Papers by Author: Kuo Lan Su

Paper TitlePage

Authors: Yung Chin Lin, Yung Chien Lin, Kun Song Huang, Kuo Lan Su
Abstract: A novel application to mechanical optimal design is presented in this paper. Here, an evolutionary algorithm, called mixed-integer differential evolution (MIHDE), is used to solve general mixed-integer optimization problems. However, most of real-world mixed-integer optimization problems frequently consist of equality and/or inequality constraints. In order to effectively handle constraints, an evolutionary Lagrange method based on MIHDE is implemented to solve the mixed-integer constrained optimization problems. Finally, the evolutionary Lagrange method is applied to a mechanical design problem. The satisfactory results are achieved, and demonstrate that the evolutionary Lagrange method can effectively solve the optimal mechanical design problem.
Authors: Kuo Lan Su, Sheng Wen Shiau, Yi Lin Liao, J.H. Guo
Abstract: The paper develops gas detection modules for the intelligent building. The modules use many gas sensors to detect environment of the home and building. The gas sensors of the detection modules are classified two types. One is competitiveness gas detection module, and uses the same sensors to detect gas leakage. The other is complementation gas detection module, and uses variety sensors to classify multiple gases. The paper uses Bayesian estimation algorithm to be applied in competitiveness gas detection module and complementation gas detection module, and implement the proposed algorithm to be nice for variety gas sensor combination method. In the competitiveness gas detection module, we use two gas sensors to improve the proposed algorithm to be right. In the complementation gas detection module, we use a NH3 sensor, an air pollution sensor, an alcohol sensor, a HS sensor, a smoke sensor, a CO sensor, a LPG sensor and a nature gas sensor, and can classify variety gases using Bayesian estimation algorithm. The controller of the two gas detection modules is HOLTEK microchip. The modules can communicate with the supervised computer via wire series interface or wireless RF interface, and cautions the user by the voice module. Finally, we present some experimental results to measure know and unknown gas using the two gas detection modules on the security system of the intelligent building.
Authors: J. Hung Guo, Kuo Lan Su, Yi Lin Liao
Abstract: The article presents a power detection and prediction system (PDPS) using fusion algorithms to be applied in the on-line power monitoring of the target device. The system contains multiple power detection units, a data integration unit, a target device, a power source and a main controller. Each power detection unit measures the assigned power source on real-time, and uses four current sensors to measure the current variety. We use fusion algorithms to be applied in current and voltage detection. We can calculate the real-time power values according to the estimated values of current and voltage measurement values. The main controller predicts the power loading for each power detection unit using auto-regression algorithm, and calculates the error value between the prediction value and the measurement value for each detection value, and compares the relation value on various condition.
Authors: Kuo Lan Su, Yung Chin Lin, Yi Lin Liao, J. Hung Guo
Abstract: The article develops a vision based auto-recharging system for mobile robots, and programs a new docking processing to enhance successful rate. The system contains a docking station and a mobile robot. The docking station contains a docking structure, a control device, a charger and a detection device and a wireless RF interface. The mobile robot contains a power detection module (voltage and current), an auto-switch, a wireless RF interface, a control system and a camera. The docking structure is designed with one active degree of freedom and two passive degrees of freedom. The active degree of freedom can move forward to contact the recharging connect points that are arranged in the mobile robot. The two passive degrees of freedom can rotation in the Z-axis and use compression spring moving on various docking condition. In image processing, the mobile robot uses a webcam to capture the real-time image; and transmits the image signal to the computer via USB interface, and uses Otsu algorithm to recognize the position of the docking station. In the experiment results, the system had been successfully guided the mobile robot moving to the docking station using the proposed method.
Authors: Kuo Lan Su, Jr Hung Guo, Kuo Hsien Hsia
Abstract: The purpose of this paper is to develop an intelligent mobile robot using image processing technology. The mobile robot is composed of a visual tracking system, a loading platform, a balance control system, a PC-based controller, four ultrasonic sensors and a power system. We develop a PC based control system for image processing and path planning. The mobile robot can track a moving target and adjust the loading platform by the balance control system simultaneously. The Image processing based on OpenCV use two different tracking methods, MTLT (Match Template Learning Tracking) and TLD (Tracking, Learning and Detection), to track moving targets. The efficiencies of both methods for tracking the moving target on the mobile robot are compared in this paper. The loading platform control system uses HOLTEK Semiconductor Company's HT66F Series 8-bit microprocessor as the processor, and receives the feedback data from the FAS-A inclinometer sensor. The controller of the loading platform uses the PID control law according to the feedback signals of the inclinometer sensor, and controls the rotation speed of the platform motor to tune the balance level. Keywords— Intelligent mobile robot, Image processing, OpenCV, MTLT, TLD, HOLTEK, FAS-A inclinometer sensor, PID control.
