• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.53 no.1
• /
• pp.40-49
• /
• 2004

In this paper, we propose a new architecture of Genetic Algorithms(GAs)-based Self-Organizing Polynomial Neural Networks(SOPNN), discuss a comprehensive design methodology and carry out a series of numeric experiments. The conventional SOPNN is based on the extended Group Method of Data Handling(GMDH) method and utilized the polynomial order (viz. linear, quadratic, and modified quadratic) as well as the number of node inputs fixed (selected in advance by designer) at Polynomial Neurons (or nodes) located in each layer through a growth process of the network. Moreover it does not guarantee that the SOPNN generated through learning has the optimal network architecture. But the proposed GA-based SOPNN enable the architecture to be a structurally more optimized network, and to be much more flexible and preferable neural network than the conventional SOPNN. In order to generate the structurally optimized SOPNN, GA-based design procedure at each stage (layer) of SOPNN leads to the selection of preferred nodes (or PNs) with optimal parameters- such as the number of input variables, input variables, and the order of the polynomial-available within SOPNN. An aggregate performance index with a weighting factor is proposed in order to achieve a sound balance between approximation and generalization (predictive) abilities of the model. A detailed design procedure is discussed in detail. To evaluate the performance of the GA-based SOPNN, the model is experimented with using two time series data (gas furnace and NOx emission process data of gas turbine power plant). A comparative analysis shows that the proposed GA-based SOPNN is model with higher accuracy as well as more superb predictive capability than other intelligent models presented previously.

• Journal of the Korean Society for Precision Engineering
• /
• v.21 no.7
• /
• pp.76-82
• /
• 2004

The solutions of neighboring optimal control are typically obtained using the sweep method or transition matrices. Due to the numerical integration, however, the gain matrix can become infinite as time go to final one in the transition matrices, and the Riccati solution can become infinite when the final time free. To overcome these disadvantages, this paper proposes the pseudospectral Legendre method which is to first discreteize the linear boundary value problem using the global orthogonal polynomial, then transforms into an algebraic equations. Because this method is not necessary to take any integration of transition matrix or Riccati equation, it can be usefully used in real-time operation. Finally, its performance is verified by the numerical example for the space vehicle's orbit transfer.

• Journal of Institute of Control, Robotics and Systems
• /
• v.22 no.6
• /
• pp.429-434
• /
• 2016

This paper introduces a stabilization condition for polynomial fuzzy systems that guarantees $H_{\infty}$ performance under the imperfect premise matching. An $H_{\infty}$ control of polynomial fuzzy systems attenuates the effect of external disturbance. Under the imperfect premise matching, a polynomial fuzzy model and controller do not share the same membership functions. Therefore, a polynomial fuzzy controller has an enhanced design flexibility and inherent robustness to handle parameter uncertainties. In this paper, the stabilization conditions are derived from the polynomial Lyapunov function and numerically solved by the sum-of-squares (SOS) method. A simulation example and comparison of the performance are provided to verify the stability analysis results and demonstrate the effectiveness of the proposed stabilization conditions.

• Journal of Korean Institute of Industrial Engineers
• /
• v.27 no.4
• /
• pp.379-383
• /
• 2001

The restricted shortest path problem is known to be weakly NP-hard and solvable in pseudo-polynomial time. Four fully polynomial approximation schemes (FPAS) are available in the literature, and most of these are based on pseudo-polynomial algorithms. In this paper, we propose a new FPAS that can be easily derived from a combination of a set of standard techniques. Although the complexity of the suggested algorithm is not as good as the fastest one available in the literature, it is practical in the sense that it does not rely on the bound tightening phase based on approximate binary search as in Hassin's fastest algorithm. In addition, we provide a review of standard techniques of existing works as a useful reference.

• Proceedings of the KIEE Conference
• /
• 2000.11d
• /
• pp.800-802
• /
• 2000

In this paper, we propose the Self-Organizing Networks(SON) based on competitive Fuzzy Polynomial Neuron(FPN) for the optimal design of nonlinear process system. The SON architectures consist of layers with activation nodes based on fuzzy inference rules. Here each activation node is presented as FPN which includes either the simplified or regression Polynomial fuzzy inference rules. The proposed SON is a network resulting from the fusion of the Polynomial Neural Networks(PNN) and a fuzzy inference system. The conclusion part of the rules, especially the regression polynomial uses several types of high-order polynomials such as liner, quadratic and modified quadratic. As the premise part of the rules, both triangular and Gaussian-like membership functions are studied. Chaotic time series data used to evaluate the performance of our proposed model.

