The purpose of this thesis is to present a comprehensive and practical approach for the time-optimal motion planning and control of a general purpose industrial manipulator. In particular, the case of point-to-point path unconstrained motions is considered, with special emphasis towards strategies suitable for efficient on-line implementations. From a dynamic model description of the plant, and using an advanced graphical robotics simulation environment, the control algorithms are formulated. Experimental work is then conducted to verify the proposed algorithms, by interfacing the industrial manipulator to the master controller, implemented on a personal computer. The full rigid-body non-linear dynamics of the open-chain manipulator have been accommodated into the modelling, analysis and design of the control algorithms. For path unconstrained motions, this leads to a model-based regulating strategy between set points, which combines conventional trajectory planning and subsequent control tracking stages into one. Theoretical insights into these two robot motion disciplines are presented, and some are experimentally demonstrated on a CRS A251 industrial arm. A critical evaluation of current approaches which yield optimal trajectory planning and control of robot manipulators is undertaken, leading to the design of a control solution which is shown to be a combination of Pontryagin's Maximum Principle and state-space methods of design. However, in a real world setting, consideration of the relationship between optimal control and on-line viability highlights the need to approximate manipulator dynamics by a piecewise linear and decoupled function, hence rendering a near-time-optimal solution in feedback form. The on-line implementation of the proposed controller is presented together with a comparison between simulation and experimental results. Furthermore, these are compared with measurements from the industrial controller. It is shown that the model-based near-optimal-time feedback control algorithms allow faster manipulator motions, with an average speed-up of 14%, clearly outperforming current industrial controller practices in terms of increased productivity. This result was obtained by setting an acceptable absolute error limit on the target location of the joint (position and velocity) to within [2.0E-02 rad, 8.7E-03 rad/s], when the joint was regarded at rest.

The reported work lies in the area of automation in the construction industry, and involves multi-disciplinary engineering studies. In particular, sewer renovation methods, computer vision (CV) and robotics are all included. More specifically, the key objective of the research programme was to investigate the characteristics of retrofit components suited to mounting on an industrial / proprietary sewer tractor. The overall aim was the provision of a non-man-entry (NME) sewer renovation robot to undertake reconnection of lateral junctions, following a cured-in-place (CIP) relining process. The programme primarily involved theoretical studies of the requisite sensory and kinematic components, incorporation of a novel computer vision sensing system and production of a chainage measurement system and robotic drill task arm. The theory was supported by laboratory testing using a modified proprietary tractor, with emphasis placed on promoting applications of information technology driven systems (i.e. CV) to construction-industry tasks within hazardous environments involving significant health issues. The use of such techniques in the construction industry is rare. Chapter 1 reviews the context and history of sewer maintenance/dereliction in the UK. NME sewers are the most common type and are, by definition, difficult to maintain. Renovation, typically employing CIP liners, is therefore a cost-effective alternative to replacement. Lateral connections are, inevitably, blocked off during the relining process; it is suggested that application of a robust robotic system to the task of reconnecting them is novel and offers clear potential within such a hazardous environment. Chapters 2 and 3 develop the underlying theoretical models of the CV and kinematic systems respectively. The novel CV work (provided by third party specialists employing the TINA CV research environment) was incorporated by the author to provide detection and classification of lateral junctions, crucially noting the particular properties of direct and reflected illumination. Classification aspects include estimation of lateral/NME intersection angle and closure-to-target distance from the robot. The author proposes a separate procedure for estimating lateral diameter. A chainage measurement system, using a rotary encoder and inclinometer, was developed to determine invert path distance travelled. This allows for the inevitable wander and thereby gives the system robustness. The novel application of GRASP (a robotic modelling and simulation design tool) to NME environments, provided the ability to model arm designs without the need for the production of more than one expensive physical prototype. A mathematical solution for determining the requisite arm kinematics is presented. Chapter 4 details the hardware requirements of the robotic system components, whilst Chapters 5 and 6 present the laboratory evaluation results for the kinematic and CV systems respectively. The abilities of the CV system qualitatively to detect laterals under reflected illumination, and to provide quantitative classification data, are demonstrated. The chainage measurement system is assessed under a variety of initialisation conditions to determine suitability to task, and the ability of the robotic arm to physically simulate lateral reconnection is investigated. Chapter 7 discusses the specification for an industrially-applicable prototype, based on the findings herein. Appropriate comparisons with the pre-prototype system are made, including cost. Finally, Chapter 8 draws conclusions and makes suggestions for further work. Supporting documentation is provided in Chapter 9 and the Appendices.

