• The Journal of Korea Robotics Society
• /
• v.17 no.1
• /
• pp.16-24
• /
• 2022

This paper presents a method that can reduce the computational cost of the hierarchical quadratic programming (HQP)-based robot controller. Hierarchical controllers can effectively manage articulated robots with many degrees of freedom (DoFs) to perform multiple tasks. The HQP-based controller is one of the generic hierarchical controllers that can provide a control solution guaranteeing strict task priority while handling numerous equality and inequality constraints. However, according to a large amount of computation, it can be a burden to use it for real-time control. Therefore, for practical use of the HQP, we propose a method to reduce the computational cost by decreasing the size of the decision variable. The computation time and control performance of the proposed method are evaluated by real robot experiments with a 15 DoFs dual-arm manipulator.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2022.05a
• /
• pp.45-47
• /
• 2022

Mobile edge computing (MEC), which enables mobile terminals to offload computational tasks to a server located at the user's edge, is considered an effective way to reduce the heavy computational burden and achieve efficient computational offloading. In this paper, we study a multi-user MEC system in which multiple user devices (UEs) can offload computation to the MEC server via a wireless channel. To solve the resource allocation and task offloading problem, we take the total cost of latency and energy consumption of all UEs as our optimization objective. To minimize the total cost of the considered MEC system, we propose an DRL-based method to solve the resource allocation problem in wireless MEC. Specifically, we propose a Asynchronous Advantage Actor-Critic (A3C)-based scheme. Asynchronous Advantage Actor-Critic (A3C) is applied to this framework and compared with DQN, and Double Q-Learning simulation results show that this scheme significantly reduces the total cost compared to other resource allocation schemes

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.25 no.12
• /
• pp.1868-1873
• /
• 2021

This paper proposes the Hough transform(HT) using an extended lookup table(LUT) to reduce the computational burden of the HT that is a typical straight line detection algorithm, and analyzes its performance. The conventional HT also uses a LUT to the calculation of the parameter 𝜌 of all straight lines passing through an edge pixel of interest(ePel) in order to reduce the computational burden. However, the proposed HT adopts an extended LUT that can be applied to straight lines across the ePel as well as its peripheral edge pixels to induce more computational reduction. This paper proves the validity of the proposed algorithm mathematically and also verifies it through simulation. The simulation results show that the proposed HT reduces the multiplication computation from 49.6% up to 16.1%, depending on the image and the applied extended LUT, compared to the conventional HT.

• The Journal of the Acoustical Society of Korea
• /
• v.15 no.2E
• /
• pp.50-59
• /
• 1996

In this paper, an adaptive precompensator, which can reduce the distortion of a Wiener system effectively, is proposed. The previous techniques for adaptive precompensation, based on the Volterra series modeling to compensate the distortion of a nonlinear system, are not suitable for real-time implementation due to high computational burden and slow convergence burden and slow convergence rate. This paper presents an adaptive precompensation technique for the class of nonlinear subsystem, referred to as Wiener system. An adaptive algorithm for adjusting the parameters of a precompensator, structured by a hammerstein model, is derived using the stochastic gradient method. Also, an adaptive precompensatin technique which can effectively reduce nonlinear distortion in μ-law type of saturation characteristics is proposed. The validity of the proposed algorithm is confirmed through simulation by applying it to known Wiener systmes and a typical loudspeaker model.

• Proceedings of the KIEE Conference
• /
• 1991.07a
• /
• pp.771-776
• /
• 1991

In this paper. an indirect adaptive controller for manipulator which is composed of two controller structure is considered. One is feedforward controller in which the dynamics equation solved and the other is feedback controller in which the output error compensated. This controller has a good performance, but the computation burden of the feed forward controller keep from real time control. At this point, we proposed the two time adaptive controller where the sampling time of the feedforward controller is quite longer than that of the feedback controller. By the computer simulation, this proposed two time adaptive controller shows good performance in the view of accuracy in spite of decreasing computational burden.

