategic level it leads.

TeleoperacijarobotovmanipulatorjevssenzorjiRGB-Dsezdajizvajav glavnemz uporaboinverznihkinematičnihtehnik.Vpoglavju9 so namestoavtorjiopisaniintuitiven načinzateleoperacijoindustrijo. MicrosoftKinectkotvidnisenzorzastrojnidelinogrodjerobotskioperacijskisistem(ROS)delujezaprogramskideldaizpolniteleoperativnanaloga.Dobljensistemjezelomodularen.

Chapter 1

A proper kinematics analysis is required to determine the current and target position of the robot arm. In this research, a geometric approach is used to determine the inverse kinematics of a robot arm.

Figure 1. Mobile humanoid robot system
Figure 1. Mobile humanoid robot system

Placing Motion Simulation Results

In task 1, the robot arm has to perform a simple motion of placing a bottle on the table, as shown in Figure 9(a) , and in task 2, a reaching and holding motion is selected ( Figure 9(b) ). If all three criteria have equal priority in generating the motion of the robot arm, NN2 performs best.

Figure 10 (c) shows NN3 solution which is similar to a single objective optimization, minimizing the  total acceleration (f 3 ) of the robot hand motion
Figure 10 (c) shows NN3 solution which is similar to a single objective optimization, minimizing the total acceleration (f 3 ) of the robot hand motion

Reaching and Holding Motion

Results show that the robot hand reached the target position in 90% of the trials out of 20 times. In this environment, the successful rate of the robot reaching the goal position is reduced.

Figure 15. Selected neural controllers for (a) right hand picking (b) right hand holding (c) left hand  picking (b) left hand holding
Figure 15. Selected neural controllers for (a) right hand picking (b) right hand holding (c) left hand picking (b) left hand holding

Chapter 2

Interaction with the environment makes no sense when the final effect of the manipulator is in contact with any surface, otherwise there is free movement. A wrist force sensor, this sensor recorded only inertial forces due to end-effector acceleration.

Figure 1. Human-robot cooperative system (Instrumentalist robot)
Figure 1. Human-robot cooperative system (Instrumentalist robot)

It is said that the movement constraints imposed on the cooperative system are homoge- neous if the scleronomic constraint of movement of a cooperative system given by (43), we can

Then there is the whole dynamics of a cooperative system described by nonlinear ordinary differential equations over another set of algebraic equations representing the constraints, resulting in a system of differential algebraic equations. In such systems, the position of each robot's end-effector is in contact with an object that is supposed to be rigid, and motion constraints appear. Since there is no relative movement between the end parts of the individual manipulator and the object, we established geometric constants that connect the end points of the manipulators.

Therefore, the end effector of the i-th manipulator can be defined in terms of the positions of the remaining manipulators. This is a logical consequence, since otherwise we could lose the point of contact between any of the manipulators and the object. Given the holonomic homogeneous constraint (43), the constrained velocity of the i-th manipulator is defined as: .

Restricted position variable is given by the integral of the restricted velocity, that is

The desired positions must conform to the constraint at any time to ensure that contact is maintained between the end effector and the object. Because of this, position and velocity errors are included in the control law and in the nominal reference signal that ensures that the rigid object is in contact with the end effectors of the robots involved in the cooperative system. The dynamic model of each cooperative system robot is obtained from the Euler-Lagrange formulation.

The robots are formed by rotational joints. ■

The links of the cooperative manipulators are rigid. ■

Considering a single contact point between each robot manipulator and the object, the movement constraints’ number on the cooperative system is equal to the number of robot manipu-

The movement constraints imposed on the cooperative system are holonomic and homogeneous. ■

While the end-effector of the manipulators is in contact with the surface, it is ϕi( )qi =0,. Let pi ∈n×1 denote the vector of the position of the ith end-effector coordinate frame. The dynamic behavior is defined as the grip's response to time before changes or disturbances in the position trajectories and the force applied to the contact point on the object's surface.

Planning is the selection of an appropriate trajectory for the robotic end-effector to approach the body and the correct selection of contact points between the object and the end-effectors. This polynomial trajectory ensures that the end effectors are placed smoothly near the object at finite time, Figure 18. Choosing the appropriate trajectory to bring the robotic end effector closer to the body and choosing the correct contact points between the object and the end effectors.

To ensure a proper grip between the body and the end-effectors of the robot manipulators. Distributed Impedance Control of Multiple Robotic Systems, Proceeding of the IEEE International Conference on Robotics and Automation.

Figure 6. Free body diagram
Figure 6. Free body diagram

Chapter 3

  • Garrido

A very efficient algorithm that was used to calculate the Voronoi Diagram (2D) in this work is due to (Breu, 1995). This method is iterative, starting at the source point of the wave (or waves) where T i j( , )0 0. The first step consists in calculating the Voronoi Diagram of the 2D or 3D map of the environment (that is, the cells equidistant from the obstacles).

More precisely, a skeletonization of the image, largely based on the (Breu, 1995) methodology (2D) and (Gagvani, 1997) (3D), is applied to get the Voronoi diagram, as shown in Figure 3 (left). This thickening of all lines of the Voronoi diagram is completed within the same amount. Map of the chamber used in the first experiment (left) Voronoi diagram (right) Thickened Voronoi diagram.

