Riyanto T. Bambang

Development of Architectures for Internet Telerobotics Systems

This paper presents our experience in developing and implementing Internet telerobotics system. Internet telerobotics system refers to a robot system controlled and monitored remotely through the Internet. A robot manipulator with five degrees of freedom, called Mentor, is employed. Client-server architecture is chosen as a platform for our Internet telerobotics system. Three generations of telerobotics systems have evolved in this research. The first generation was based on CGI and two tiered architectures, where a client presents a Graphical User Interface to the user, and utilizes the user’s data entry and actions to perform requests to robot server running on a different machine. The second generation was developed using Java. We also employ Java 3D for creating and manipulating 3D geometry of manipulator links, and for constructing the structures used in rendering that geometry, resulting in 3D robot movement simulation presented to the users (clients) through their web browser. Recent development in our Internet telerobotics includes object recognition through image captured by a camera, which poses challenging problem, giving the undeterministic latency of the Internet. The third generation is centered around the use of CORBA for development platform of distributed internet telerobotics system, aimed at distributing task of telerobotics system.

DSP based RBF neural modeling and control for active noise cancellation

This paper presents active control of acoustic noise using radial basis function (RBF) networks and a digital signal processor (DSP) real-time implementation. The neural control system consists of two stages: first, identification (modeling) of the secondary path of active noise control using RBF networks and its learning algorithm, and secondly neural control of the primary path based on the neural model obtained in the first stage. A tapped delay time is introduced in front of the neural controller, and another tapped delay line is inserted between controller neural networks and model neural networks. An algorithm referred to as FX-RBF is proposed to account for secondary path effects of the control system arising in active noise control. The resulting algorithm turns out to be the filtered-X version of the standard RBF learning algorithm. We address centralized and decentralized controller configurations and their DSP implementation is carried out. The effectiveness of the neural controller is demonstrated by applying the algorithm to active noise control within a 3-D enclosure to generate quiet zones around error microphones. Results of real-time experiments show 10-30 dB noise attenuation, better than those obtained by classical least mean-square techniques such as FX-LMS.

Active noise cancellation using recurrent radial basis function neural networks

Active noise cancellation using neural networks is addressed, with the aim being to derive an architecture/algorithm combination which provides spatiotemporal properties for faster convergence while maintaining a nonlinear dynamics approximation capability. Radial basis function neural networks, with feedback loops connecting the output and input of hidden neurons, are employed. A new learning algorithm suited for active noise cancellation, which is referred to as FX-LRRBF, is proposed. The structure/algorithm is implemented in real-time on a floating point DSP and experimentally carried-out to model the secondary path, which is required for attenuating acoustic noise.

Information extraction from scientific paper using rhetorical classifier

Time constraints often lead a reader of scientific paper to read only the title and abstract of the paper, but reading these parts is often ineffective. This study aims to extract information automatically in order to help the readers get structured information from a scientific paper. The information extraction is done by rhetorical classification of each sentence in a scientific paper. Rhetoric information is the intention to be conveyed to the reader by the author of the paper. This research used corpus-based approach to build rhetorical classifier. Since there was a lack of rethorical corpus, we constructed our own corpus, which is a collection of sentences that have been labeled with rhetorical information. Each sentence represented as a vector of content, location, citation, and meta-discourses features. This collection of feature vectors is used to build rhetorical classifiers by using machine learning techniques. Experiments were conducted to select the best learning techniques for rhetorical classifier. Training set consists of 7239 labeled sentences, and the testing set consists of 3638 labeled sentences. We used WEKA (Waikato Environment for Knowledge Analysis) and LibSVM libraries. Learning techniques being considered were Naive Bayes, C4.5, Logistic, Multi-Layer Perceptron, PART, Instance-based Learning, and Support Vector Machines (SVM). The best performers are the SVM and Logistic classifier with accuracy of 0.51. By applying one-against-all strategy, the SVM accuracy can be improved to 0.60.

Filtered-X Adaptive Neuro-Fuzzy Inference Systems for Nonlinear Active Noise Control

A new method for active noise control is proposed and experimentally demonstrated. The method is based on Adaptive Neuro-Fuzzy Inference Systems (ANFIS), which is introduced to overcome nonlinearity inherent in active noise control. A new algorithm referred to as Filtered-X ANFIS algorithm suitable for active noise control is proposed. Real-time experiment of Filtered-X ANFIS is performed using floating point Texas Instruments C6701 DSP. In contrast to previous work on ANC using computational intelligence approaches which concentrate on single channel and off-line adaptation, this research addresses multichannel and employs online adaptation, which is feasible due to the computing power of the DSP.

