This paper presents and investigates a neural network structure which can perform general fuzzy inference. This system consists of a number of parallel neural network units which are called ''flexible inferenc...
详细信息
This paper presents and investigates a neural network structure which can perform general fuzzy inference. This system consists of a number of parallel neural network units which are called ''flexible inference cells'' (FICs). Each FIC implements a single-input/single-output (SISO) IF-THEN rule of a fuzzy knowledge base. The assumption of SISO fuzzy rules allows the implementation of any complex fuzzy inference algorithm (for control or other decision making purposes), since any MIMO (multi-input/multi-output) rule can be decomposed into an equivalent set of MISO (multi-input/single-output) rules, and a MISO rule can be decomposed to a set of SISO rules. The paper discusses the assumptions and requirements for the proposed neurofuzzy structure, and classifies the FICs into four categories. Some results derived by simulation using 3125 exemplar patterns produced computationally are provided.
Neural networks or connectionist models are massively parallel interconnections of simple neurons that work as a collective system, can emulate human performance and provide high computation rates. On the other hand, ...
详细信息
Neural networks or connectionist models are massively parallel interconnections of simple neurons that work as a collective system, can emulate human performance and provide high computation rates. On the other hand, fuzzy systems are capable to model uncertain or ambiguous situations that are so often encountered in real life. One way for implementing fuzzy systems is through utilizations of the expert system architecture. Recently, many attempts have been made to ''fuse'' fuzzy systems and neural nets in order to achieve better performance in reasoning and decision making processes. The systems that result from such a fusion are called neuro-fuzzy inference systems and possess combined features. The purpose of the present paper is to propose such a neuro-fuzzy system by extending and improving the system of Keller et al. (1992). The present system makes use of Hamacher's fuzzy intersection function and Sugeno's complement function. After a brief outline of the operation of the system its features are established with the aid of four theorems which are fully proved. The capabilities of the system are shown by a set of simulation results derived for the case of trapezoidal fuzzy sets. These results are shown to be better than the ones obtained with the original neuro-fuzzy system of Keller et al.
作者:
Tzes, ALe, KPolytechnic University
Control/Robotics Research Laboratory Mechanical Engineering Department 333 Jay Street Brooklyn NY 11201 U.S.A.
The identification problem of transfer functions associated with flexible structures is addressed in this paper. An adaptive infinite impulse response filter with data preprocessing in the frequency domain is used to ...
详细信息
The identification problem of transfer functions associated with flexible structures is addressed in this paper. An adaptive infinite impulse response filter with data preprocessing in the frequency domain is used to identify directly the poles of the transfer function and the coefficients of the numerator polynomial. The parallel form realisation is utilised to model the structure of the identified transfer function. The stability of the identified system can easily be monitored and the sensitivity of its coefficients is lower compared to that of the classical direct form. Simulation studies are used to illustrate the proposed algorithm. (C) 1996 Academic Press Limited
A new control system is described for the PUMA 560 industrial robotic manipulator based on transputer networks, where both the hardware and software designs are detailed. A Transputer Interface Board (TIE) establishin...
详细信息
A new control system is described for the PUMA 560 industrial robotic manipulator based on transputer networks, where both the hardware and software designs are detailed. A Transputer Interface Board (TIE) establishing a transputer link to the 6503 microprocessors of the PUMA arm joints has been designed, built and tested successfully. In addition to hardware implementation, software testing for this new system had been accomplished. The new system can communicate with the PUMA lower level controller at a much shorter period (i.e. 1.75 ms) than the default 28 ms. Genetic Algorithms are used to plan the PUMA robot minimum-time motion trajectory, which is not possible by the traditional exhaustive search method. Real-time experiments have been carried out based on the new PUMA control platform, and show a very good match with simulations. controlled by the new system, PUMA perfoms better when it is interfaced with the shorter time period. Copyright (C) 1996 Elsevier Science Ltd.
A magnetic suspension system with linear actuators and permanent magnets (instead of electromagnets) has been developed. In this system the bearing forces are controlled by adjusting the air gaps between permanent mag...
详细信息
A magnetic suspension system with linear actuators and permanent magnets (instead of electromagnets) has been developed. In this system the bearing forces are controlled by adjusting the air gaps between permanent magnets and the levitated object (a circular disk). This system uses less energy and exhibits none of the over-heating of coils typical of systems employing electro-magnets. This paper describes a disk suspension system with three degrees of freedom. The system can operate either with a decentralized (distributed) feedback control system or with a centralized (lumped) one. Feedback gain matrices are calculated for each type of control system. The performances of the two types of control systems are compared by numerical simulations and experiment. This work is part of an effort to develop a magnetic levitation conveyer system.