The values of a given manipulator's dynamics coefficients need to be accurately identified in order to employ model-based algorithms in the control of its motion. This thesis details the development of a novel form of adaptive network which is capable of accurately learning the coefficients of systems, such as manipulator inverse dynamics, where the algebraic form is known but the coefficients' values are not. Empirical motion data from a pair of PUMA 560s has been processed by the Context-Sensitive Linear Combiner (CSLC) network developed, and the coefficients of their inverse dynamics identified. The resultant precision of control is shown to be superior to that achieved from employing dynamics coefficients derived from direct measurement. As part of the development of the CSLC network, the process of network learning is examined. This analysis reveals that current network architectures for processing analogue output systems with high input order are highly unlikely to produce solutions that are good estimates throughout the entire problem space. In contrast, the CSLC network is shown to generalise intrinsically as a result of its structure, whilst its training is greatly simplified by the presence of only one minima in the network's error hypersurface. Furthermore, a fine-tuning algorithm for network training is presented which takes advantage of the CSLC network's single adaptive layer structure and does not rely upon gradient descent of the network error hypersurface, which commonly slows the later stages of network training.

No description available.

Object recognition has been one of the main areas of research into computer vision in the last 20-30 years. Until recently most of this research has been performed on scenes taken using static monocular, binocular or even trinocular cameras. It is believed, however, that by adding the ability to move the look point and concentrate on a region of interest a more robust and efficient method of vision can be achieved. Recent studies into the ability to provide human-like vision systems for a more active approach to vision have lead to the development of a number of robot controlled vision systems. In this thesis the development of one such system at the University of Surrey, the stereo robot head "Getafix" is described. The design, construction and development of the head and its control system have been undertaken as part of this project with the aim of improving current vision tasks, in particular, that of object recognition. In this thesis the design of the control systems, kinematics and control software of the stereo robot head will be discussed. A number of simple commissioning experiments are also shown, using the concepts of the robot control developed herein. Camera lens control and calibration is also described. A review of classical primitive based object recognition systems is given and the development of a novel generic cylindrical object recognition strategy is shown. The use of this knowledge source is demonstrated with other vision processes of colour and stereo. The work on the cylinder recognition strategy and the stereo robot head are finally combined within an active vision framework. A purposive active vision strategy is used to detect cylindrical structures, that would otherwise be undetectable by the cylindrical object detection algorithm alone.

In a complex and flexible manufacturing environment tasks maybe dynamically reconfigured. In this situation a robot needs to plan paths automatically to avoid obstacles and rendezvous with changing target points. A novel path planning system is presented which takes into account both kinematic and dynamic constraints. The main part of the system comprises a robot "Path Planner" and "Path Adapter", both using a dynamic "World Model" updated by a vision system. The Path Planner contains a geometric model of the static environment and the robot. Given a task, the Path Planner calculates an efficient collision free path. This is passed to the control computer where a trajectory is generated. Pre-determination of optimum paths using established techniques frequently involve unacceptably high time penalties. To overcome this problem the automatic path refinement techniques employed avoid the necessity for optimality before beginning a movement. Repeated improvements to the sub optimal paths initially generated by the Path Planner are made until the robot is ready to begin the new path. Algorithms are presented which give a rapid solution for simplified obstacle models. The algorithms are robust and are especially suitable for repetitive robot tasks. With the Path Planner, the robot structure is modelled as connected cylinders and spheres and the range of robot motion is quantised. The robot path, calculated initially only takes account of geometric, kinematic and obstacle constraints. Although this path is sub optimal, the calculation time is short. The path avoids obstacle and seeks the "shortest" path in terms of total actuator movement. Several of the new path planning methods presented employ a local method, taking a "best guess" at a path through a 2-D space for two joints and then calculating a path for the third joint such that obstacles are avoided. A different approach is global and depends on searching a 3-D graph of quantised joint space. The Path Planner works in real time. If there is enough time available a "Path Adapter" modifies the planned path in an effort to improve the path subject to selected criteria. The Path Adapter considers dynamic constraints. The first robot path improvement method depends on detecting the joint motor currents in order to minimise changes in joint direction, the other is based on a set of adaptive rules based on simplified dynamic software models of the robot stored within the planning computer. The adapted path is passed to the control computer. The static model of the robot work-cell is held in computer memory as several solid polyhedral. With the aid of a vision system, this model is updated as new obstacles enter or leave the work-place. Overlapping spheres and 2-D slices in joint space are used to model obstacles. In this form the vision system can be updated quickly and the obstacle data can de accessed efficiently by the path planning and path improvement algorithms.

