Mu-Song Chen

Dayeh University · Department of Electrical Engineering

Due to the prevalence of IoV technology, big data has increasingly been promoted as a revolutionary development in a variety of applications. Indeed, the received big data from IoV is valuable particularly for those involved in analyzing driver’s behaviors. For instance, in the fleet management domain, fleet administrators are interested in fine-gr...

Usually, most of execution time of match-resolve-act reasoning cycle is spent in the matching phase. This issue has prevented the applicability of rule base systems. In this paper, the parallelism of α- and β-networks constructions in Rete algorithm have been realized on Graphics Processing Unit (GPU). It is possible to speed up the reasoning time...

Gaussian Mixture Models are among the most statistically mature methods which are used to make statistical inferences as well as performing unsupervised clustering. Formally, a gaussian mixture model corresponds to the mixture distribution that represents the probability distribution of observations in the data set. In this paper, a probabilistic c...

A queueing model is generally designed with sufficient capacity or resources to ensure that the system is stable, while preserving quality of service. However, the multi-queue system with finite capacity and timing constraints in an overload condition are more often encountered and discussed in a variety of real-life problems. In such a situation,...

The Long Term Evolution (3GPP-LTE) combines the Multi-input multi-output (MIMO) antenna and the Orthogonal Frequency Division Multiple Access (OFDMA) techniques to accomplish high speed data transmission. Comparisons of zero forcing (ZF), minimum mean square error (MMSE) and sphere decoding (SD) equalization methods with Turbo code are given under...

The setup of interactive communication between arm strength training machine and the people will make exercise and rehabilitation therapy become more friendly. The employment of electromyographic not only can help physical therapy but also can achieve more effective rehabilitation. Both of the system hardware and software of the arm strength traini...

The M/G/1 model is the fundamental basis of the queueing system in many network systems. Usually, the study of the M/G/1 is limited by the assumption of single queue and infinite capacity. In practice, however, these postulations may not be valid, particularly when dealing with many real-world problems. In this paper, a two-stage state-space approa...

Due to timing constraints, uncertain natures, dynamical characteristics, and lack of exact mathematical model of the message scheduling system, an analytical solution of the optimal scheduling sequence is not easy to obtain. Regarding this issue, the concept of message scheduling controller (MSC) is presented in this article. This study makes use o...

This paper presents a simple method to design the speed control system of motor drive for an electric bicycle. By properly selecting the current controller time delay and speed controller time delay, the crossover frequency of current controller and the 3-dB corner frequency of PI controller can be found, respectively. The rotor frequency is design...

The aim of this paper is to design and implement arm strength training machine with electromyographic (EMG) biofeedback based on a microcontroller system. The hardware design based on a microcontroller is analyzed and discussed. The software programming is developed in MPLAB integrated development environment from the Microchip Technology Inc. and...

The interfaces of NCAP and physical sensor network have been realized on the IEEE 1451 middleware for exploring their practicality of IEEE 1451 protocol family based on the well-developed CANOpen and Zigbee. The functions of IEEE 1451 middleware have request-reply command translation, node discovery, surveillance, failure diagnosis, and break detec...

To improve the drawback of inconvenient and bulk weight of exercise machine, an arm strength training machine based on PMSM drive controller which has the advantages of user friendly, light weight, movable, and convenient, is designed and implemented in this paper. The hardware design based on a microcontroller is analyzed and discussed. The softwa...

The Model Driven Architecture (MDA) methodology requires several intelligent operation stages, such as the computation independent model transformation (CIMT), the platform independent model transformation (PIMT), and the platform specific model transformation (PSMT), to progressively transform an abstract model to a physical system. The special Un...

In recent years, the highly growth and development of world economy results in the natural resources being gradually run out and the environment further directly and indirectly being polluted more severe. Consequently, any kind of alternative energy resource have been developed, harvested and designed. An electric bicycle based on a blushless dc mo...

Remote sensing is an important tool in a variety of scientific researches which can help to study and solve many practical environmental problems. Classification of remote sensing image, however, is usually complex in many respects that a lot of different ground objects show mixture distributions in space and change with temporal variations. Theref...

In this paper, we present to utilize Genetic Algorithms (GAs) as tools to model control processes. Two different crossover operators are combined during evolution to maintain population diversity and to sustain local improvement in the search space. In this manner, a balance between global exploration and local exploitation is reserved during genet...