Prototype based methods are commonly used in cluster analysis and the results may be highly dependent on the prototype used. In this paper, we propose a fuzzy clustering method that involves adaptively expanding conve...
详细信息
Prototype based methods are commonly used in cluster analysis and the results may be highly dependent on the prototype used. In this paper, we propose a fuzzy clustering method that involves adaptively expanding convex polytopes. Thus, the dependency on the use of prototypes can be eliminated. Also, the proposed method makes it possible to effectively represent an arbitrarily distributed data set without a priori knowledge of the number of clusters in the data set. Specifically, nonlinear membership functions are utilized to determine whether a new cluster is created or which vertex of the cluster should be expanded. For this, the membership function of a new vertex is assigned according to not only a distance measure between an incoming pattern vector and a current vertex, but also the amount of how much the current vertex has been modified. Therefore, cluster expansion can be only allowed for one cluster per incoming pattern. Several experimental results are given to show the validity of our method.
This paper presents a design experience of a supervisory control system for coordination of multiple robotic devices. To effectively program job commands, a Petri net-type graphical robot language(PGRL) is proposed, w...
详细信息
This paper presents a design experience of a supervisory control system for coordination of multiple robotic devices. To effectively program job commands, a Petri net-type graphical robot language(PGRL) is proposed, where some functions for coordination among tasks such as concurrency and synchronization, can be easily programmed. Each task of PGRL is described by employing formal model languages, which are composed of three modules, sensory, data handling, and action module. It is expected that by using our proposed PGRL and formal model languages, one can efficiently describe a job or task, and hence can easily operate a complex real-time concurrent system. The proposed control system has been implemented by using VME-based 32-bit microprocessor boards and a real-time multitasking operating system (VxWorks), and is shown to successfully work for robotic jobs.
A robust internal force-based impedance control scheme for coordinating manipulators is introduced. Internal force-based impedance control enforces a relationship between the velocity of each manipulator and the inter...
详细信息
A robust internal force-based impedance control scheme for coordinating manipulators is introduced. Internal force-based impedance control enforces a relationship between the velocity of each manipulator and the internal force on the manipulated objects and requires no knowledge of the object dynamic model. Each manipulator's nonlinear dynamics is compensated by a robust auxiliary controller which is insensitive to robot-model uncertainty and payload variation. The controller is only weakly-dependent on each manipulator's inertia matrix. Stability of the system is analyzed. The scheme is computationally inexpensive and suitable for general-purpose microcomputer implementation. Rigorous experimental investigations are performed and the results presented which validate the proposed concepts.
A common approach to control multi-fingered grippers during stable grasping is the stiffness control scheme. "Internal" stiffness parameters and suitable references for internal grasping forces-often determi...
详细信息
A common approach to control multi-fingered grippers during stable grasping is the stiffness control scheme. "Internal" stiffness parameters and suitable references for internal grasping forces-often determined heuristically-ensure a stable grasp. In this paper, optimal internal forces are obtained in real-time by linearly constrained gradient flows on the smooth manifold of positive definite matrices. This optimization approach is a generic one for any number of fingers in contact with the object. Experiments with the 3-fingered Darmstadt hand show the simplicity and efficiency of the approach.
This paper proposes a decentralised compensation scheme for unstructured uncertainties and modelling errors of robotic manipulators. The scheme employs a central decoupler and independent joint neural network compensa...
详细信息
This paper proposes a decentralised compensation scheme for unstructured uncertainties and modelling errors of robotic manipulators. The scheme employs a central decoupler and independent joint neural network compensators. Recursive Newton Euler formulas are used to decouple robot dynamics to obtain a set of equations in terms of input and output of each joint. Each joint sub-system is then controlled separately by neural network controllers which suppress the effects of uncertainties associated with the model. Multi-layered perceptrons using back-propagation learning algorithm are employed as the adaptive elements in the control scheme. The effectiveness of the proposed scheme is demonstrated by a simulation experiment on PUMA 560. Simulation results show that this control scheme achieves fast and precise robot motion control under substantial model inaccuracies. Properties of the neural compensation technique are compared with those of a globally stable adaptive controller.
暂无评论