Browsing by Author "Hu, JS"

Now showing 1 - 20 of 20

Active impedance control of linear one-dimensional wave equations
(1999-02-15)
Hu, JS
;
Lin, JF
;
電控工程研究所
;
Institute of Electrical and Control Engineering
Active impedance control of linear one-dimensional wave equations is investigated. The proposed control algorithms are based on the concepts of wave propagation and impedance matching. Two control objectives are considered. The first objective is to obtain total reflection, and the second objective is to achieve total absorption (e.g. matched impedance). Both control laws utilize some interesting properties of the wave-type partial differential equation and do not require information about the disturbance and boundary conditions. The resulting controllers, although boundary-independent, contain an infinite number of poles on the imaginary axis and the closed-loop systems are not internally stable. A simple modification is added to the control laws and the stability is analysed. An example of active sound cancellation in ducts with a moving medium is given to demonstrate one of the control algorithms.
Active noise cancellation in ducts using internal model-based control algorithms
(1996-03-01)
Hu, JS
;
交大名義發表
;
電控工程研究所
;
National Chiao Tung University
;
Institute of Electrical and Control Engineering
A feedback control algorithm based on the internal model principle is proposed to solve the active noise control (ANC) problem in ducts. Because of its infinite open-loop gain at selected frequencies, the controller is able, at the cancellation speaker position, to block noise transmission of these frequencies. In other words, the speaker acts like a totally reflective boundary, Stability analysis of the closed-loop system is performed using the Nyquist criterion, Experimental results confirm the theoretical analysis, showing an average noise level reduction of 30 to 40 dB at the target frequencies.
An object-oriented interface for control and automation software in open architecture systems
(2000)
Ou, YC
;
Horng, SH
;
Hu, JS
;
電控工程研究所
;
Institute of Electrical and Control Engineering
In this paper, a unified framework for control system and automation software design is proposed. Under this framework, supervisory synthesis methodology with DES specification can be used in designing control software not only for a single machine but also for machines' integration. In the structural design, basic classes of our proposed framework were model by Object-Oriented Analysis (OOA) and design techniques presented by the UML (Unified Modeling Language) notation. Secondly, in the mechanistic design, design pattern was used to improve the modality and flexibility of the framework. At last, distributed Object technique using CORBA was used for implementation, and a simple example was presented.
Analysis and design of 1-bit noise-shaping quantizer using variable structure control approach
(2004)
Hu, JS
;
Yu, SH
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper introduces a straightforward analysis of a 1-bit noise-shaping quantizer using variable structure control methods. The method using the concept of equivalent control to access the bound on state signals when the sliding condition is reached. Simulation examples and an experimental circuit using off-the-shelf electronic components are given to demonstrate the theoretical analysis.
Analysis and design of a single-bit noise-shaping quantizer
(2003)
Yu, SH
;
Hu, JS
;
電控工程研究所
;
Institute of Electrical and Control Engineering
Noise-shaping quantization is the use of feedback to improve the effective resolution of a coarse quantizer. Based on the theory of sliding modes, this study presents a simple and effective approach to analyzing and designing a single-bit noise-shaping quantizer. The derived stability conditions and accurate estimates of the state bounds make an exact design possible.
Application of model-matching techniques to feedforward active noise controller design
(1998-01-01)
Hu, JS
;
Yu, SH
;
Hsieh, CS
;
電控工程研究所
;
Institute of Electrical and Control Engineering
In this paper, three digital model-matching techniques in H-2, H-infinity, and l(1) performance measures are applied to design digital feedforward controllers for active noise cancellation in ducts. Different measures account for different optimization objectives in terms of physical signals. The distributed nature and high-bandwidth requirements of the control system result in a large set of parameters in plant description and these design techniques proved to be useful insolving the controllers numerically. Experiments were conducted using a floating-point digital signal processor that produced broad-band noise reduction. Design variations and noise reduction effects in terms of human perception are also discussed. It is experimentally proved that using model-matching designs, the causality principle originally raised by Paul Lueg does not have to be satisfied in order to actively reduce the noise level.
