• 제어로봇시스템학회:학술대회논문집
• /
• 1996.10b
• /
• pp.1400-1403
• /
• 1996

In this study, we applied the model predictive control(MPC) to Multi-loop control structure. Since MPC has many advantage for MIMO process and constraints handling, it induces the better performance to apply MPC to multi-loop control. And we suggest the advanced method to reduce the calculation load using the wavelet transform. It shows the possibility to substitute the existing PID control based structure with MPC.

• Proceedings of the KIEE Conference
• /
• 2001.07d
• /
• pp.2669-2671
• /
• 2001

In this paper, we propose design a Multi-Fuzzy Inference model structure. In order to determine structure of the proposed Multi-Fuzzy Inference model, HCM clustering method is used. The parameters of membership function of the Multi-Fuzzy are identified by genetic algorithms. A aggregate performance index with a weighting factor is used to achieve a sound balance between approximation and generalization abilities of the model. We use simplified inference and linear inference as inference method of the proposed Multi-Fuzzy model and the standard least square method for estimating consequence parameters of the Multi-Fuzzy. Finally, we use some of numerical data to evaluate the proposed Multi-Fuzzy model and discuss about the usefulness.

• Speech Sciences
• /
• v.13 no.2
• /
• pp.75-84
• /
• 2006

Various methods have been proposed to overcome the problem of speech recognition in the noisy conditions. Among them, the model compensation methods like the parallel model combination (PMC) and Jacobian adaptation (JA) have been found to perform efficiently. The JA is quite effective when we have hidden Markov models (HMMs) already trained in a similar condition as the target environment. In a previous work, we have proposed an improved method for the JA to make it more robust against the changing environments in recognition. In this paper, we further improved its performance by compensating the delta-mean vectors and covariance matrices of the HMM and investigated its feasibility in the multi-model structure for the noisy speech recognition. From the experimental results, we could find that the proposed improved the robustness of the JA and the multi-model approach could be a viable solution in the noisy speech recognition.

• Structural Engineering and Mechanics
• /
• v.43 no.3
• /
• pp.371-394
• /
• 2012

Accurate finite element (FE) models are needed in many applications of Civil Engineering such as health monitoring, damage detection, structural control, structural evaluation and assessment. Model accuracy depends on both the model structure (the form of the equations) and the model parameters (the coefficients of the equations), and can be generally improved through that process of experimental reconciliation known as model updating. However, modelling errors, including (i) errors in the model structure and (ii) errors in parameters excluded from adjustment, may bias the solution, leading to an updated model which replicates measurements but lacks physical meaning. In this paper, an application of ambient-vibration-based model updating to a large-scale benchmark prototype of a building structure is reported in which both types of error are met. The error in the model structure, originating from unmodelled secondary structural elements unexpectedly working as resonant appendages, is faced through a reduction of the experimental modal model. The error in the model parameters, due to the inevitable constraints imposed on parameters to avoid ill-conditioning and under-determinacy, is faced through a multi-model parameterization approach consisting in the generation and solution of a multitude of models, each characterized by a different set of updating parameters. Results show that modelling errors may significantly impair updating even in the case of seemingly simple systems and that multi-model reasoning, supported by physical insight, may effectively improve the accuracy and robustness of calibration.

• 제어로봇시스템학회:학술대회논문집
• /
• 1986.10a
• /
• pp.321-325
• /
• 1986

A new control scheme for the state space model of the robot system using the theory of optimal multi-variable structure is presented in this paper. It is proposed to optimize multi-dimensional variable structure systems for obtaining the required stabilizing signal by minimizing a performance index with respect to the state vector in the sliding mode. It is concluded the proposed variable structure controller yields better system dynamic performance than that obtained by using the only linear optimal controller inthat responses for a step disturbance have a shorter setting time, no matter what overshoot values and rising time.

