• Proceedings of the KIEE Conference
• /
• 1998.07g
• /
• pp.2363-2366
• /
• 1998

The intent of this paper is to describe a neural network structure called dynamic neural processor(DNP), and examine how it can be used in developing a learning scheme for computing robot inverse kinematic transformations. The architecture and learning algorithm of the proposed dynamic neural network structure, the DNP, are described. Computer simulations are provided to demonstrate the effectiveness of the proposed learning using the DNP.

• 제어로봇시스템학회:학술대회논문집
• /
• 1990.10b
• /
• pp.1173-1177
• /
• 1990

Cerebellar Model Linear Associator Net(CMLAN), a kind of neuro-net based adaptive control function generator, was applied to the problem of direct inverse calibration of three and six d.o.f. POMA 560 robot. Since CMLAN autonomously maps and generalizes a desired system function via learning on the sampled input/output pair nodes, CMLAN allows no knowledge in system modeling and other error sources. The CMLAN based direct inverse calibration avoids the complex procedure of identifying various system parameters such as geometric(kinematic) or nongeometric(dynamic) ones and generates the corresponding desired compensated joint commands directly to each joint for given target commands in the world coordinate. The generated net outputs automatically handles the effect of unknown system parameters and dynamic error sources. On-line sequential learning on the prespecified sampled nodes requires only the measurement of the corresponding tool tip locations for three d.o.f. manipulator but location and orientation for six d.o.f. manipulator. The proposed calibration procedure can be applied to any robot.

• 제어로봇시스템학회:학술대회논문집
• /
• 2000.10a
• /
• pp.431-431
• /
• 2000

The modeling of 5-bar linkage robot manipulator dynamics by means of a mathematical and neural architecture is presented. Such a model is applicable to the design of a feedforward controller or adjustment of controller parameters. The inverse model consists of two parts: a mathematical part and a compensation part. In the mathematical part, the subsystems of a 5-bar linkage robot manipulator are constructed by applying Kawato's Feedback-Error-Learning method, and trained by given training data. In the compensation part, MLP backpropagation algorithm is used to compensate the unmodeled dynamics. The forward model is realized from the inverse model using the inverse of inertia matrix and the compensation torque is decoupled in the input torque of the forward model. This scheme can use tile mathematical knowledge of the robot manipulator and analogize the robot characteristics. It is shown that the model is reasonable to be used for design and initial gain tuning of a controller.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.17 no.6
• /
• pp.807-812
• /
• 2007

Recently, Extreme learning machine(ELM), a novel learning algorithm which is much faster than conventional gradient-based learning algorithm, was proposed for single-hidden-layer feedforward neural networks. The initial input weights and hidden biases of ELM are usually randomly chosen, and the output weights are analytically determined by using Moore-Penrose(MP) generalized inverse. But it has the difficulties to choose initial input weights and hidden biases. In this paper, an advanced method using the bacterial foraging algorithm to adjust the input weights and hidden biases is proposed. Experiment at results show that this method can achieve better performance for problems having higher dimension than others.

• Smart Structures and Systems
• /
• v.24 no.2
• /
• pp.233-242
• /
• 2019

In this paper, an inverse approach based on uniform load surface (ULS) is presented for structural damage localization and quantification. The ULS is excellent approximation for deformed configuration of a structure under distributed unit force applied on all degrees of freedom. The ULS make use of natural frequencies and mode shapes of structure and in mathematical point of view is a weighted average of mode shapes. An objective function presented to damage detection is discrepancy between the ULS of monitored structure and numerical model of structure. Solving this objective function to find minimum value yields damage's parameters detection. The teaching-learning based optimization algorithm has been employed to solve inverse problem. The efficiency of present damage detection method is demonstrated through three numerical examples. By comparison between proposed objective function and another objective function which make use of natural frequencies and mode shapes, it is revealed present objective function have faster convergence and is more sensitive to damage. The method has good robustness against measurement noise and could detect damage by using the first few mode shapes. The results indicate that the proposed method is reliable technique to damage detection in structures.