Application of model-matching techniques to feedforward active noise controller design
(1998-01-01)
Hu, JS
;
Yu, SH
;
Hsieh, CS
;
電控工程研究所
;
Institute of Electrical and Control Engineering
In this paper, three digital model-matching techniques in H-2, H-infinity, and l(1) performance measures are applied to design digital feedforward controllers for active noise cancellation in ducts. Different measures account for different optimization objectives in terms of physical signals. The distributed nature and high-bandwidth requirements of the control system result in a large set of parameters in plant description and these design techniques proved to be useful insolving the controllers numerically. Experiments were conducted using a floating-point digital signal processor that produced broad-band noise reduction. Design variations and noise reduction effects in terms of human perception are also discussed. It is experimentally proved that using model-matching designs, the causality principle originally raised by Paul Lueg does not have to be satisfied in order to actively reduce the noise level.
Assessing weights of product attributes from fuzzy knowledge in a dynamic environment
(2004-04-01)
Hu, YC
;
Hu, JS
;
Chen, RS
;
Tzeng, GH
;
科技管理研究所
Fuzzy knowledge of consumers' frequent purchase behaviors can be extracted from transaction databases. To effectively supporting decision makers, it is necessary to use fuzzy knowledge to assess weights or degrees of consumers' attentiveness to product attributes. From the standpoint of habitual domains, frequent purchase behaviors can be viewed as ideas that are contained in the reachable domain of customers. In addition, this reachable domain is changeable with time, due to the dynamic environment. This paper thus proposes a two-phase learning method with adaptive capability. The first phase builds a fuzzy knowledge base by discovering frequent purchase behaviors from transaction databases; the second phase finds weights of product attributes by a single-layer perceptron neural network. Indeed, customers are asked to evaluate alternatives and attributes through questionnaire. Then, each alternative can be transformed into a piece of input training data for the neural network by the fuzzy knowledge base and part-worths of attributes' levels. After completing the training task, we can find weights from connection weights. Simulation results demonstrate that the proposed methods can use fuzzy knowledge to effectively find customers' attentive degrees of attributes. (C) 2002 Elsevier B.V. All rights reserved.
Asymptotic rejection of periodic disturbances with fixed or varying period
(2001-09-01)
Yu, SH
;
Hu, JS
;
電控工程研究所
;
Institute of Electrical and Control Engineering
A constructive derivation of repetitive control is obtained, through attempting to derive a control law for asymptotic rejection of periodic disturbances. This derivation not only reveals a close relationship between iterative operator inversion and repetitive control, but also suggests a unified design method for a learning control algorithm. Also, based on the observation, digital repetitive control can be generalized to reject periodic disturbance whose period is not exactly an integer multiple of the sampling interval, This study introduces a delay filter in the digital repetitive control law, which optimally interpolates the signal between samples, thus effectively reconstructing the signal of the previous period and making the learning process of repetitive control successful. The proposed optimal delay filter can be updated easily according to different signal periods. Thus it is specifically suitable for on-line timing when the signal period is changing. Compared with the available tuning methods, the proposed tuning method has excellent steady-state performance while maintaining fast transient and system robustness. The simulations on active noise cancellation within a duct confirm the superiority of this tuning method.
Controller design for active noise cancellation headphones using experimental raw data
(2001-12-01)
Yu, SH
;
Hu, JS
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This work presents a practical method of designing controllers for active noise cancellation (ANC) headphones. Without attempting system identification and perturbation modeling, the headphone system is directly described by a set of frequency-response data. In frequency domain, the controller synthesis problem is formulated as a constrained optimization problem, where the Ha performance objective is minimized with various frequency-dependent constraints. The fixed-order robust controller is thus designed to achieve maximum noise attenuation with acceptable stability margins. Further, the method is able to accurately constrain noise amplification outside control bandwidth due to waterbed effect of nonminimum phase plants. This feature is very important for the ANC design to maintain an overall quality of noise reduction.
