Feature Extraction Of Acoustic Emission Signals For Crack Detection Of The Jacket Offshore Structure Models

Fixed offshore platforms are widely employed in the offshore oil-gas exploration and jacket offshore platforms are the common ones of fixed offshore platforms. Tubular joint fatigue failures have been commonly regarded as the design problem for jacket offshore platforms. When a fatigue fracture occurs in the node of jacket offshore platforms, an early diagnosis is the key in hostile ocean environments.Therefore it is extremely significant to detect the crack of offshore platforms using the acoustic emission (AE) technique. As we all know, extracting features is the key of the AE technique. To extract features effectively, signal processing-based methods are widely used today. Due to the fact that most of the AE signals present non-stationary and nonlinear properties, it is essential to choose appropriate signal processing methods that are suitable for non-stationary and nonlinear signals to extract AE signals features.The time-frequency analysis methods are widely studied in AE signals processing because they can provide both time and frequency domain information of a signal simultaneously. However, the time-frequency analysis methods such as windowed Fourier transforms(WFT) and wavelet transform have their own limitations. Recently, a novelty for non-stationary signals named as Local Wave Analysis (LWA), has been put forward and confirmed to be superior to the other signal processing methods in many applications. Supported by National Natural Science Foundation, this dissertation introduces LWA into AE signals processing, whose aim is to extract feature of AE signals by using LWA. In this paper we show the possibility of using local-wave to analyze the time-frequency feature of the acoustic emission signals produced by the crack in the offshore structure model.In the investigation, we used a local wave decomposition technique, allowing time series of acoustic emission signal being decomposed into a small number of intrinsic mode function components (IMF). Under the Hilbert transformation process, IMF can be translated into an expression called Hilbert spectra, which exhibits the amplitude-frequency-time distribution of the data. The marginal spectra, which present the energy-frequency distribution of the data, were obtained by integrating the Hilbert spectra with time. The feature of the offshore structure simulation acoustic emission signals could be extracted by applying local wave analysis. The characteristics of the crack acoustic emission signals in the offshore structure, was found which indicated the acoustic emission occurrence by using the local-wave analyzing. Consequently, the experimental results show that the proposed approach is able to effectively capture the significant information reflecting the acoustic emission in the offshore structure, and thus has good potential in the field of acoustic emission signal feature extraction.This thesis presents a new approach to characterize the acoustic emission signals of the structure cracking in the process of loading based on the Approximate Entropy (ApEn), which is a statistical measure that quantifies the regularity of a time series. The conception and nature are introduced. Successful application has been achieved to analyze the simulating acoustic emission signals and the acoustic emission signals produced by the crack in the steel tube. The results show that ApEn has obvious high ability to quantify the complexity of signals, thereby providing a new effective tool for the acoustic emission signals processing.A new approach through combining the recently developed Local Wave method with the Approximate Entropy to characterize the acoustic emission signals was studied in the thesis. Firstly, local wave method is used to decompose the acoustic emission signal into a number of intrinsic mode functions (IMFs), and then calculate the ApEn of IMFs to describe their complexity, detect the occurrence and the development and quantify the characteristic of the acoustic emission signals. The effectiveness of the proposed methods has been demonstrated by using the acoustic emission signals from the steel tube cracking during a quasi-static loadings test. The experimental results show that the proposed approach can effectively capture the significant information reflecting the acoustic emission, and thus has good potential in the field of acoustic emission signal feature extraction.This dissertation presents a new approach, which is combined the recently developed Local Wave method with the neural network to characterize and identify the acoustic emission signals of offshore structures. Local wave method is used to decompose the acoustic emission signals of offshore structures into a number of intrinsic mode functions, and then energy feature parameter extracted from IMFs could be served as input parameter of neural networks to identify the acoustic emission signals of offshore structures. The experimental analysis results from the acoustic emission signals of offshore structures model show that the approach of neural network based on local wave extracting energy parameter as feature can effectively recognise the offshore structures AE signals, and thus providing a new effective tool for the acoustic emission signal feature extraction identification of offshore structures.This paper presents an identification platform of acoustic emission signals of offshore structures supported by the open country-wide interconnected database and based on local wave method, which can manage a lot of acoustic emission signal experimental data of offshore structures and study the nature of the acoustic emission signals by multi-recognition arithmetic. By using the program tools such as PowerBuilder and Matlab, in combination with database technique, the data and algorithm of acoustic emission signals of offshore structures were combined by designing several interfaces. The results show that the identification platform has powerful functions with easy operation, and has more practical values. And the system can offer convenience for managing and identifying the experimental data and the practical data of acoustic emission signals of offshore structures.