• The Journal of Korea Robotics Society
• /
• v.15 no.4
• /
• pp.398-403
• /
• 2020

Robot's throwing control is difficult to accurately calculate because of air resistance and rotational inertia, etc. This complexity can be solved by using machine learning. Reinforcement learning using reward function puts limit on adapting to new environment for robots. Therefore, this paper applied deep reinforcement learning using neural network without reward function. Throwing is evaluated as a success or failure. AI network learns by taking the target position and control policy as input and yielding the evaluation as output. Then, the task is carried out by predicting the success probability according to the target location and control policy and searching the policy with the highest probability. Repeating this task can result in performance improvements as data accumulates. And this model can even predict tasks that were not previously attempted which means it is an universally applicable learning model for any new environment. According to the data results from 520 experiments, this learning model guarantees 75% success rate.

• Journal of information and communication convergence engineering
• /
• v.5 no.4
• /
• pp.362-368
• /
• 2007

A single-phase uninterruptible power supply system equipped with a superconducting magnet energy storage unit is proposed to achieve a simple circuit configuration and higher system reliability. It reduces a number of switching devices by applying a common-arm scheme. Removing some switches or substituting passive elements for active switches can increase the sophistication and reduces degree of freedom in control strategy. However, high-performance DSP controller can execute the complicated control task with no additional cost. Operational principles to normal, stored-energy, and bypass mode are discussed in detail. The validity of the proposed system is verified by experimental results.

• Journal of Power Electronics
• /
• v.17 no.1
• /
• pp.127-135
• /
• 2017

Recently the researches on modular multi-level converter (MMC) are being highlighted because high quality and efficient power transmission have become key issues in high voltage direct current (HVDC) systems. This paper proposes an improved pre-charging method for the sub-module (SM) capacitor of MMC-based HVDC systems, which operates in the nearest level control (NLC) modulation and does not need additional circuits or pulse width modulation (PWM) techniques. The feasibility of the proposed method was verified through computer simulations for a scaled 3-phase 10kVA MMC with 12 SMs per each arm. Hardware experiments with a scaled prototype have also been performed in the lab to confirm the simulation results.

• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.119.2-119
• /
• 2002

In this paper a passive compliant joint (PCJ) for human-friendly service robots is presented. The PCJ is composed of a magneto-rheological damper and a spring for elasticity. In the magneto-rheological damper, a braking force proportional to the electric current is generated. It works as a viscous damper by controlling the electric current according to the angular velocity of spring displacement. The compliance property and position control of the PCJ are verified through experiments and simulations to a single DOF simple arm with the PCJ.

• Journal of Institute of Control, Robotics and Systems
• /
• v.15 no.7
• /
• pp.728-733
• /
• 2009

This paper presents the compensation of mechanical deflection error in SCARA robot. End of robot gripper is deflected by weight of arm and pay-load. If end of robot gripper is deflected constantly regardless of robot configuration, it is not necessary to consider above mechanical deflection error. However, deflection in end of gripper varies because that moment of each axis varies when robot moves, it affects the relative accuracy. I propose the compensation method of deflection error using neural network. FEM analysis to obtain the deflection of gripper end was carried out on various joint angle, the results is used in neural network teaming. The result by simulation showed that maximum relative accuracy reduced maximum 9.48% on a given working area.

• Journal of Mechanical Science and Technology
• /
• v.14 no.2
• /
• pp.188-196
• /
• 2000

In this paper, our aim is to develop a model for two cooperating flexible manipulators handling a rigid object by using lumped parameters. This model is in turn analyzed on MATLAB. In order to validate the model, a precise simulation model is developed using $ADAMS^{TM}$ (Automatic Dynamic Analysis of Mechanical System). Moreover, to clarify the discussion, the motions of a dual-arm experimental flexible manipulator are considered. Using the developed model, we control a robotic system with a symmetric hybrid position/force control scheme. Finally, experiments and simulations are performed, and a comparison of simulation results with experimental results is given to a rerify the validity of our model.

• Journal of Institute of Control, Robotics and Systems
• /
• v.13 no.7
• /
• pp.671-676
• /
• 2007

In this paper, a novel neuro-fuzzy controller is presented. The generic fuzzy controller is compensated by a neural network controller so that an overall control structure forms a neuro-fuzzy controller. The proposed neuro-fuzzy controller solves the difficulty of selecting optimal fuzzy rules by providing the similar effect of modifying fuzzy rules simply by changing crisp input values. The performance of the proposed controller is tested by controlling humanoid robot arms. The humanoid robot arm is analyzed and implemented. Experimental studies have shown that the performance of the proposed controller is better than that of a PID controller and of a generic fuzzy PD controller.

• Proceedings of the KIEE Conference
• /
• 1997.11a
• /
• pp.80-82
• /
• 1997

A robust position control system for a BLDC motor using new sliding mode control strategy is presented. Using the new variable structure system, reaching phase problem is eliminated and transient response is largely improved by design of nonlinear sliding surface. The design of the sliding mode position controller is robust in motor parameter, load variations and disturbance. Experiment results show that the proposed approach can achieve accurate position motor tracking in face of large parameter variations and external disturbances, such as a robot arm, etc.

• The Transactions of the Korean Institute of Electrical Engineers P
• /
• v.63 no.2
• /
• pp.69-73
• /
• 2014

In this paper, remote control and surveillance monitoring system of HID street-lamp using low-cost type of Zigbee communication module was developed. A plurality of capacitors with different capacity are attached to magnetic ballast for HID lamps. Changes in the amount of illumination can be adjusted by the changes in capacity so that the electric energy can be reduced. In Zigbee communication module, a ATmega128L microprocessor is used and in main CPU, ARM9 Core is used to construct a low-cost smart system. In surveillance monitoring systems, the system was configured to remote monitoring of presence of error in the lamp abnormalities, the status of leakage current and abnormalities of ballasts and capacitors and the validity of the proposed empirical experiment was confirmed.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.223-223
• /
• 2000

Embedded system area has brought an innovation and has been spread rapidly by the growth of the Internet, wireless telephony and multimedia recently. Many embedded systems are required to be real-time systems in that it needs multi-tasking and priority based scheduling. This paper introduces a real-time system that was developed with web server ability for control and monitoring system employing a real-time operating system. It discusses the design model, structure, and applications of web server system. We used SNDS100 board which has a 32-bit RISC microcontroller of ARM7TDMI core as a hardware platform. MicroC/OS kernel was used as Real-time operating system that supports a preemptive and multitasking functions. We developed a hierarhchical control and monitoring system that not only reduced system and management costs, but also enhanced reusability and portability.