Distributed embedded real-time Ethernet platform for robots control
(2005)
Wu, LW
;
Hu, JS
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This work proposes a distributed embedded control system architecture based on real-time Ethernet, to provide a uniform environment for integration of real-time control and information exchange. A novel TCP/IP-compatible Hardware Real-Time Protocol (HRTP) is proposed for real-time capability. This structure focuses on the design of the control network inside a robot. Using embedded Ethernet simplifies the integration of a real-time communications network of an intelligence-type robot system. Furthermore, a quadruped robot was designed and implemented to test the proposed framework.
Feedforward active noise cancellation in ducts with on-line identification of sensor and actuator dynamics
(2000)
Hu, JS
;
Lin, CF
;
Hsiao, TS
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper proposed an adaptive control algorithm using multiple feedforward sensors and the wave propagation principle to eliminate noise in the downstream area of a duct. The algorithm is able to separate the dynamics of the sensor and actuator from the duct. Two experiments were conducted to show the effectiveness of the control law.
Feedforward active noise controller design in duets without independent noise source measurements
(2000-05-01)
Hu, JS
;
Lin, JF
;
電控工程研究所
;
Institute of Electrical and Control Engineering
Feedforward control architectures have been widely used in active noise control systems to achieve broad-band noise reductions. The basic principle of the control algorithm requires that the feedforward signal, usually the noise source, be independent from the actuator's output. Failure to meet this requirement implies that the overall system contains an acoustic feedback loop and stability and robustness issues become important. This paper investigates this problem in a one-dimensional sound field (e.g., ducts) using a unidirectional sound wave as the feedforward signal. Controller design based on a distributed parameter model is studied. Several experiments are conducted to illustrate the design procedure as well as verify the effectiveness of broad-band noise reductions.
Feedforward and feedback control strategy for active noise cancellation in ducts
(1996-06-01)
Hu, JS
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper presents the theoretical work about active noise cancellation in ducts. The proposed control system is designed based on the assumption of a one-dimensional sound field. The controller consists of a feedforward block which serves as a noise observer. The feedback portion of the control algorithm is to minimize residual disturbances. Closed-loop stability of the MIMO (multiple-input-multiple-output) system is analyzed and the result shows that the dynamic influenced by the space-feedforward and feedback controllers can be decoupled. Both semi-infinite and finite-length ducts are considered in this study and simulation examples are given to illustrate the effectiveness of the proposed controllers.
Feedforward and feedback control strategy for active noise cancellation in ducts
(1996-06-01)
Hu, JS
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper presents the theoretical work about active noise cancellation in ducts. The proposed control system is designed based on the assumption of a one-dimensional sound field. The controller consists of a feedforward block which serves as a noise observer. The feedback portion of the control algorithm is to minimize residual disturbances. Closed-loop stability of the MIMO (multiple-input-multiple-output) system is analyzed and the result shows that the dynamic influenced by the space-feedforward and feedback controllers can be decoupled. Both semi-infinite and finite-length ducts are considered in this study and simulation examples are given to illustrate the effectiveness of the proposed controllers.