• 제어로봇시스템학회:학술대회논문집
• /
• 2005.06a
• /
• pp.1911-1916
• /
• 2005

In this study, the basic motor control system had been investigated. The Discrete-Time Feedback Error Learning (DTFEL) method is used to control this system. This method is anologous to the original continuous-time version Feedback Error Learning(FEL) control which is proposed as a control model of cerebellum in the field of computational neuroscience. The DTFEL controller consists of two main parts, a feedforward controller part and a feedback controller part. Each part will deals with different control problems. The feedback controller deals with robustness and stability, while the feedforward controller deals with response speed. The feedforward controller, used to solve the tracking control problem, is adaptable. To make such the tracking perfect, the adaptive law is designed so that the feedforward controller becomes an inverse system of the controlled plant. The novelty of FEL method lies in its use of feedback error as a teaching signal for learning the inverse model. The PD control theory is selected to be applied in the feedback part to guarantee the stability and solve the robust stabilization problems. The simulation of each individual part and the integrated one are taken to clarify the study.

• Electronics and Telecommunications Trends
• /
• v.34 no.6
• /
• pp.100-107
• /
• 2019

Recently, reinforcement learning (RL) has expanded from the research phase of the virtual simulation environment to a wide range of applications, such as autonomous driving, natural language processing, recommendation systems, and disease diagnosis. However, RL is less likely to be used in these complex real-world environments. In contrast, inverse reinforcement learning (IRL) can obtain optimal policies in various situations; furthermore, it can use expert demonstration data to achieve its target task. In particular, IRL is expected to be a key technology for artificial general intelligence research that can successfully perform human intellectual tasks. In this report, we briefly summarize various IRL techniques and research directions.

• KIISE Transactions on Computing Practices
• /
• v.23 no.10
• /
• pp.599-604
• /
• 2017

A useful application of smart assistants is to predict and suggest users' daily behaviors the way real assistants do. Conventional methods to predict behavior have mainly used explicit schedule information logged by a user or extracted from e-mail or SNS data. However, gathering explicit information for smart assistants has limitations, and much of a user's routine behavior is not logged in the first place. In this paper, we suggest a novel approach that combines explicit schedule information with patterns of routine behavior. We propose using inference based on a Markov decision process and learning with a reward function based on inverse reinforcement learning. The results of our experiment shows that the proposed method outperforms comparable models on a life-log dataset collected over six weeks.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.4C
• /
• pp.403-413
• /
• 2009

We compare three predistortion methods to prevent signal distortion and spectral re-growth due to the high PAPR (peak-to-average ratio) of OFDM signal and the non-linearity of high-power amplifiers. The three predistortion methods are pth order inverse, indirect learning architecture and look up table. The pth order inverse and indirect learning architecture methods requires less memory and has a fast convergence because these methods use a polynomial model that has a small number of coefficients. Nevertheless the convergence is fast due to the small number of coefficients and the simple computation that excludes manipulation of complex numbers by separate compensation for the magnitude and phase. The look up table method is easy to implement due to simple computation but has the disadvantage that large memory is required. Computer simulation result reveals that indirect learning architecture shows the best performance though the gain is less than 1 dB at $BER\;=\;10^{-4}$ for 64-QAM. The three predistorters are adaptive to the amplifier aging and environmental changes, and can be selected to the requirements for implementation.

• 제어로봇시스템학회:학술대회논문집
• /
• 2004.08a
• /
• pp.1711-1716
• /
• 2004

This paper presents identification and control of a 10-m antenna via accelerometers and angle encoder data. Artificial Neural Networks can be used effectively for the identification and control of nonlinear dynamical system such as a large flexible antenna. Some identification results are shown and compared with the results of conventional prediction error method. And we use a neural network inverse model for control the large flexible antenna. In the neural network inverse model, a neural network is trained, using supervised learning, to develop an inverse model of the antenna. The network input is the process output, and the network output is the corresponding process input. The control results show the validation of the ANN approach for identification and control of the 10-m flexible antenna.