Saša Nikolić

A CLASS OF ALMOST ORTHOGONAL FILTERS

In this paper we define a new class of the almost orthogonal filters. These filters are a generalization of the classical orthogonal filters commonly used in the circuit theory, control system theory, signal processing and process identification. Almost orthogonal filters generate the series of almost orthogonal Legendre functions on the interval (0, ∞). They can be successfully used for the analysis, synthesis and designing of imperfect technical systems. They can also be used for approximation of arbitrary functions on (0, ∞). A new method for obtaining the models of imperfect systems is presented as well. Simulations were performed in order to validate the theoretical results and the comparison of new filters with classical filters is given.

Modelling of dynamical systems based on almost orthogonal polynomials

A new class of the almost orthogonal filters is described in this article. These filters are a generalization of the classical orthogonal filters commonly used in the circuit theory, control system theory, signal processing, and process identification. Almost orthogonal filters generate the series of almost orthogonal Legendre functions over the interval (0, ∞). It is well known that all real systems suffer from some imperfections, so the models of these systems should reflect this fact. A new method for obtaining an imperfect system model is proposed. This method uses an almost orthogonal filter, which is based on almost orthogonal functions. Experiments with modular servo drive were performed to validate theoretical results and demonstrate that the method described in the article is suitable for modelling of imperfect systems.

Digital linear phase notch filter design based on IIR all-pass filter application

A new procedure for design of linear phase IIR notch filters is presented in this paper. Resulted filter is realized using parallel connection of two IIR all-pass filters with approximately linear phase. Adequate amplitude characteristic is obtained by approximation of all-pass filter phase. Amplitude characteristic satisfies all prescribed demands but it is not equiripple no matter both all-pass filters have equiripple phase. Algorithm exhibits fast convergence and easy initial values determination.

Approximation based on orthogonal and almost orthogonal functions

Abstract In this paper, we define a class of almost orthogonal rational functions of Legendre type in a new manner. Relations of these functions with classical exponentional functions orthogonal over interval (0, ∞), as well as classical polynomials orthogonal over (0, 1) are explained. Defining relations of these functions can be used for designing almost orthogonal filters. These filters are generators of orthogonal signals and can be successfully applied in finding the best signal approximation in the sense of the mean square error. The filters orthogonal property enables building of physical (in this case electrical) models of dynamical systems (the sources of signals to be approximated) either with less components for the same model accuracy or higher accuracy for the same number of components than the other known models. New filters represent further improvement of previously designed filters, by the same authors, in the sense of simplicity, higher accuracy, lesser approximation time and even a possibility to approximate signals generated by systems with built-in imperfections. Series of experiments were performed to analyze the dependence of approximation accuracy and the number of filters sections.

Multi-channel descriptors and ensemble of Extreme Learning Machines for classification of remote sensing images

Abstract We present a method for real-time scene classification which achieves high accuracy without a time consuming descriptor learning step and kernelized classifiers. Robustness of the classification is achieved by combining the powerful multi-channel Gabor-based descriptors and an ensemble of Extreme Learning Machines (ELM). We first extend the recently introduced Binary Gabor Patterns (BGP) to multi-channel images. This is done by extracting BGP over several color channels and embedding an additional compact color layout descriptor. Then we propose an effective method for the aggregation of multiple ELMs into a single classification system, which leads to significant classification accuracy improvements. The experimental evaluation demonstrates that multi-channel color information constantly improves classification results. The integration of multiple ELMs into an ensemble using the proposed aggregation strategy significantly outperforms linear SVM in terms of accuracy, and reaches results similar to the non-linear SVM while operating in real time. Therefore, an ensemble of ELMs with the proposed aggregation strategy could be used as an efficient alternative to the non-linear SVM for remote sensing image classification tasks.

Optimal fuzzy sliding mode control with a time-varying sliding surface

It is well known that sliding mode control can achieve good transient performance and system robustness. However, the sensitivity of the sliding mode control before reaching the sliding surface and the presence of chattering may introduce problems to the tracking control of the system. In this paper, a new sliding surface design approach is proposed that varies the sliding surface in a nonlinear and time-varying fashion. Slope constant of the sliding surface is varied by fuzzy controller based on the ratio between the derivate of error and error itself. Optimal parameters of the fuzzy controller are determined using genetic algorithms. Proposed optimal fuzzy sliding mode control has been applied to an electrical servo drive system and the performed experiments have verified efficiency and improvements of a new control method.

Systems modeling based on Legendre polynomials

A new method for obtaining models of continuous systems, based on Legendre orthogonal filters, is proposed in this paper. Recent results in the field of orthogonal functions were used to improve accuracy of modeling. First, Legendre orthogonal filters were designed on the basis of orthogonal functions. Then, adjustable models were formed. Models parameters were optimized using genetic algorithm. As a case study, an experimental simple DC drive was considered and described in details. Simulations were performed out to approve theoretical results and demonstrate that the method described in the paper is very suitable for modeling continuous systems. It achieves excellent results in the sense of model accuracy and modeling algorithm speed.

Transitional Butterworth-Chebyshev recursive digital filters

A design for transitional Butterworth–Chebyshev recursive digital filters is proposed. These filters show gradual transition and, at the same time, maintain the equiripple peaks. An analysis of the cutoff slope of these filters is also presented. It is shown that the first type of filter is suitable for the design of wideband low-pass filters, and the second one for the design of narrowband low-pass filters.

Fuzzy sliding mode control for anti-lock braking systems

A combination of two control methods, sliding mode and fuzzy, for the wheel slip control in anti-lock braking system, is presented in this paper. The fuzzy block is used to determine the values of key parameters important for establishing switching function dynamics. It is demonstrated, via performed experiments, that proposed control algorithm gives good system performances.

Hierarchical ELM ensembles for visual descriptor fusion

Abstract Extreme Learning Machines (ELM) have been successfully applied to variety of classification problems by utilizing a single descriptor type. However, a single descriptor may be insufficient for the visual classification task, due to the high level of intra-class variability coupled with low inter-class distance. Although several studies have investigated methods for combining multiple descriptors by ELM, they predominantly apply a simple concatenation of descriptors before classifying them. This type of descriptor fusion may impose problems of descriptor compatibility, high dimensionality and restricted accuracy. In this paper, we propose a hierarchical descriptors fusion strategy at the decision level (“late-fusion”), which relies on ELM ensembles (ELM-E). The proposed method, denoted as H-ELM-E, effectively combines multiple complementary descriptors by a two-level ELM-E based architecture, which ensures that a more informative descriptors will gain more impact on the final decision. In the first level, a separate ELM-E classifier is trained for every image descriptor. In the second level, the output scores from the previous level are aggregated into the mid-level representation which is conducted to an additional ELM-E classifier. The exhaustive experimental evaluation confirmed that the proposed hierarchical ELM-E based strategy is superior to the single-descriptor methods as well as “early fusion” of multiple descriptors, for the visual classification task. Additionally, it was shown that significant accuracy improvement is achieved by integrating ensembles of ELM as a basic classifier, instead of using a single ELM.