Frequency domain microphone array calibration and beamforming for automatic speech recognition
(2005-09-01)
Hu, JS
;
Cheng, CC
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This investigation proposed two array beamformers SPFDBB (Soft Penalty Frequency Domain Block Beamformer) and FDABB (Frequency Domain Adjustable Block Beamformer). Compared with the conventional beamformers, these frequency-domain methods can significantly reduce the computation power requirement in ASR (Automatic Speech Recognition) based applications. Like other reference signal based techniques, SPFDBB and FDABB minimize microphone's mismatch, desired signal cancellation caused by reflection effects and resolution due to the array's position. Additionally, these proposed methods are suitable for both near-field and far-field environments. Generally, the convolution relation between channel and speech source in time domain cannot be modeled accurately as a multiplication in the frequency domain with a finite window size, especially in ASR applications. SPFDBB and FDABB can approximate this multiplication by treating several frames as a block to achieve a better beamforming result. Moreover, FDABB adjusts the number of frames on-line to cope with the variation of characteristics in both speech and interference signals. A better performance was found to be achievable by combining these methods with an ASR mechanism.
Optimal MISO repetitive control system design using mixed time and frequency domain criteria
(1998-03-01)
Hu, JS
;
Yu, SH
;
Huang, SC
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper presents an optimal repetitive controller design approach when both frequency and time domains criteria are considered. These criteria include limited control force, convergence rate, and robustness. In particular, the design formulation is based on MISO (multiple-input-single-output) plants which aim at solving nodal point problems encountered in flexible dynamic systems. Nodal points represent physical constraints for feedback control, especially when their frequencies are within the control bandwidth. If multiple control paths with distinct nodal frequencies are available, the proposed method is able to design controllers to compensate the effect and reach certain design objectives. The ellipsoid algorithm is used to calculate controllers' parameters. Experiments were conducted to demonstrate the proposed method.
Optimal synthesis of a fractional delay FIR filter in a reproducing kernel Hilbert space
(2001-06-01)
Yu, SH
;
Hu, JS
;
電控工程研究所
;
Institute of Electrical and Control Engineering
Based on a bandlimited signal model, the optimal fractional delay finite impulse response (FIR) filter and corresponding interpolating error bound is derived in a reproducing kernel Hilbert space. The resulting optimal filtering is a projection onto a prescribed finite dimensional subspace, The connection of this filtering accuracy to the delay time and the filter order is investigated via error analysis.
Robust beamforming of microphone array using H infinity adaptive filtering technique
(2006-03-01)
Hu, JS
;
Liu, WH
;
Cheng, CC
;
電控工程研究所
;
Institute of Electrical and Control Engineering
In ASR (Automatic Speech Recognition) applications, one of the most important issues in the real-time beamforming of microphone arrays is the inability to capture the whole acoustic dynamics via a finite-length of data and a finite number of array elements. For example, the reflected source signal impinging from the side-lobe direction presents a coherent interference, and the non-minimal phase channel dynamics may require an infinite amount of data in order to achieve perfect equalization (or inversion). All these factors appear as uncertainties or un-modeled dynamics in the receiving signals. Traditional adaptive algorithms such as NLMS that do not consider these errors will result in performance deterioration. In this paper, a time domain beamformer using H-infinity filtering approach is proposed to adjust the beamforming parameters. Furthermore, this work also proposes a frequency domain approach called SPFDBB (Soft Penalty Frequency Domain Block Beamformer) using H-infinity filtering approach that can reduce computational efforts and provide a purified data to the ASR application. Experimental results show that the adaptive H-infinity filtering method is robust to the modeling errors and suppresses much more noise interference than that in the NLMS based method. Consequently, the correct rate of ASR is also enhanced.
Robust environmental change detection using PTZ camera via spatial-temporal probabilistic modeling
(2005)
Hu, JS
;
Su, TM
;
電控工程研究所
;
Institute of Electrical and Control Engineering
This paper presents a novel approach to detect environmental change using a Pan-Tilt-Zoom (PTZ) camera. Conventional pixel-based comparison methods require precision PTZ mechanisms and a large memory for background scene storage. The proposed method represents each scene captured by the camera using a spatial-temporal statistical model with much less parameters. Although the detail of the environment covered by the camera is lost, this model is efficient and robust in detecting the changes in the environment. The effectiveness of the method is demonstrated by experiments in a complex environment and relationship between the environmental change and the feature space parameters are observed.