• Journal of Information Processing Systems
• /
• v.17 no.3
• /
• pp.441-452
• /
• 2021

Federated learning provides an efficient integrated model for distributed data, allowing the local training of different data. Meanwhile, the goal of multi-task learning is to simultaneously establish models for multiple related tasks, and to obtain the underlying main structure. However, traditional federated multi-task learning models not only have strict requirements for the data distribution, but also demand large amounts of calculation and have slow convergence, which hindered their promotion in many fields. In our work, we apply the rank constraint on weight vectors of the multi-task learning model to adaptively adjust the task's similarity learning, according to the distribution of federal node data. The proposed model has a general framework for solving optimal solutions, which can be used to deal with various data types. Experiments show that our model has achieved the best results in different dataset. Notably, our model can still obtain stable results in datasets with large distribution differences. In addition, compared with traditional federated multi-task learning models, our algorithm is able to converge on a local optimal solution within limited training iterations.

• International conference on construction engineering and project management
• /
• 2007.03a
• /
• pp.509-519
• /
• 2007

Private financing is playing an increasing role in public infrastructure construction projects worldwide. However, private investors/operators are exposed to the financial risk of low profitability due to the inaccurate estimation of facility demand, operation income, maintenance costs, etc. From the operator's perspective, a sound and thorough financial feasibility study is required to establish the appropriate capital structure of a project. Operators tend to reduce the equity amount to minimize the level of risk exposure, while creditors persist to raise it, in an attempt to secure a sufficient level of financial involvement from the operators. Therefore, it is important for creditors and operators to reach an agreement for a balanced capital structure that synthetically considers both profitability and repayment capacity. This paper presents an optimal capital structure model for successful private infrastructure investment. This model finds the optimized point where the profitability is balanced with the repayment capacity, with the use of the concept of utility function and multi-objective GA (Generic Algorithm)-based optimization. A case study is presented to show the validity of the model and its verification. The research conclusions provide a proper capital structure for privately-financed infrastructure projects through a proposed multi-objective model.

• International Journal of High-Rise Buildings
• /
• v.7 no.4
• /
• pp.389-396
• /
• 2018

This paper presents a numerical investigation on the structural response of a multi-story building subjected to spreading multi-compartment fires. A recently proposed simple fire model has been used to simulate two spreading multi-compartment fire scenarios in a 10-story steel-framed office building. By assuming simple temperature rising and distribution profiles in the fire exposed structural components (steel beams, steel column and concrete slabs), finite element simulations using a three-dimensional structural model has been carried out to study the failure behavior of the whole structure in two multi-compartment fire conditions and also in a standard fire condition. The structure survived the standard fire but failed in the multi-compartment fire. Whilst more accurate fire models and heat transfer models are needed to better predict the behaviors of structures in realistic fires, the current study based on very simple models has demonstrated the importance and necessity of considering spreadingmulti-compartment fires in fire resistance design of multi-story buildings.

• 제어로봇시스템학회:학술대회논문집
• /
• 2002.10a
• /
• pp.102.2-102
• /
• 2002

In this paper, the improved Multi-FNN (Fuzzy-Neural Networks) model is identified and optimized using HCM (Hard C-Means) clustering method and optimization algorithms. The proposed Multi-FNN is based on FNN and use simplified and linear inference as fuzzy inference method and error back propagation algorithm as learning rules. We use a HCM clustering and genetic algorithms (GAs) to identify both the structure and the parameters of a Multi-FNN model. Here, HCM clustering method, which is carried out for the process data preprocessing of system modeling, is utilized to determine the structure of Multi-FNN according to the divisions of input-output space using I/O process data. Also, the parame...

• Journal of the Korean Society of Manufacturing Technology Engineers
• /
• v.26 no.1
• /
• pp.30-35
• /
• 2017

The major parts of a vacuum cleaner are molded. The vacuum cleaner works in multi-load conditions. Therefore, the designer needs to optimize the structure and injection molding conditions simultaneously. Here, the main factor of design is the rib shape and thickness. The greater the rib thickness, the greater the stiffness of the structure. However, it causes an increase in weight. On the other hand, the lower the rib thickness, the greater the increase in the injection pressure. However, the weight will be reduced. Therefore, the designer needs to optimize the rib shape and thickness for structure stiffness and injection molding. In order to solve this problem, we propose an optimization method using D.O.E and a response surface model, which is a multi-objective optimization method using the multi-objective genetic algorithm.