Figure 1. The function  T x y ( , )  gives a cone-shaped surface. The height  T  gives the set of points reached  at time  T
Figure 1. The function T x y ( , ) gives a cone-shaped surface. The height T gives the set of points reached at time T

Chapter 4

  • A. Fountas
  • M. Vaxevanidis
  • I. Stergiou
    • Basic Structure of Genetic Algorithms
    • Basics of Evolutionary Algorithms
    • GA-EA Performance: Exploration and Exploitation
    • Drawbacks of Genetic and Evolutionary Algorithms
    • Virus Theory of Evolution: Brief Overview
    • Virus-Evolutionary Genetic Algorithm Architecture
    • Software Development for Automated Machining Modeling
    • Infrastructure and Optimization Workflow
    • Implementation and Algorithm Settings
    • Optimization Results

The neighborhood of a candidate solution in the solution space must correspond to the neighborhood of the respective string in the string space. In the first category, selection is based on an individual's fitness value compared to the overall fitness value of the total population. Selection scheme for ranking: This scheme is based on the ranking of the fitness value of the individual.

Expected value" selection scheme: This scheme is based on the expected value of the individual's fitness. Depending on the number of breakpoints among individuals, the crossover mechanism combines the strings of the genotype. One is the host population, which basically works in the same way as the population in GA.

Figure 2. Graphical depiction of a function’s exploration and exploitation regions
Figure 2. Graphical depiction of a function’s exploration and exploitation regions

A conveyor belt forwards the raw part to a certain position and pushes the previously completed part to storage

Using a virtual configuration composed by CNC machine tool models, robot arms and conveyor belts (Figure 16); everything prepared and tested by CAD/CAM, it is possible to define scheduling times for all industrial operations. Conveyor belt speed can also be known, so the time a part stays on it or moves on it can be easily calculated. If a manufacturing cell is simple, a process planning study is more than enough to maximize productivity while reducing idle times.

On the contrary, if a production cell is more complicated, special strategies can be implemented to tackle with optimal productivity cycles and beneficial performance. The process of flexible production systems scheduling optimization can be handled by appropriate artificial intelligence algorithms (Kubota et al., 1996) such as the one presented in this study. A conveyor belt advances the raw part to a specific position and pushes the previously finished part into storage.

A robotic takes up the raw part and carries it to the CNC machine tool vice

The vice automatically grips the raw part and robot moves away to safe position

Machining processes begin. Robot waits until all machining phases are completed

Robot moves towards the vice. Vice jaws open and robot picks up the completed part

Robot places the completed part to a specific position on the conveyor belt

Steps 1 to 6 are repeated until batch production is complete

If the CNC machining center is in process before the robot system, the robot stands by until the machining is completed and then loads a new part setup (accessories and raw material). Also note that the transport time of the robot system is ignored compared to the NC machining time on a machining center and the loading/unloading time of raw materials and final parts respectively is constant. Thus, the set of i-th genes on each string in the individual is the palette of the fixtures on the i-th CNC machining center.

Constraints can be introduced into the optimization problem such that overlapping of the same fixture type on the same pallet is not allowed. The cost of evaluating the optimization methodology depends on the morphology and complexity of the work, however;. A New Genetic Approach for the Traveling Salesman Problem, Proceeding of the 1st IEEE Conference on Evolutionary Computing, 1, (pp: 7-12).

Chapter 5

  • INTRODUCTION
  • APPROACH COMPARISON
  • DISCRETE EVENT SYSTEMS
    • The Automats
    • The Networks of Queue
    • The Petri Nets
  • THE SIMULATION OF DISCRETE EVENT SYSTEMS
    • The Production Systems
    • The Evolution of Production Systems
    • The Simulation of Production Flows
  • EXISTING TOOLS FOR THE SIMULATION OF INDUSTRIAL PROCESSES To be reactive, must be competitive, is why manufacturers today have a vision to high production tech-
    • Tools of modeling and simulation « ARENA© »
    • Tool of modeling and simulation « WITNESS »
    • Tool of modeling and simulation « QUEST »
    • Reviewer and Our Proposed Approach
  • VIRTUAL REALITY AND ITS CONTRIBUTION TO THE INDUSTRY
    • Virtual Reality
    • Virtual and Augmented Reality for Industrial Design
    • Virtual and Augmented Reality for Numerical Production and Virtual Training
    • Virtual, Augmented Reality and the Simulation of Industrial Processes The use of existing tools today for the simulation of various industrial processes, requires a good knowl-
  • CONTRIBUTION TO THE SIMULATION OF VIRTUAL MANUFACTURING
    • Architectural Design
    • Functioning of System
    • Technical Environment
  • PERSPECTIVES
  • CONCLUSION

A production system provides the necessary mechanism to carry out production to achieve the goal of the system. The establishment of the evolution of production systems was initiated by the automobile industry (Womack, Jones, and Ross, 1990). Various techniques are used to simulate models, such as queuing networks, discrete event simulation, operational analysis or Petri nets.

And by Klingstam, the virtual reality tools will do part of the future of the production system simulation (Klingstam & Gullander, 1999). ARENA contains a base collection of more than 60 graphic components in the general context of the system. These systems are often used for the simulation of the manufacturing and the implementation of a plant of the factory in an industrial platform.

Figure 1. Example of a model of Petri nets
Figure 1. Example of a model of Petri nets

Hình ảnh

Figure 1. Mobile humanoid robot system
Figure 2. Left and right hand design of the mobile humanoid robot
Figure 5. Mobile platform
Figure 10 (c) shows NN3 solution which is similar to a single objective optimization, minimizing the  total acceleration (f 3 ) of the robot hand motion
+7

Tài liệu tham khảo

Tài liệu liên quan

Wind turbine model of DFIG is a doubly fed induction generator or a variable speed, doubly fed asynchronous generator with rotor-side converter called DFIG wind