The establishment of interactive communication between arm strength training machine (ASTM) and the people will make the exercise and rehabilitation therapies become more friendly. This paper presents a friendly human interface for ASTM, in such a way to reduce the occurrence for improper operation of exercise machine. The system hardware consists...

In queueing system, the mean waiting times of messages are important measures to characterize the quality of service (QoS) under various requirements. In a time-critical system, message transactions which cannot meet deadline constraints might lead to catastrophic consequences. Currently, the waiting time estimations using the first-come-first-serv...

In recent years, the highly growth and development of world economy results in the natural resources being gradually run out and the environment further directly and indirectly being polluted more severe. Consequently, any kind of alternative energy resource have been developed, harvested and designed. An electric bicycle based on a blushless dc mo...

This paper presents the brushless DC (BLDC) motor drive with direct torque control (DTC) for a washing machine over variable speed range. The system hardware and software programming are analyzed and designed. The flux controller and torque controller are derived and realized accordingly. A prototype of BLDC motor drive based on the DSP controller...

A high power factor brushless DC motor (BLDC) is designed and implemented in this paper. In order to increase the power factor of a BLDC drive, an active power factor controller is employed to improve the high input current harmonics generated from the power diodes as well as the switching of the inverter. The detailed design of the power factor co...

Transductive support vector machine (TSVM) is one kind of transductive inference process, which combines labeled samples with unlabelled samples to derive the decision rules for classification tasks. Compared with the classical SVM, the transductive SVM is more robust and can achieve better performance. However, there are some disadvantages still b...

In order to cultivate talents and raise the design ability for power converters at engineering education on campus, the design methodology of a flyback dc/dc converter is introduced step by step in this paper. First of all, the fundamental of switching power converters is analyzed and described in this paper. Later, the power converter is modeled a...

In this paper, we propose an innovative approach for face recognition based on collaborative image similarity assessment (CISA). In the proposed method, the test sample is first represented by a linear combination of all the training samples for each face class. The classification task is then conducted using the similarity measures including struc...

A high power factor brush less DC motor (BLDC) is designed and implemented in this paper. In order to increase the power factor of a BLDC drive, an active power factor controller is employed to improve the high input current harmonics created from the power diodes as well as the switching of the inverter. The detailed design of the power factor con...

The Controller Area Network (CAN) is a communication bus for message transaction in real-time environments. A real-time system typically consists of several classes of messages and a scheduler is responsible to allocate network resources to fulfill timing constraints. Given sufficient bandwidth, the static scheduling algorithms can meet the bounded...

The M/M/1 model is the most frequently used model in queueing problems. The research on the M/M/1 model with single queue and finite/infinite buffer length is already well developed. However, combinations of multiple queues with limited buffer capacity are rarely discussed. The reason for this deficiency is the lack of queue selection probability w...

In this paper, we present a novel method for gait recognition of different people groups based on Fourier descriptors (FDs) and support vector machine (SVM). The proposed method involves the procedures of background modeling, extraction of gait silhouettes by background subtraction, shadow removal, representation of gait silhouettes using the FDs,...

In this paper, we propose a method of eye detection based on skin color analysis under varying illumination. The proposed method consists of several phases, including color conversion, skin color segmentation and face mask calculation, facial feature extraction and eye candidate determination, and detection of human eyes. To eliminate the effect of...

The design of a flyback converter based on simulation is adopted in this paper. From the simulation results, a flyback converter is realized and implemented. The experimental results show the verification for the design and implementation of the flyback converter. This paper concludes that the design of a flyback converter based on the simulation,...

This paper presents an active power factor controller for a PMSM drive to improve the high input current harmonics created from the power diodes as well as the switching of the inverter. The detailed design is analyzed and implemented by a motor drive prototype. The experimental results verify the feasibility of PMSM motor drive with designed APFC...

The M/G/1 model with multiple queues and finite capacity is very common and can be found very often in our daily life. Interesting measures of this model can be the waiting times of messages, mean queue lengths, etc. Usually, these problems are solved by the transformed approach to determine model parameters. Instead of the transformed approach, we...

In time critical system, message scheduling plays an important role to arbitrate fair service among all competing messages, where messages are conditioned on different timing constraints. There already exist many algorithms, including static and dynamic scheduling, to resolve these problems. In this paper, we extend our recent work by presenting a...