• Proceedings of the KIEE Conference
• /
• 2000.07d
• /
• pp.2946-2948
• /
• 2000

In this paper, we proposed the Fuzzy Polynomial Neural Networks(FPNN) model with fuzzy activation node. The proposed FPNN structure is generated from the mutual combination of PNN(Polynomial Neural Networks) structure and fuzzy inference system. The premise of fuzzy inference rules defines by triangular and gaussian type membership function. The fuzzy inference method uses simplified and regression polynomial inference method which is based on the consequence of fuzzy rule expressed with a polynomial such as linear, quadratic and modified quadratic equation are used. The structure of FPNN is not fixed like in conventional Neural Networks and can be generated. The design procedure to obtain an optimal model structure utilizing FPNN algorithm is shown in each stage. Gas furnace time series data used to evaluate the performance of our proposed model.

• Bulletin of the Korean Mathematical Society
• /
• v.46 no.1
• /
• pp.183-198
• /
• 2009

In this paper, we derive a characterization of orthonormal balanced multiwavelets of order p in terms of the continuous moments of the multiscaling function $\phi$. As a result, the continuous moments satisfy the discrete polynomial preserving properties of order p (or degree p - 1) for orthonormal balanced multiwavelets. We derive polynomial reproduction formula of degree p - 1 in terms of continuous moments for orthonormal balanced multiwavelets of order p. Balancing of order p implies that the series of scaling functions with the discrete-time monomials as expansion coefficients is a polynomial of degree p - 1. We derive an algorithm for computing the polynomial of degree p - 1.

• Korean Journal of Remote Sensing
• /
• v.19 no.5
• /
• pp.363-370
• /
• 2003

This paper describes a rectification procedure that relies on a polynomial model derived from the imaging geometry without loss of accuracy. By using polynomial model, one can effectively eliminate the iterative process to find an image pixel corresponding to each output grid point. With the imaging geometry and ephemeris data, a geo-location polynomial can be constructed from grid points that are produced by solving three equations simultaneously. And, in order to correct the local distortions induced by the geometry and terrain height, a distortion model has been incorporated in the procedure, which is a function of incidence angle and height at each pixel position. With this function, it is straightforward to calculate the pixel displacement due to distortions and then pixels are assigned to the output grid by re-sampling the displaced pixels. Most of the necessary information for the construction of polynomial model is available in the leader file and some can be derived from others. For validation, sample images of ERS-l PRI and Radarsat-l SGF have been processed by the proposed method and evaluated against ground truth acquired from 1:25,000 topography maps.

• Journal of the Korea Society of Computer and Information
• /
• v.19 no.11
• /
• pp.175-181
• /
• 2014

This paper suggests O(m) polynomial time heuristic algorithm to obtain the solution for the driver scheduling problem, DSP, that has been classified as NP-complete problem. The proposed algorithm gets the initial assignment of n minimum number of drivers from given m schedules. Nextly, this algorithm gets the minimum total time (TC) using 5 rules of swap and insert for decrease of over times (OT) and idle times (IT). Although this algorithm is a heuristic polynomial time algorithm with O(m) time complexity rules to be find a optimal (or approximate) solution, this algorithm is equal to metaheuristic methods for the 5 experimental data. To conclude, this paper shows the DSP is not NP-complete problem but Polynomial time (P)-problem with polynomial time rules.

• Journal of KIISE:Computer Systems and Theory
• /
• v.31 no.10
• /
• pp.562-574
• /
• 2004

We address the problem of energy-optimal voltage scheduling for fixed-priority hard real-time systems. First, we prove that the problem is NP-hard. Then, we present a fully polynomial time approximation scheme (FPTAS) for the problem. for any $\varepsilon$>0, the proposed approximation scheme computes a voltage schedule whose energy consumption is at most (1+$\varepsilon$) times that of the optimal voltage schedule. Furthermore, the running time of the proposed approximation scheme is bounded by a polynomial function of the number of input jobs and 1/$\varepsilon$. Experimental results show that the approximation scheme finds more efficient voltage schedules faster than the best existing heuristic.