• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.504-507
• /
• 1997

It is well known that Variable Structure Controller(VSC) is robust to parameters variation and disturbance but its performance depends on the design parameters such as switching gain and slope of sliding surface. This paper proposes a more robust VSC that is composed of local VSC's. Each local VSC considers the local system dynamics with narrow parameter variation and disturbance. First we optimize the local VSC's by use of Evolution Strategy, and next we use Artificial Neural Network to generalize the local VSC's and construct the overall VSC in order to cover the whole range of parameter variation and disturbance. Simulation on BLDC motor current control shows that the proposed VSC is superior to the conventional VSC.

• 제어로봇시스템학회:학술대회논문집
• /
• 1997.10a
• /
• pp.1820-1823
• /
• 1997

For the Cartesian space position controlled robot, it is required to have the accurate mapping from the Cartesian space to the joint space in order to command the desired joint trajectories correctly. since the actual mapping from Cartesian space to joint space is obtained at the joint coordinate not at the actuator coordinate, uncertainty in Jacobian can be present. In this paper, two feasible neural network schemes are proposed to compensate for the kinematic Jacobian uncertainties. Uncertainties in Jacobian can be compensated by identifying either actuator Jacobian off-line or the inverse of that in on-line fashion. the case study of the stenciling robot is examined.

• Journal of Institute of Control, Robotics and Systems
• /
• v.2 no.2
• /
• pp.108-114
• /
• 1996

We present proofs of the stability and convergence of Self-organizing feature map (SOFM) neural network with time-invarient learning rate and binary reinforcement function. One of the major problems in Self-organizing feature map neural network concerns with learning rate-"Kalman Filter" gain in stochsatic control field which is monotone decreasing function and converges to 0 for satisfying minimum variance property. In this paper, we show that the stability and convergence of Self-organizing feature map neural network with time-invariant learning rate. The analysis of the proposed algorithm shows that the stability and convergence is guranteed with exponentially stable and weak convergence properties as well.s as well.

• Journal of Institute of Control, Robotics and Systems
• /
• v.9 no.5
• /
• pp.361-367
• /
• 2003

We propose an algorithm to detect the face region using a variable ellipsoidal mask and a neural network. Since outlines of human faces are similar to ellipsoid, the ellipsoidal mask that has the fixed ratio of major and minor axis can be used to detect the candidate area. The positions of eyes and lips are extracted in this candidate area, and then the morphological analysis is applied to make features which are consist of six parameters, such as the geometrical ratio of eyes and lips. A back-propagation neural network is used as a classifier to determine the most possible face region. The experimental result is conducted to verify its efficiency compared with those of previous works.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.1334-1337
• /
• 2004

This paper presents a novel method to speed up neural network (NN) based face detection systems. NN-based face detection can be viewed as a classification and search problem. The proposed method formulates the search problem as an integer nonlinear optimization problem (INLP) and expands the basic particle swarm optimization (PSO) to solve it. PSO works with a population of particles, each representing a subwindow in an input image. The subwindows are evaluated by how well they match a NN-based face filter. A face is indicated when the filter response of the best particle is above a given threshold. To achieve better performance, the influence of PSO parameter settings on the search performance was investigated. Experiments show that with fine-adjusted parameters, the proposed method leads to a speedup of 94 on 320${\times}$240 images compared to the traditional exhaustive search method.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.1272-1277
• /
• 2003

This paper presents a method for combining classifier which is constructed by fuzzy and neural network classifiers and uses classifier fusion algorithms and selection algorithms. The input space of combing classifier is divided by the extended hyperbox region proposed in this paper to guarantee non-overlapped data property. To fuse the fuzzy classifier and the neural network classifier, we propose the fusion parameter for the overlapped data. In addition, the adaptive learning algorithm also proposed to maximize classifier performance. Finally, simulation examples are given to illustrate the effectiveness of the method.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.2461-2466
• /
• 2003

This paper presents a simple sensor fusion algorithm using neural network for navigation of mobile robots with obstacle avoidance and trap recovery. The multiple sensors input sensor data to the input layer of neural network activating the input nodes. The multiple sensors used include optical encoders, ultrasonic sensors, infrared sensors, a magnetic compass sensor, and GPS sensors. The proposed sensor fusion algorithm is combined with the VFH(Vector Field Histogram) algorithm for obstacle avoidance and AGPM(Adaptive Goal Perturbation Method) which sets adaptive virtual goals to escape trap situations. The experiment results show that the proposed low-level fusion algorithm is effective for real-time navigation of mobile robot.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.118-121
• /
• 2003

In the conventional boundary estimation in EIT (Electrical Impedance Tomography), the interface between anomalies and background is expressed in usual as Fourier series and the boundary is reconstructed by obtaining the Fourier coefficients. This paper proposes a method for the boundary estimation, where the boundary of anomaly is approximated as the interpolation of front points located discretely along the boundary and is imaged by tracking the points in the region of interest. In the solution to the inverse problem to estimate the front points, the multi-layer neural network is introduced. For the verification of the proposed method, numerical experiments are conducted and the results indicate a good performance.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.1305-1308
• /
• 2003

A new technique is presented to construct a predictive model of plasma etch process. This was accomplished by combining a backpropagation neural network (BPNN) and a genetic algorithm (GA). The predictive model constructed in this way is referred to as a GA-BPNN. The GA played a role of controlling training factors simultaneously. The training factors to be optimized are the hidden neuron, training tolerance, initial weight magnitude, and two gradients of bipolar sigmoid and linear functions. Each etch response was optimized separately. The proposed scheme was evaluated with a set of experimental plasma etch data. The etch process was characterized by a $2^3$ full factorial experiment. The etch responses modeled are aluminum (A1) etch rate, silica profile angle, A1 selectivity, and dc bias. Additional test data were prepared to evaluate model appropriateness. The GA-BPNN was compared to a conventional BPNN. Compared to the BPNN, the GA-BPNN demonstrated an improvement of more than 20% for all etch responses. The improvement was significant in the case of A1 etch rate.

• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.479-482
• /
• 2003

We propose a real time gas monitoring system for classifying various gases with different concentrations. Using thermal modulation of operating temperature of two sensors, we extract patterns of gases from the voltage across the load resistance. We adopt the relative resistance as a pre-processing method and an ART2 neural network as a pattern recognition method. The proposed method has been implemented in a real time embedded system with tin oxide gas sensors, TGS 2611, 2602 and an MSP430 ultra-low power microcontroller in the test chamber.