• Journal of Korean Association for Spatial Structures
• /
• v.21 no.4
• /
• pp.49-56
• /
• 2021

This study aims to develop a form-finding algorithm for a single-layered pneumatic membrane. The initial shape of this pneumatic membrane, which is an air-supported type pneumatic membrane, is to find a state in which a given initial tension and internal pneumatic pressure are in equilibrium. The algorithm developed to satisfy these conditions is that a nonlinear optimization problem based on the force method considering the deformed shape is formulated, and, it's able to find the shape by iteratively repeating the process of obtaining a solution of the governing equations. An computational technique based on the Gauss-Newton method was used as a method for obtaining solutions of nonlinear equations. In order to verify the validity of the proposed form-finding algorithm, a single-curvature pneumatic membrane example and a double-curvature air pneumatic membrane example were adopted, respectively. In the results of these examples, it was possible to well observe the step-by-step convergence process of the shape of the pneumatic membrane, and it was also possible to confirm the change in shape according to the air pressure. In addition, the calculation results of the shape and internal force after deformation due to initial tension, air pressure, and self-weight were obtained.

• Journal of the Korean Society for Precision Engineering
• /
• v.12 no.6
• /
• pp.120-127
• /
• 1995

A nonlinear empirical state-space model of the Artificial Neural Network(ANN) has been developed. The nonlinear model structure incorporates characteristic, so as to enable identification of the transient response, as well as the steady-state response of a dynamic system. A hybrid feedfoward/feedback neural network, namely a Local Time Delayed Recurrent Multi-layer Perception(RMLP), is the model structure developed in this paper. RMLP is used to identify nonlinear dynamic system in an input/output sense. The feedfoward protion of the network architecture provides with the well-known curve fitting factor, while local recurrent and cross-talk connections provides the dynamics of the system. A dynamic learning algorithm is used to train the proposed network in a supervised manner. The derived dynamic learning algorithm exhibit a computationally desirable characteristic; both network sweep involved in the algorithm are performed forward, enhancing its parallel implementation. RMLP state-space and its associate learning algorithm is demonstrated through a simple examples. The simulation results are very encouraging.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.13 no.11
• /
• pp.5379-5388
• /
• 2012

In fuzzy modeling for nonlinear process, the fuzzy rules are typically formed by selection of the input variables, the number of space division and membership functions. The Generation of fuzzy rules for nonlinear processes have the problem that the number of fuzzy rules exponentially increases. To solve this problem, complex nonlinear process can be modeled by generating the fuzzy rules by means of fuzzy division of input space. Therefore, in this paper, rules of non-fuzzy inference systems are generated by partitioning the input space in the scatter form using HCM clustering algorithm. The premise parameters of the rules are determined by membership matrix by means of HCM clustering algorithm. The consequence part of the rules is represented in the form of polynomial functions and the consequence parameters of each rule are identified by the standard least-squares method. And lastly, we evaluate the performance and the nonlinear characteristics using the data widely used in nonlinear process. Through this experiment, we showed that high-dimensional nonlinear systems can be modeled by a very small number of rules.

• Smart Structures and Systems
• /
• v.16 no.6
• /
• pp.1133-1145
• /
• 2015

A new neural network (NN) predictive controller (NNPC) algorithm has been developed and tested in the computer simulation of active control of a nonlinear structure. In the present method an NN is used as a predictor. This NN has been trained to predict the future response of the structure to determine the control forces. These control forces are calculated by minimizing the difference between the predicted and desired responses via a numerical minimization algorithm. Since the NNPC is very time consuming and not suitable for real-time control, it is then used to train an NN controller. To consider the effectiveness of the controller on probability of damage, fragility curves are generated. The approach is validated by using simulated response of a 3 story nonlinear benchmark building excited by several historical earthquake records. The simulation results are then compared with a linear quadratic Gaussian (LQG) active controller. The results indicate that the proposed algorithm is completely effective in relative displacement reduction.

• Journal of the Korean Society of Fisheries and Ocean Technology
• /
• v.32 no.2
• /
• pp.164-176
• /
• 1996

The main intention of this paper is to develop and compare the algorithm based on finite element procedures for nonlinear transient dynamic analysis which has combined effects of material and geometric nonlinearities. Incremental equilibrium equations based on the principle of virtual work are derived by the finite element approach. For the elasto - plastic large deformation analysis of shells and the determination of the displacement-time configuration under time-varying loads, the explicit, implicit and combined explicit-implicit time integration algorithm is adopted. In the time structure is selected and the results are compared with each others. Isoparametric 8-noded quadrilateral curved elements are used for shell structure in the analysis and for geometrically nonlinear elastic behaviour, a total Lagrangian coordinate system was adopted. On the other hands, material nonlinearity is based on elasto-plastic models with Von-Mises yield criteria. Thus, the combined explicit-implicit time integration algorithm is benefit in general case of shell structure, which is the result of this paper.