The devised multi-agent development platform has been compo sited by abstraction layers of PAL, DAL, MWL, and MAL. Their functionalities have agent communication language interpreter, interaction protocol, and abstraction hardware platform that are in tum applied to support the realization of intelligent agent nodes for a multi-agent application sy...

Due to limited resource contentions and deadline constraints, messages on the controller area network (CAN) are competing for service from the common resources. This problem can be resolved by assigning priorities to different message classes to satisfy time-critical applications. Actually, because of the fluctuation of network traffic or an ineffi...

Message scheduling is a critical research issue in the Controller Area Network (CAN). Messages that fail to meet predefined deadlines may deteriorate system performance significantly. In this paper, we propose an online Message-Scheduling Controller (MSC), which is realised by the Radial Basis Function (RBF) network. The MSC exploits two novel lear...

Optimal message scheduling is one of the key issues in the field of controller area network (CAN) bus system. There are numerous approaches related to this issue. Most of them are essentially based on priority-based strategies. In 1, we utilized Radial Basic Function (RBF) network 2 as a message scheduling controller to dynamically schedule message...

In this paper, we propose a generalized fuzzy inference system (GFIS) in noise image processing. The GFIS is a multi-layer neuro-fuzzy structure which combines both Mamdani model and TS fuzzy model to form a hybrid fuzzy system. The GFIS can not only preserve the interpretability property of the Mamdani model but also keep the robust local stabilit...

Ultrasound has many important applications in nondestructive evaluation. Since ultrasonic signals can penetrate many materials, they are widely used for detection of flaws. However, the use of ultrasound is limited by interference, such as geometric interface, coarse grain, mode conversion and ringing. Hence it is practically difficult to perform e...

A technique for improving the topology of a trained neural network, used for an inversion or classification problem, is presented. The technique models the multilayer perceptron as a power series, which allows us to (1) remove units from the network which are well-approximated by zero-degree or first-degree polynomials, (2) measure the effect of re...

In this paper, a complex-valued neural network based on the Kalman filter is presented for channel equalization in a communication system. The complex-valued decoupled extended Kalman filter algorithm is derived which provides a faster convergence rate and better performance than those of complex back-propagation algorithms. Computer simulation res...

Fuzzy model identification is an application of fuzzy inference system for identifying unknown functions, for a given set of sampled data. The most important thing for fuzzy identification task is to decide the parameters of membership functions (MFs) used in fuzzy systems. A lot of efforts (Chung and Lee, 1994; Jang, 1993; Sun and Jang, 1993) have...

One of the important problems to be solved for fuzzy inference systems is to tune the free parameters for solving the given task. In this paper, we propose to combine the RPROP adaptive learning algorithm, which is much faster than the gradient descent type, with the recursive-least-squares-error technique for tuning parameters of fuzzy membership...

Classification of terrain cover using polarimetric radar is an area of considerable interest and research. This paper describes the application of neural network approach to the classification of a fully polarimetric SAR image. The structure and connection weights of the network are decided based on the cascade correlation algorithm. The procedures...

We compare several popular training algorithms for tuning parameters of fuzzy membership functions (MFs). The algorithms compared are gradient descent (GD), resilient propagation (RPROP), Quickprop (QP), and Levenberg-Marquardt (LM) algorithms. These algorithms are combined with RLSE (recursive least squares estimate) to improve the efficiency of a...

The paper presents a constructive method, which combines the architectural feature of the cascade correlation algorithm (CCA) and genetic algorithms for building the neural network and training the corresponding connection weights. Comparisons between the proposed method and the cascade correlation algorithm are made by applying it to SAR image cla...

ANALYSIS AND DESIGN OF THE MULTI-LAYER PERCEPTRON USING POLYNOMIAL BASIS FUNCTIONS Publication No._________ Mu-Song Chen, Ph.D. The University of Texas at Arlington, 1991 Supervising Professor : Michael. T. Manry In this dissertation, the theory of polynomial basis functions is developed as a means for the design and analysis of the multi-layer per...

Presents a genetic algorithm based system for evolving neural networks. New genetic operators, which combine a heuristic approach and pseudo gradient information, are designed to enhance the performance of genetic algorithms. In this way, the extension or contraction of search region can be more adaptive to the characteristics of the neural network...

