Chap1 Control challenges in robotics and automation¶

Introduction: Automation¶

Automation¶

Automation of repeat functions(via measurement+control) at the process level(control algorithm)

Transfer of demanding guidance and monitoring tasks to the human being in the strategy level(high-level process control techniques)

Model¶

The control algorithm bases on a mathematical decription of the process, a model, reflecting the process as a function of I/O parameters and environmental influences

Model is an abstract replica of the system that serves to plan and predict the process behavior

Basic concepts of control/automation¶

Control approach forms the basis of automation!

control path: the part of a system(process) in which one or more state variables( 状态变量 )/process parameters( 过程参数 ) are to be influenced to control the system

disturbances: disturbances of different nature intervene into the control path, which can affect the state variables

closed-loop control: a type of control when a state variable(system output) is permanently measured by a sensor and compared with the desired value, the reference

open-loop control(non-feedback control): a type of control when the control action from the controller(system input) is independent of the system output

Automation hierarchy( 自动化层次结构 ) ¶

process level -> operation level -> coordination level -> strategy level (from low to high)

At the higher levels, the reference variables are determined at an abstract level, the software has increasing intelligence and complexity; At the lower levels, the process control is specified step by step down to the single state values of the process, the software has increasing real-time capability

Introduction: Robotics¶

Industrial Robots¶

2^nd and 3^rd generation robots need advanced contorl approaches: artificial neural networks and fuzzy logic

Basic structure and components of a robot¶

mechanical subsystems

• Vehicle or platform: coarse positioning

• Manipulator: fine positioning

• Endeffector: specific application(mimicking a human)

Robots are characterized by

• Kinematics/Dynamics( 运动学与动力学 )

• Actuators/Sensors( 驱动器与传感器 )

• Control system

• SW/User interface

Kinematics( 运动学 )/Dynamics( 动力学 ) ¶

In robotics, kinematics is used to describe the motion and the trajectories( 运动和轨迹 ) of an endeffector( 末端执行器 ) in relation to the position of each joint(link)( 关节 ) of the robot

The direct kinematics( 正运动学 ) problem: to compute the position(the trajectory) of the endeffector from specified values of the joint parameters(position/angles)

The inverse kinematics( 逆运动学 ) problem: to compute the joint parameters(position/angles) that produce a desired endeffector position(trajectory)

Transformation mattrices( 变换矩阵 ) are usually used to solve the kinematics problem: simple linear matrix calculations

Dynamics is a branch of classical mechanics which deals with the motion of bodies under the action of forces/torques( 扭矩 )

Actuators/Sensors¶

The sensors are used to measure the state variables of the robot and its workspaces, provide feedback necessary for closed-loop robot control

The actuators(drives of the robot) are responsible for moving a mechanism as reaction of the control system to sensor feedback

Control system¶

The tasks of the control system include: trajectory planning( 轨迹规划 ), collision avoidance( 碰撞避免 )， position/force control

Application¶

the possible areas of allication of fuzzy logic and neural networks:

• General control problems, process monitoring( 过程检测 )

• Planning and optimization( 规划和优化 )

• Identification and modeling( 辨识和建模 )

• Evaluation of sensor data/failure diagnosis( 传感数据评估和故障诊断 )

• Navigation: path planning, collision avoidance( 导航 )

• Expert systems, decision making, classification of features( 专家系统 )

• Pattern recognition and image processing( 模式识别和图像处理 )

• Speech recognition, machine translation( 语音识别和机器翻译 )

评论区