• 제어로봇시스템학회:학술대회논문집
• /
• 2001.10a
• /
• pp.71.2-71
• /
• 2001

This paper deals with nonlinear system modelling using neural network and genetic algorithm. Application of neural network to control and identification is actively studied because of their approximating ability of nonlinear function. It is important to design the neural network with optimal structure for minimum error and fast response time. Genetic algorithm is getting more popular nowadays because of their simplicity and robustness. In this paper, We optimize neural network structure using genetic algorithm. The genetic algorithm uses binary coding for neural network structure and search for optimal neural network structure of minimum error and response time. Through extensive simulation, Optimal neural network structure is shown to be effective for ...

• Journal of the Computational Structural Engineering Institute of Korea
• /
• v.20 no.6
• /
• pp.731-742
• /
• 2007

This paper presents a new nonlinear analysis algorithm, that is, adaptive Newton-Raphson iteration method, The presented algorithm is based on the existing Newton-Raphson method, and the concept of it can be summarized as calculating the equivalent load for stiffness(ELS) and adapting this to the initial global stiffness matrix which has already been calculated and saved in initial analysis and finally calculating the correction displacements for the nonlinear analysis, The key characteristics of the proposed algorithm is that it calculates the inverse matrix of the global stiffness matrix only once irresponsive of the number of load steps. The efficiency of the proposed algorithm depends on the ratio of the active Dofs - the Dofs which are directly connected to the members of which the element stiffness are changed - to the total Dofs, and based on this ratio by using the proposed algorithm as a complementary method to the existing algorithm the efficiency of the nonlinear analysis can be improved dramatically.

• Journal of Advanced Navigation Technology
• /
• v.4 no.1
• /
• pp.45-49
• /
• 2000

A GPS measurement solution, in general, is obtained as a least squares solution since the measurement includes errors such as clock errors, ionospheric and tropospheric delays, multipath effect etc. Because of the nonlinearity of the measurement equation, we utilize the nonlinear Newton algorithm to obtain a least squares solution, or mostly, use its linearized algorithm which is more convenient and effective. In this study we developed a fixed point algorithm and proved its availability to replace the nonlinear Newton algorithm and the linearized algorithm. A nonlinear Newton algorithm and a linearized algorithm have the advantage of fast convergence, while their initial values have to be near the unknown solution. On the contrary, the fixed point algorithm provides more reliable but slower convergence even if the initial values are quite far from the solution. Therefore, two types of algorithms may be combined to achieve better performance.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 2003.09b
• /
• pp.95-98
• /
• 2003

In this paper, an adaptive fuzzy neural control of unknown nonlinear systems based on the rapid learning algorithm is proposed for optimal parameterization. We combine the advantages of fuzzy control and neural network techniques to develop an adaptive fuzzy control system for updating nonlinear parameters of controller. The Fuzzy Neural Network(FNN), which is constructed by an equivalent four-layer connectionist network, is able to learn to control a process by updating the membership functions. The free parameters of the AFN controller are adjusted on-line according to the control law and adaptive law for the purpose of controlling the plant track a given trajectory and it's initial values are off-line preprocessing, In order to improve the convergence of the learning process, we propose a rapid learning algorithm which combines the error back-propagation algorithm with Aitken's $\delta$$\^$2/ algorithm. The heart of this approach ls to reduce the computational burden during the FNN learning process and to improve convergence speed. The simulation results for nonlinear plant demonstrate the control effectiveness of the proposed system for optimal parameterization.

• The Journal of Korea Institute of Information, Electronics, and Communication Technology
• /
• v.5 no.4
• /
• pp.224-231
• /
• 2012

In this paper, we introduce a fuzzy inference systems based on fuzzy c-means clustering algorithm for fuzzy modeling of nonlinear process. Typically, the generation of fuzzy rules for nonlinear processes have the problem that the number of fuzzy rules exponentially increases. To solve this problem, the fuzzy rules of fuzzy model are generated by partitioning the input space in the scatter form using FCM clustering algorithm. The premise parameters of the fuzzy rules are determined by membership matrix by means of FCM clustering algorithm. The consequence part of the rules is expressed in the form of polynomial functions and the coefficient parameters of each rule are determined by the standard least-squares method. And lastly, we evaluate the performance and the nonlinear characteristics using the data widely used in nonlinear process.