H ∞ controller synthesis for Networked Control Systems with time delay system approach

In modern control systems, physical plant, controller, sensors and actuators are difficult to be located at the same place, and hence these components need to be connected over network media. When feedback control system is closed via a communication channel, then the control system is classified as a Networked Control System (NCS). The introduction of communication network gives significant advantages, such as reduced installation and maintenance cost and increased system agility. In the other hand, the use of communication network will lead to intermittent losses or delays of the information and may decrease the performance and cause instability. This paper investigates the problem of H∞ controller synthesis for (NCS). It is well known that H∞-norm constraint can be used to provide a prespecified disturbance attenuation level, and alternatively to analyze robust stability of dynamical system.The NCS is modelled as a time delay system. The physical plant and controller are in continuous time. Two network features are considered: signal transmission delay and data packet dropout. Our objective is focused on the design of state feedback controller which guarantee asymptotic stability of the closed-loop systems. The proposed methods are given in the terms of Linear Matrix Inequality (LMI). If this LMI conditions feasible, a desired controller can be readily constructed. Finally, we consider an unstable system for numerical example. It is shown that the state feedback controller proposed here make the closed-loop system stable with or without input disturbance.

Recent progress in adaptive nonlinear active noise control

In this paper recent progress on adaptive nonlinear active noise control is presented. Particular attention is paid to a new learning algorithm for recurrent neural networks based on Adjoint Extended Kalman Filter that is developed for nonlinear active noise control. The overall control structure for active noise control is constructed using two recurrent neural networks: the first neural network is used to model secondary path of active noise control while the second one is employed to generate control signal. Recent work by authors on combined FIR and neural networks is presented for nolinear active noise control to exploit the benefit of high-order tapped delay line in FIR filter and of the nonlinearity of function expansion. Real-time experiment of the proposed algorithm using Digital Signal Processor is carried-out to show the effectiveness of the method

Nonlinear active noise control using EKF-based recurrent fuzzy neural networks

Active Noise Control (ANC) system is commonly designed and implemented using adaptation algorithm and adaptive control structure. In this paper we present theoretical and experimental result of active noise control system using Recurrent Fuzzy Neural Network (RFNN). RFNN is developed by combining fuzzy logic and neural networks, aimed at producing better control system performance than if we use neural network or fuzzy logic separately. Using a control structure with two multilayer feedforward RFNNs (one RFNN serves as a nonlinear controller while the other one operates as a nonlinear plant model), a recursive least-squares algorithm based on Adjoint Extended Kalman Filter approach is employed for the training of the controller network. Extended Kalman Filter (EKF) algorithm is introduced to develop a new algorithm with faster convergence speed by using nonlinear recursive-least square method. Experimental result using DSP demonstrates effectiveness of the proposed RFNN structure and algorithm to attenuate unwanted noise.

Quaternion-Based attitude control system design of single and cooperative spacecrafts: boundedness of solution approach

It is well known that single equilibrium orientation point in matrix rotation is represented by two equilibrium points in quaternion. This fact would imply nonefficient control effort as well as problem in guaranteeing stability of the two equilibrium points in quaternion. This paper presents a solution to design quaternion-based spacecraft attitude control system whose saturation element is in its control law such that those problems are overcome. The proposed feature of methodology is the consideration on boundedness of solution in the control system design even in the presence of unknown external disturbance. The same methodology is also used to design cooperative spacecrafts attitude control system. Through the proposed method, the most relaxed information-state topology requirement is obtained, that is, the directed graph that contains a directed spanning tree. Some numerical simulations demonstrate effectiveness of the proposed feature of methodology.

Modeling Waste Heat Recovery System of Industrial Ammonia Process Plant Using LPV Identification

Abstract This paper presents modeling of Waste Heat System of an industrial ammonia process plant. Linear Parameter Varying (LPV) identification is utilized to cover changes in process operating conditions, such as start-up, normal operation and shut-down. Recursive Least Square (RLS) based algorithm is employed in the LPV identification process. Experimental input-output signals required for identification process are taken from DCS historian data of the ammonia process plant during plant operations. The resulting LPV model is simulated and validated with respect to the measured data. Promising results are obtained in applying advanced LPV identification to cover variations of process operating conditions in an industrial process plant.