Data acquisition and learning capabilities are necessary for an intelligent system operating in unstructured, dynamically changing environments. For this purpose, a method for the effective use of multiple sensors mus...
详细信息
Data acquisition and learning capabilities are necessary for an intelligent system operating in unstructured, dynamically changing environments. For this purpose, a method for the effective use of multiple sensors must be developed. This paper shows how multisensor fusion can be accomplished by neural networks. It first summarizes the conventional fusion techniques and consequently describes the use of neural networks for sensor fusion as well as their advantages. Finally, an application is presented where a neural network is used to fuse the signals of several sensors, of different type, for material identification purposes.
We consider the feedback stabilization of nonlinear systems. The design methodology is based on a class of iterative methods, which has previously been proposed for the path planning and feedback stabilization of nonh...
详细信息
We consider the feedback stabilization of nonlinear systems. The design methodology is based on a class of iterative methods, which has previously been proposed for the path planning and feedback stabilization of nonholonomic systems. We show that this scheme guarantees the closed loop asymptotic stability for a class of nonlinear systems.
A number of approaches based on sliding mode control methodology are proposed in the literature for the position control of robotic manipulators. An important problem in this context is chattering, that is high freque...
详细信息
A number of approaches based on sliding mode control methodology are proposed in the literature for the position control of robotic manipulators. An important problem in this context is chattering, that is high frequency oscillations in the velocity. In this paper, a novel approach is considered which eliminates chattering if the controller parameters are set suitably. A fuzzy adaptation scheme is devised for the online adaptation of the parameters of this method. Simulation results show that the fuzzy rule based adaptation scheme ensures good controller performance without chattering.
This paper discusses control behavior integration and bucket action fusion for excavation control of a robotic front-end-loader type machine, To utilize the experience and expertise from skilled human operators, a fuz...
详细信息
This paper discusses control behavior integration and bucket action fusion for excavation control of a robotic front-end-loader type machine, To utilize the experience and expertise from skilled human operators, a fuzzy-logic based control approach is developed, A hierarchical excavation control architecture decomposes excavation goals to tasks, then tasks to behaviors, and finally behaviors to actions, The excavation actions are primitive and can be executed directly by an excavation machine, Finite state machines are used to specify the coordination and integration of behaviors for task execution and actions for behavior implementation, A simple strategy for action fusion is proposed based on fuzzy logic reasoning and the COA defuzzification method. Finally, laboratory experiments are conducted using a PUMA 560 robot arm and a Zebra force/torque sensor in a simulated rock excavation environment. Experimental results indicate that the proposed approach in this paper has led to more efficient task execution than our previous approach.
Most robust control methods of robot manipulators guarantee small tracking errors by applying sufficiently high feedback gains. Infinite gains are required for zero tracking errors under such controllers. However, in ...
Most robust control methods of robot manipulators guarantee small tracking errors by applying sufficiently high feedback gains. Infinite gains are required for zero tracking errors under such controllers. However, in practice, feedback gains could be severely limited by hardware factors. Robust control schemes without using high gains are desirable. In this paper, we propose a globally asymptotically stable robust control scheme by combining integral control with a robust saturation control law. The proposed robust control method takes advantage of both saturation control and integral control techniques, while the disadvantages attributed to them are remedied by each other. Global asymptotic stability of an n-link robot system with parametric model uncertainties is achieved. The proposed control scheme has been implemented on a direct drive robot arm. Experiments were conducted to evaluate the proposed control method and to compare it with pure saturation and integral control methods, and experimental results are reported herein. (C) 1996 Elsevier Science Ltd.
This paper addresses the problem of robust hybrid position and force control of robot manipulators. Variable structure control with sliding mode is used to implement the hybrid control strategy. Two variable structure...
详细信息
This paper addresses the problem of robust hybrid position and force control of robot manipulators. Variable structure control with sliding mode is used to implement the hybrid control strategy. Two variable structure control algorithms are developed in task space. One of the algorithms is based on hierarchical control method, and the other is developed for control of robot manipulators used to carried out both unconstrained and constrained tasks.
Using data gathered over many years, people working in sheet metal fabrication have compiled lists of errors that are frequently made by designers in the field. These errors (or violations) are common to the design of...
Using data gathered over many years, people working in sheet metal fabrication have compiled lists of errors that are frequently made by designers in the field. These errors (or violations) are common to the design of sheet metal components used in a wide variety of products from computer chassis to automobiles. This paper outlines the efforts to use this information to aid in the development of a rule-based design advisor for sheet metal fabrication. The design advisor is a module of Promod-S, a concurrently engineered, application-orientated product modeler that is being developed by researchers at the automation and robotics Laboratory of the University of Massachusetts at Amherst. The paper presents algorithms for both identification and correction of design violations. The correction approach includes methods for updating an entire part after a feature has been modified.
In this paper, a methodology for robot position/force controller design is discussed. Unlike pure position controllers for robotic manipulators, the position/force controller consists of two parts: position and force ...
详细信息
In this paper, a methodology for robot position/force controller design is discussed. Unlike pure position controllers for robotic manipulators, the position/force controller consists of two parts: position and force subcontrollers. The principle of consistence, as a guideline for design of position /force controllers, is proposed, which requires that both sub-controllers not interact with each other. A systematic approach, based on the principle of consistency, is submitted to be a platform for the design of position/force controllers.
In this paper, fundamental principles of design of position/force controller for robot manipulators executing constrained task are enunciated. The principles are based on stability considerations, and are addressing t...
详细信息
In this paper, fundamental principles of design of position/force controller for robot manipulators executing constrained task are enunciated. The principles are based on stability considerations, and are addressing the dynamic interaction between position and force subcontrollers. The principles could serve as design guidelines that would guarantee overall stability and reduction of the interactions between the subcontrollers. The concept of consistency introduced in this paper forms the cornerstone of proper implementation of position/force (P/F) controllers, in terms of reduction of the interaction between the two subcontrollers. The principle of consistency indicates that the direct interaction is canceled via the design of a stable P/F controller, however the indirect interaction can only be removed if a Jacobian condition is satisfied, independent of the subcontrollers design.
For pt.I see ibid. In this part we focus on the fuzzy system identification stage. The Fuzzy C-Means (FCM) clustering algorithm is investigated. Two indices are introduced for assignment of number of rules and level o...
详细信息
For pt.I see ibid. In this part we focus on the fuzzy system identification stage. The Fuzzy C-Means (FCM) clustering algorithm is investigated. Two indices are introduced for assignment of number of rules and level of "fuzziness". An efficient method is also suggested for choosing the initial location of cluster centers. The classification problem is addressed in fuzzy modeling; and an efficient strategy is proposed for assignment of dominant input variables and their membership functions.
暂无评论