• 제어로봇시스템학회:학술대회논문집
• /
• 1993.10a
• /
• pp.14-19
• /
• 1993

In this paper, we proposed an indirect learning and direct adaptive control schemes using neural networks, i.e., composite adaptive neural control, for a class of continuous nonlinear systems. With the indirect learning method, the neural network learns the nonlinear basis of the system inverse dynamics by a modified backpropagation learning rule. The basis spans the local vector space of inverse dynamics with the direct adaptation method when the indirect learning result is within a prescribed error tolerance, as such this method is closely related to the adaptive control methods. Also hash addressing technique, similar to the CMAC functional architecture, is introduced for partitioning network hidden nodes according to the system states, so global neuro control properties can be organized by the local ones. For uniform stability, the sliding mode control is introduced when the neural network has not sufficiently learned the system dynamics. With proper assumptions on the controlled system, global stability and tracking error convergence proof can be given. The performance of the proposed control scheme is demonstrated with the simulation results of a nonlinear system.

• Proceedings of the Korean Institute of Electrical and Electronic Material Engineers Conference
• /
• 2004.07b
• /
• pp.1234-1238
• /
• 2004

A Controlled neural networks are proposed in order to measure nonlinear environments in adaptive and in realtime. The structure of it is similar to recurrent neural networks: a delayed output as the input and a delayed error between tile output of plant and neural networks as a bias input. In addition, we compute the desired value of hidden layer by an optimal method instead of transfering desired values by backpropagation and each weights are updated by RLS(Recursive Least Square). Consequently, this neural networks are not sensitive to initial weights and a learning rate, and have a faster convergence rate than conventional neural networks. This new neural networks is Error Estimated Neural Networks. We can estimate nonlinear models in realtime by the proposed networks and control nonlinear models. To show the performance of this one, we have various experiments. And this controller call prove effectively to be control in the environments of various systems.

• Proceedings of the KIEE Conference
• /
• 2003.07e
• /
• pp.49-51
• /
• 2003

This paper presents a speed control system of induction motor using neural network. The speed control of induction motor was designed to NNC(Neural Network Controller) and NNE(Neural Network Estimator) used backpropagation, the NNE was constituted to be get an error value of output of an induction motor and conspire an input/output. NNC is controled to be made the error of reference speed and actual speed decrease, and in order to determine the weighting of NNC can be back propagated through the NNE, and it is adapted to the outside circumstances and system characters with learning ability.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.32B no.3
• /
• pp.503-511
• /
• 1995

A batch-mode algorithm is proposed to increase the speed of learning in the error backpropagation algorithm with variable learning rate and variable momentum parameters in classification problems. The objective function is normalized with respect to the number of patterns and output nodes. Also the gradient of the objective function is normalized in updating the connection weights to increase the effect of its backpropagated error. The learning rate and momentum parameters are determined from a function of the gradient norm and the number of weights. The learning rate depends on the square rott of the gradient norm while the momentum parameters depend on the gradient norm. In the two typical classification problems, simulation results demonstrate the effectiveness of the proposed algorithm.

• 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 Society for Precision Engineering
• /
• v.20 no.2
• /
• pp.146-153
• /
• 2003

Ultrasonic inspection of defects has been focused on the existence of defect in structural material and need has much time and expenses in inspecting all the coordinates (x, y) on material surface. Neural networks can have an application to coordinates (x, y) of defects by multi-point inspection method. Ultrasonic inspection modeling is optimized by neural networks Neural networks has trained training example of absolute and relative coordinate of defects, and defect pattern. This method can predict coordinates (x, y) of defects within engineering estimated mean error $\psi$.

• Journal of the Korean Institute of Telematics and Electronics B
• /
• v.31B no.8
• /
• pp.211-218
• /
• 1994

In this paper we develope a hybrid learning rule to improve the robustness of multi-layer Perceptions. In most neural networks the activation of a neuron is deternined by a nonlinear transformation of the weighted sum of inputs to the neurons. Investigating the behaviour of activations of hidden layer neurons a new learning algorithm is developed for improved robustness for multi-layer Perceptrons. Unlike other methods which reduce the network complexity by putting restrictions on synaptic weights our method based on error-backpropagation increases the complexity of the underlying proplem by imposing it saturation requirement on hidden layer neurons. We also found that the additional gradient-descent term for the requirement corresponds to the Hebbian rule and our algorithm incorporates the Hebbian learning rule into the error back-propagation rule. Computer simulation demonstrates fast learning convergence as well as improved robustness for classification and hetero-association of patterns.

• Proceedings of the KIEE Conference
• /
• 2000.07d
• /
• pp.2380-2382
• /
• 2000

Fuzzy system which are based on membership functions and rules, can control nonlinear, uncertian, complex system well. However, Fuzzy controller has problems: It is difficult to design a stable for amateur. To update the then-part membership functions of the fuzzy controller can be designed using the error back-propagation algorithm to be minimized error. Then we could be optimized the system choosing a good performance index. The proposed fuzzy controller based on neural network is applied to control an inverted pendulum for demonstration of the robustness of proposed methodology.

• Proceedings of the KIEE Conference
• /
• 1996.11a
• /
• pp.407-410
• /
• 1996

In this Paper, we propose the feedback method having neural network to control the chaotic signals to periodic signals. This controller has very simple structure, it is immune to small parameter variations, the precise access to system parameters is not required and it is possible to follow ones of its inherent periodic orbits or the desired orbits without error, The controller consist of linear feedback gain and neural network. The learning of neural network is achieved by error-backpropagation algorithm. To prove and analyze the proposed method, we construct a software tool using c-language.

• Tunnel and Underground Space
• /
• v.19 no.3
• /
• pp.203-212
• /
• 2009

The prediction of near field behavior around an underground high-level radioactive waste repository is important for the repository design as well as the safety assessment. In this study, a sensitivity analysis for seven parameters consisted of design parameters and material properties was carried out using a three-dimensional finite difference code. From the sensitivity analysis, it was found that the effects of borehole spacing, tunnel spacing, cooling time and rock thermal conductivity were more significant than the other parameters. For getting a statistical distribution of buffer and rock temperatures around the repository, an artificial neural network, backpropagation, was applied. The reliability of the trained neural network was tested with the cases with randomly chosen input parameters. When the parameter variation is within ${\pm}10%$, the prediction from the network was found to be reliable with about a 1% error. It was possible to calculate the temperature distribution for many cases quickly with the trained neural network. The buffer and rock temperatures showed a normal distribution with means of $98^{\circ}C$ and $83.9^{\circ}C$ standard deviations of $3.82^{\circ}C$ and $3.67^{\circ}C$, respectively. Using the neural network, it was also possible to estimate the required change in design parameters for reducing the buffer and rock temperatures for $1^{\circ}C$.