China
Africa
Explore chapters and articles related to this topic
Types of Robols and Their Integration into Computer-Integrated Manufacturing Systems
Published in Ulrich Rembold, Robot Technology and Applications, 2020
Sensor systems, tactile and nontactile, internal or external, perceive the interaction between the robot and the environment. A sensor processing system interpreting sensor data from the actual operation enhances the autonomy of the robot. The robot control system translates the sensor data and motion commands into signals for the servo controllers and the electromechanical transducers, to instruct the mechanical system to follow the desired trajectory. The control system tries to minimize the positioning or path errors. The programming system allows the specification of a given task with high level language instructions, graphic test facilities, and decision support (in many cases, with the aid of a user-friendly function menu). Programming may be done interactively using simulation techniques or automatically with the aid of an action sequence planner. On this level, the use of artificial intelligence techniques is of great advantage. Advanced autonomous robot systems perform action sequence planning, program execution, mission supervision, and on-line error recovery. Thus, goal-oriented behavior, reaction to unexpected events, and handling of exceptions are expected. Passive and active learning strategies enable the robot to acquire skills using experience from the past. Building such a system requires a strong interaction among all components and a clear separation of the control levels that are performing their specific tasks.
Model-Based Fault Accommodation Control of Robotic Systems
Published in Sunan Huang, Kok Kiong Tan, Poi Voon Er, Tong Heng Lee, Intelligent Fault Diagnosis and Accommodation Control, 2020
Sunan Huang, Kok Kiong Tan, Poi Voon Er, Tong Heng Lee
Robotic control systems include many components, such as sensors, actuators, joints and motors. These components are required to function according to some specifications and control requirements in order for the overall system to operate precisely and reliably. Especially, when a failure occurs, the controlled system should diagnose the failure and continue to maintain the operation so that the system avoids a total collapse in function [61, 62]. Various approaches to fault detection have been reported during the last two decades. It has been shown that the use of adequate process models can allow early fault detection with normal measurable variables [63]. For robotic systems, the model-based analysis of fault diagnosis has received considerable attention ([33, 46]). In [46], an expert system model is developed for fault detection. In [33], a dynamical model is presented to detect incipient fault. However, the model-based fault detection schemes depend on the assumption that a mathematical characterization of the robotic system is available. In practice, this is not true since it is difficult to obtain an exact model. By using neural network (NN) approximation, robust fault detection schemes for robotic manipulators have been developed in [34, 64].
Computer Vision in Multi-Robot Cooperative Control
Published in C.W. de Silva, Mechatronic Systems, 2007
The objective of our multi-robot transportation project in the Industrial Automation Laboratory is to develop a physical mechatronic system where a group of intelligent robots work cooperatively to transport an object to a goal location and orientation in an unknown dynamic environment. Obstacles may be present and even appear randomly during the transportation process. Robot control, multi-agent technologies, and machine learning are integrated into the developed physical platform to cope with the main challenges of the problem. A schematic representation of the first version of the developed system is shown in Figure 14.2. The latest system in the laboratory uses state-of-the-art mobile robots together with fixed-base articulated robots (one state-of-the-art commercial robot and two prototypes developed in our laboratory).
Towards Knowledge Sharing Oriented Adaptive Control
Published in Cybernetics and Systems, 2022
Guixian Li, Yufeng Xu, Haoxi Zhang, Edward Szczerbicki
Nowadays, machine learning is becoming increasingly popular for solving robot control problems (Zhou et al. 2019). It helps robots perform various tasks well in uncertain and complex conditions (Bokeno et al. 2018; Liang 2019; Sajedi and Liang 2019). However, training a robot for particular tasks are still time-consuming and expensive (Ding et al. 2016; Liang, Zheng, and Zhang 2018).