Genetic algorithms (GAs) are adaptive methods, which can be employed to solve search and optimization problems. The GA relies on genetic operators to exchange gene between individuals for generating better offspring. An important issue to execute GA efficiently is to maintain population diversity and to sustain local improvement in the search stage...

Table-form document recognition has many applications in office automation. An algorithm is proposed in the paper for automatic form processing. A high order correlation method was originally developed for point target detection in three-dimensional space. It computes the spatio-temporal cross-correlations of consecutive data to extract track infor...

A technique for improving the topology of a trained neural network, used for an inversion or classification problem, is presented. The technique models the multilayer perceptron as a power series, which allows us to (1) remove units from the network which are well-approximated by zero-degree or first-degree polynomials, (2) measure the effect of re...

We show that the pattern storage capability of the Gabor polynomial is much higher than the commonly used lower bound on multi-layer perceptron (MLP) pattern storage. We also show that multi-layer perceptron networks having second and third degree polynomial activations can be constructed which efficiently implement Gabor polynomials and therefore...

In this letter, we present a new technique for modeling the multilayer perceptron (MLP) neural network, in which input and hidden units are represented by polynomial basis functions (PBF's). The MLP output is expressed as a linear combination of the PBF's and can therefore be expressed as a polynomial function of its inputs. Thus the MLP is isomorp...

A technique for analyzing multilayer perceptron (MLP) neural networks is presented in which each hidden unit is modeled as a power series of the net function. This allows (1) pruning useless hidden units, (2) measuring the effect of removing a hidden layer, and (3) determining the degree of the overall polynomial discriminant (PD) which approximate...

In a recent paper by M. Chen and M. Maury (1990), it was shown that multilayer perceptron neural networks can be used to form products of any number of inputs, thereby constructively proving universal approximation. This result is extended, and a method for the analysis and synthesis of single-input, single-output neural subnetworks is described. G...

Presents a technique for analyzing backpropagation neural networks. Each hidden unit in the network is modeled as a power series of the net function. This approach allows determination of the degree of the overall polynomial discriminant, which approximates the network, potentially revealing the complexity of the decision boundary for the training...

Develops a polynomial basis function approach for modeling BP (backpropagation) neural networks. This method leads directly to a constructive proof of the BP approximation theorem. In addition, the basis vector approach provides a means to synthesize the BP neural network output as a polynomial function. An algorithm for pruning the useless basis v...

Summary form only given, as follows. A method for approximating an existing N-input backpropagation neural network (NN) using an N-dimensional (N-D) polynomial discriminant (PD) function is discussed. First, the hidden unit activation functions are approximated by polynomials. Then, after multiplying out the resulting composition of polynomials, th...

In this dissertation, the theory of polynomial basis functions is developed as a means for the design and analysis of the multi-layer perceptron (MLP) neural networks. Methods and algorithms are presented for designing the MLP network system using polynomial models. The theory enables us to develop an approximation theorem for the MLP network, to m...

Several investigators have constructed back-propagation (BP) neural networks by assembling smaller, pre-trained building blocks. This approach leads to faster training and provides a known topology for the network. The authors carry this process down one additional level, by describing methods for mapping given functions to sub-blocks. First, polyn...

A representation theorem is developed for backpropagation neural networks. First, it is assumed that the function to be approximated, F ( x ) for the vector x , is continuous and has finite support, so that it can be approximated arbitrarily well by a multidimensional power series. The activation function, sigmoid or otherwise, is then approximated...

Due to limited resource contentions and deadline constraints, messages on the Controller Area Network (CAN) are competing for service from the common resources. Actually, because of the fluctuation of network traffic or an inefficient use of resources, the existing static or dynamic priority policies may not guarantee flexibility for different kind...

The recognition of table form documents is useful in office automation and file management. This paper presents a new approach for automatic document classification using high order correla- tion (HOC) method. HOC was originally used to recursively compute the cross-correlations be- tween consecutive data in order to extract moving target tracks in...

Face recognition is a great challenge for face images of underexposure. In this paper, a novel method for face recognition based on the transformation of histogram matching is proposed to tackle such difficulties. The histogram matching method is first used to enhance the face features such as eyes and mouth to be more discriminative. Then, dimensi...