• Journal of the Korean Society for Precision Engineering
• /
• v.12 no.6
• /
• pp.128-136
• /
• 1995

The problem addressed in this paper is that of the on-line computational burden of time-optimal control laws for quick, strongly nonlinear systems like revolute robots. It will be demonstrated that a large amount of off-line computation can be substituted for most of the on-line burden in cases of time optimization with constrained inputs if differential point-to- point specifications can be relaxed to cell-to-cell transitions. These cells result from a coarse discretization of likely swaths of state space into a set of nonuniform, contiguous volumes of relatively simple shapes. The cell boundaries approximate stream surfaces of the phase fluid and surfaces of equal transit times. Once the cells have been designed, the bang- bang schedules for the inputs are determined for all likely starting cells and terminating cells. The scheduling process is completed by treating all cells into which the trajectories might unex- pectedly stray as additional starting cells. Then an efficient-to-compute control law can be based on the resulting table of optimal strategies.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.3
• /
• pp.570-575
• /
• 2019

Active muffler is implemented by applying active noise control technique to reduce exhaust noise of automobile muffler. Conventional Filtered_x LMS algorithm has a problem that the degree of control filter becomes very large and convergence deteriorates when acoustic feedback is present. The recursive LMS algorithm can compensate for this problem because it can be easily diverted in the adaptive filter adaptation process. In this paper, the structure of the primary path and the secondary path transfer function is designed as the IIR filter to improve the convergence performance and the computational burden, and the stabilization filter algorithm is applied to secure stability which is a disadvantage of the IIR filter structure. The stabilization filter algorithm plays a role of pulling the pole into the unit circle to prevent the pole of the transfer function corresponding to the acoustic feedback from diverging during the adaptation process. In this way, the computational burden of the active muffler system and the convergence performance can be improved. In order to show the usefulness of the proposed system, we compared the performance of the proposed Filtered_x LMS algorithm with the performance of the proposed system for the exhaust sound of a diesel engine, which is a variable environment. Compared to conventional algorithm, proposed algorithm's computational burden is less than half, and convergence performances are more than 4 times.

• Journal of Satellite, Information and Communications
• /
• v.10 no.1
• /
• pp.22-28
• /
• 2015

In this paper, we propose a new computationally efficient joint iterative multi-input multi-output (MIMO) detection scheme using a soft interference cancellation and minimum mean squared-error (SIC-MMSE) method. The critical computational burden of the SIC-MMSE scheme lies in the multiple inverse operations of the complex matrices. We find a new way which requires only a single matrix inversion by utilizing the Taylor series expansion of the matrix, and thus the computational complexity can be reduced. The computational complexity reduction increases as the number of antennas is increased. The simulation results show that our method produces almost the same performances as the conventional SIC-MMSE with reduced computational complexity.

• Wind and Structures
• /
• v.36 no.6
• /
• pp.423-434
• /
• 2023

Aerodynamic force coefficients are generally obtained from traditional wind tunnel tests or computational fluid dynamics (CFD). Unfortunately, the techniques mentioned above can sometimes be cumbersome because of the cost involved, such as the computational cost and the use of heavy equipment, to name only two examples. This study proposed to build a deep neural network model to predict the aerodynamic force coefficients based on data collected from CFD simulations to overcome these drawbacks. Therefore, a series of CFD simulations were conducted using different geometric parameters to obtain the aerodynamic force coefficients, validated with wind tunnel tests. The results obtained from CFD simulations were used to create a dataset to train a multilayer perceptron artificial neural network (ANN) model. The models were obtained using three optimization algorithms: scaled conjugate gradient (SCG), Bayesian regularization (BR), and Levenberg-Marquardt algorithms (LM). Furthermore, the performance of each neural network was verified using two performance metrics, including the mean square error and the R-squared coefficient of determination. Finally, the ANN model proved to be highly accurate in predicting the force coefficients of similar bridge sections, thus circumventing the computational burden associated with CFD simulation and the cost of traditional wind tunnel tests.

• 한국연소학회:학술대회논문집
• /
• 1997.06a
• /
• pp.113-122
• /
• 1997

The conditional moment closure formulation considering the molecular and turbulent diffusion is derived. A simplified solution procedure is proposed to reduce the computational burden due to the increased dimensionality of the conditionally averaged variables. A conditionally averaged variable is expressed as a linear weighted average of the two extremes, 'no reaction' and 'equilibrium' states. The modified elliptic-type conditional moment closure formulation is implemented to simulate a two dimensional nonpremixed mixing layer reacting flow. Results show good agreement for the conditional averages of the species concentration in Bilger et al.