The work presented in this thesis attempts to improve the performance of industrial robot systems in a flexible manufacturing environment by addressing a number of issues related to external sensory feedback and sensor integration, robot kinematic positioning accuracy, and robot dynamic control performance. To provide a powerful control algorithm environment and the support for external sensor integration, a transputer based open architecture robot controller is developed. It features high computational power, user accessibility at various robot control levels and external sensor integration capability. Additionally, an on-line trajectory adaptation scheme is devised and implemented in the open architecture robot controller, enabling a real-time trajectory alteration of robot motion to be achieved in response to external sensory feedback. An in depth discussion is presented on integrating a stereo vision sensor with the robot controller to perform external sensor guided robot operations. Key issues for such a vision based robot system are precise synchronisation between the vision system and the robot controller, and correct target position prediction to counteract the inherent time delay in image processing. These were successfully addressed in a demonstrator system based on a Puma robot. Efforts have also been made to improve the Puma robot kinematic and dynamic performance. A simple, effective, on-line algorithm is developed for solving the inverse kinematics problem of a calibrated industrial robot to improve robot positioning accuracy. On the dynamic control aspect, a robust adaptive robot tracking control algorithm is derived that has an improved performance compared to a conventional PID controller as well as exhibiting relatively modest computational complexity. Experiments have been carried out to validate the open architecture robot controller and demonstrate the performance of the inverse kinematics algorithm, the adaptive servo control algorithm, and the on-line trajectory generation. By integrating the open architecture robot controller with a stereo vision sensor system, robot visual guidance has been achieved with experimental results showing that the integrated system is capable of detecting, tracking and intercepting random objects moving in 3D trajectory at a velocity up to 40mm/s.

No description available.

No description available.

The issue of Command and Control (C2) is generally associated with the management infrastructure of large scale systems for warfare, public utilities and public transportation, and is concerned with ensuring that the distributed human elements of command and control can be fully integrated into a coherent, total system. Intelligent Autonomous Systems (IASs) are a class of complex systems that perform tasks autonomously in uncertain, dynamic environments, the management of which can be viewed from the perspective of embedded command and control systems. This thesis establishes a vision for the modular construction of intelligent autonomous embedded C2 systems, which defines a complex integration problem characterised by distributed intelligence, world knowledge and control, concurrent processing on heterogeneous platforms, and real-time performance requirements. It concludes that by adopting an appropriate systems infrastructure model, based on Object Technology, it is possible to view the construction of embedded C2 systems as the integration of a temporally assembled collection of reusable components. To support this metaphor it is necessary to construct a common reference model, or standards framework, for the representation and specification of modular C2 systems. This framework must support the coherent long term development and evolution in system capability, ensuring that systems are extensible, robust and perform correctly. In this research, which draws together the themes of other published research in object oriented systems and robotics, classical AI models for intelligent systems architectures are used to specify the overall system structure, with open systems technologies supporting the interoperation of elements within the architecture. All elements of this system are modelled in terms of objects, with well defined, implementation independent interfaces. This approach enables the system to be specified in terms of an object model, and the development process to be framed in terms of object technology, defining a new approach to IAS development. The implementation of an On-board Command and Control System for an Autonomous Underwater Vehicle is used to validate these concepts. The further application of emergent industrial standards in distributed object oriented systems means that this kind of component-based integration is scaleable, providing a near-term solution to generic command and control problems, including Computer Integrated Manufacturing and large scale autonomous systems, where individual autonomous systems, such as robots, form elements of a complete, total intelligent system, for application to areas such as fully automated factories and cooperating intelligent autonomous vehicles for construction sites.