Authors: Yung Chin Lin, Kuo Lan Su, Cheng Yun Chung
Abstract: The paper proposes an adaptive fusion algorithm using competitiveness sensors for fire detection module, and uses computer simulation results to select the optimal weight values for each optic-sensor. Then we design the fire detection module using the tuned weight values of optic-sensors. The competitiveness flame sensor type is ultra-violet sensor (R2868). The controller of the module is HOLTEK microchip, and acquires the detection signals from the optic-sensors through I/O pins, and transmits the detection signals of all sensors to the computer via wire series interface. The adaptive fusion algorithm can tunes weight values according to decision output of the fusion center. The fusion algorithms of the fusion center use Bayesian estimated method to decide the fire event to be true or not. We set the improved weight values in the module for each optic-sensor. From the simulation and experimental implementation results, it demonstrates that the proposed algorithms can compute the adequate weight values.
Authors: Kuo Lan Su, Bo Yi Li, Jian Da Fong
Abstract: We present the path planning techniques of the fire escaping system using multiple mobile robots for intelligent building. The controller of the mobile robot is MCS-51 microchip, and acquires the detection signal from flame sensor through I/O pins, and receives the command from the supervised compute via wireless RF interface. The mobile robot transmits ID code, detection signal, location and orientation of the mobile robots to the supervised computer via wireless RF interface. We proposed A* searching algorithm to program escaping motion paths to guard peoples moving to the safety area using mobile robots, and develop user interface on the supervised computer for the fire escaping system. In the experimental results, the supervised computer locates the positions of fire sources by mobile robots, and programs the escaping paths on the user interface, and transmits the motion command to the mobile robots. The mobile robot guards peoples leaving the fire sources.
Authors: Yung Chin Lin, Kuo Lan Su, Chih Hung Chang
Abstract: The article programs the shortest path searching problems of the mobile robot in the complexity unknown environment, and uses the mobile robot to present the movement scenario from the start point to the target point in a collision-free space. The complexity environment contains variety obstacles, such as road, tree, river, gravel, grass, highway and unknown obstacle. We set the relative dangerous grade for variety obstacles. The mobile robot searches the target point to locate the positions of unknown obstacles, and avoids these obstacles moving in the motion platform. We develop the user interface to help users filling out the positions of the mobile robot and the obstacles on the supervised computer, such the initial point of the mobile robot, the start point and the target point. The supervised computer programs the motion paths of the mobile robot according to A* searching algorithm, flood-fill algorithm and 2-op exchange algorithm The simulation results present the proposed algorithms that program the shortest motion paths from the initial point approach to the target point for the mobile robot. The supervised computer controls the mobile robot that follows the programmed motion path moving to the target point in the complexity environment via wireless radio frequency (RF) interface.
Authors: J. Hung Guo, Kuo Lan Su
Abstract: The article mainly researches path planning and task allocation problems of multiple mobile robots using A* searching algorithm and greedy algorithm, and solve the shortest path problems such that the robots can move from the start point to reach the multiple target points in a collision-free space, and uses 2-opt exchange heuristic algorithm to improve the shortest path. In this manner, the mobile moves to the final target point through the other points, and construct the motion path using A* searching algorithm and greedy algorithm. The supervised computer control the mobile robot feedback to the start point from the final target point through the other points, and programs a shortest path using 2-opt exchange heuristic algorithm. We develop the user interface to program the motion path of mobile robots via wireless RF interface. It can displays the motion path of the mobile robot on real-time. The simulated results presents that the proposed method can finds the shortest motion path for mobile robots moving to multiple target points from the start point in a collision-free space. Finally, we implement the experiment scenario on the grid platform using the module-based mobile robot.
Authors: J. Hung Guo, Yung Chin Lin, Kuo Lan Su, Bo Yi Li
Abstract: The article designs the multiple pattern formation controls of the multi-robot system according to two arms’ gesture of the player, and uses flood fill searching algorithm and A* searching algorithm to program the motion paths. The inertia module detects two arms’ gesture of the player. We use the inertia module to be embedded in the two arms, and use mobile robots to present the movement scenario of pattern formation exchange on the grid based motion platform. We have been developed some pattern formations applying in the war game, such as rectangle pattern formation, long snake pattern formation, L pattern formation, sword pattern formation, cone pattern formation and so on. We develop the user interface for variety pattern formation exchange according to the minimum displacement on the supervised computer. The mobile robot receives the command from the supervised compute, and transmits the status of environment to the supervised computer via wireless RF interface. Players can use variety arms’ gesture to control the multiple mobile robots to executed pattern formation exchange. In the experimental results, the supervised computer can decides the arm gesture using fusion algorithms. Mobile robots can receive the pattern formation command from the supervised computer, and change the original pattern formation to the assigned pattern formation on the motion platform, and avoid other mobile robots.
Showing 1 to 10 of 12 Paper Titles