Papers by Keyword: Continuous Wavelet Transform

Abstract: A new method for premature ventricular contraction (PVC) detection and classification is presented. The proposed algorithm is constituted of two principal phases: the features extraction and reduction phase and the optimized classification phase. In the first phase, the discrete cosine transform (DCT) and the continuous wavelet transform (CWT) are applied on each ECG beat to generate an augmented features vector. For the optimized classification phase, the radial basis function (RBF) neural network classifier is trained and optimized by the bat algorithm. For the aim of performances evaluation of the proposed method, the MIT-BIH arrhythmia database has been used. Consequently, the BAT-RBF classifier yielded an overall sensitivity of 95,2% and an accuracy of 98,2%, confirming clearly the competitiveness of the proposed method compared to some recent and powerful algorithms.

Abstract: The concept of vibration based condition monitoring technology has been developing at a rapid stage in the recent years suiting to the maintenance of sophisticated and complicated machines. Nowadays, wavelet analysis based signal processing technique is applied as effective tool for condition monitoring. The experimental studies were conducted on the gear testing apparatus to obtain the vibration signal from a healthy gear and an induced faulty gear. In this paper, two different techniques using Laplace wavelet as base function are used to characterize the fault in the gear signals, specifically wavelet enveloped power spectrum and wavelet kurtosis. The wavelet parameters are optimized using genetic algorithm to select most fault related features. A comparative study detailing features of fault characterization is also given in order to understand the effectiveness of both the wavelet based signal processing methods and their fault diagnosis capability.

Abstract: Combustion engines are the components of means of transport which during service are exposed to the occurrence of excessive clearances. The diagnosing of the condition of the engine requires developing methods which could be useful during both continuous and single checks of the engine. Application of measurements and acoustic signal processing may facilitate a quick diagnosis of the engine without any additional disassembly work. As part of the study, research was conducted to develop a non-invasive method of diagnosing excessive clearance of the combustion engine. Based on the recorded acoustic signals of the combustion engine and reference signals of the momentary crankshaft position, a method was developed to diagnose an increased valve clearance in the combustion engine. In this method, processing of the recorded acoustic signals by means of the continuous wavelet transform was applied together with a comparison of their instantaneous energy, depending on the crankshaft rotation angle and timing angles, as well as the scale parameters of time and frequency distributions.

Abstract: This paper presents a wavelet-based method of identification modal parameter and damage detection in a free vibration response. An algorithm for modal parameter identification and damage detection is purposed and complex Morlet wavelet is chosen as an analysis wavelet function. This paper only focuses on identification of natural frequencies of the structural system. The method utilizes both undamaged and damage experiment data of free vibration response of the truss structure system. Wavelet scalogram is utilizes for damage detection. The change of energy components for undamaged and damage structure is investigated from the plot of wavelet scalogram which corresponded to the detection of damage.

Abstract: The division into two stages of the Continuous Wavelet Transform (CWT) computing is proposed. This is expedient in circumstances when CWT is repeated many times, e.g., for online detection of nonstationary signal singularities. It is shown that the preprocessing of wavelet coefficients in the first stage can significantly reduce computing time required in the second stage. The comparative estimation of the runtime reduction in the second stage of CWT calculation is deduced.

Abstract: Analysis the common gear fault signal, using continuous wavelet transform well time-frequency characteristics and failure mechanism gear features, combined with quasi-periodic signal rotary mechanical characteristics of the continuous wavelet transform fault signal. This method can reduce the noise source and other incentives interference, remove the specific needs of the signal to improve signal stripping effect, the use of a certain type of gear failures this step, to get a clear diagnosis results show that this method has proven a strong application space.

Abstract: In this paper, it uses continuous wavelet transform in analysing zhuji station monthly rainfall. from the transformed wavelet coefficients’ real component, variance and modulus square, we can get the main measure contained in zhuji’s monthly rainfall sequence. By anlysing the transformation process of continuous wavelet transform coefficien’s real part of various scales, we can get zhuji station’s monthly rainfall sequence wet-dry transformation process of various scales. By caculated , we found zhuji station’s monthly rainfall sequence contains 10 month , 171 month and 393 month scale. In large scales, zhuji station’s rainfall is in hemiplegia period from 2013 to 2021. in small scale, zhuji station’s rainfall shifts strongly.

Abstract: For a building under evolutionary random excitations, a new random vibration analysis based on equivalent excitation method is presented, in which, the equivalent excitation is constructed by using continuous wavelet transform. When the input power spectral density of a seismic excitation is modulated by a time varying function, the continuous wavelet coefficients are calculated firstly by using scales and the evolutionary power spectral density, then the equivalent excitation is constructed by inverse continuous wavelet transform, finally the random vibration analysis is carried out following the procedures of the equivalent excitation method. An example is given to show the whole procedure, in which the special ability of the method for dealing with the evolutionary random excitation is discussed in detail.

Abstract: It is a challenging problem for scientists and engineers to calculate the random vibration responses of a large scale engineering structure, during which, the calculating efficiency and accuracy are always the main concerns. In civil engineering, there are some even more difficult cases that have to be considered, such as a large span cable-stayed bridge under multi-dimensional seismic loads consisting of evolutionary random processes. In this paper, such a problem is analyzed by using equivalent excitation method, in which the equivalent excitations are constructed by using continuous wavelet transform, and the affections of multi-dimensional seismic excitations are investigated. Simulations of a typical cable-stayed bridge show some of the plausible features of the method presented.

Abstract: For a building under multi-dimensional seismic excitations, a new random vibration analysis based on equivalent excitation method is presented, in which, a new equation of motion of the building is used to depict its dynamics and kinematics features. Since most of the earth shocks are non-stationary random processes, the input power spectral density of the seismic excitations are modulated by a time varying function, and their equivalent excitations are constructed by using continuous wavelet transform, where the continuous wavelet coefficients are calculated by using scales and the evolutionary power spectral density, while the equivalent excitations are constructed by inverse continuous wavelet transform. The random vibration analysis is carried out following the procedures of the equivalent excitation method. A simulation example is given to investigate some affection of the multi-dimensional seismic inputs, especially the rotational one.