УДК 620.179
ADAPTED SPLIT SPECTRUM PROCESSING FOR ULTRASONIC SIGNAL IN THE PULSE ECHO TEST
T. Bouden1, F. Djerfi1, M. Nibouche2 1 LEND laboratory, Department of Automatic, Jijel University, Algeria 2 Department of Engineering Design and Mathematics, UWE University Bristol, UK Е-mail: Bouden_toufik@yahoo.com, Corresponding Author Djerfi.fares@yahoo.com andmokhtar.nibouche@uwe.ac.uk
Abstract. In this paper, an Adaptive Split Spectrum Processing technique (A-SSP) is proposed, to improve ultrasonic echoes detection. It is an arrangement of conventional Split Spectrum Processing (SSP) with an empirical method of analyzing nonlinear and non-stationary signals, called Empirical Mode Decomposition (EMD). This proposed technique allows breaking up the signal into several bands of frequencies in an adaptive way and intrinsic to the treated signal using EMD. It enables us to know the internal contents and the local changes of the ultrasonic signal and makes the detection of any desired targets more flexible for the coherent noise problem. In the combination phase of A-SSP, a linear operation for selected intrinsic mode functions and a non linear one for non selected intrinsic mode functions are used to reconstruct the signal with separated echoes. To evaluate the proposed techniques (A-SSP with different combination operations), firstly a mortar specimen with artificial defect is used to resolve the defects detection and localization problem. Secondly a paste cement specimen is also used to resolve the materials characterization problem. The signals were obtained using a technique applied in pulse-echo mode, known as the prism technique. Numerical and experimental tests were performed to verify the effectiveness and reliability of the proposed technique and to show its excellent performances.
Key words: Non-Destructive Testing (NDT), paste cement, mortar, propagation velocity, thickness, detection and localization, Split Spectrum Processing (SSP), Empirical Mode Decomposition (EMD).
ОБРАБОТКА УЛЬТРАЗВУКОВОГО СИГНАЛА С АДАПТИВНЫМ РАЗДЕЛЕНИЕМ СПЕКТРА ДЛЯ ИМПУЛЬСНОГО ЭХОМЕТОДА
Т. Боден1, Ф. Джерфи2, М. Нибуш2
1 Лаборатория LEND, Факультет автоматики, Университет Джиджель,
Алжир
2 Факультет технического проектирования и математики, Университет
UWE, Бристоль, Великобритания
E-mails: bouden_toufik@yahoo.com; djerfi.fares@yahoo.com; mokhtar.
nibouche@uwe.ac.uk
Предложен метод адаптивного разделения спектра эхосигнала (А-SSP) для улучшения детектирования у.з. эхосигналов. Это модификация традиционного метода разделения спектра эхосигнала (SSP), включающая эмпирический метод анализа нелинейных и нестационарных сигналов, называемый разделением на эмпирические моды (EMD). Предлагаемая методика позволяет адаптивно разделять сигналы на несколько частотных диапазонов и выделять внутренние особенности в обработанном сигнале, используя EMD. Это позволяет узнать внутреннее содержание и локальные изменения у.з. сигналов и делает обнаружение любых необходимых объектов более адаптированным к проблеме когерентного шума. На этапе комбинирования методом A-SSP линейные операции для выбранных внутренних собственных функций и нелинейные — для остальных внутренних собственных функций используются для реконструкции сигнала по выбранным эхоимпульсам. Чтобы оценить предложенные алгоритмы (метод А-SSP с различными операциями комбинирования) использованы образцы: во-первых, из известкового раствора с искусственным дефектом для рассмотрения проблемы обнаружения и локализации дефекта; во-вторых, из цементной пасты для решения задачи характеристики материала. Эхосигналы получены методом клина, применяемого при реализации эхоимпульсного метода. Для подтверждения эффективности и надежности предложенной методики и демонстрации ее превосходной вычислительной производительности выполнены расчетные и экспериментальные проверки.
Ключевые слова: неразрушающий контроль, цемент, известковый раствор, скорость распространения, толщина, обнаружение и локализация, метод разделения спектра эхосигнала, разделение на эмпирические моды.
1. INTRODUCTION
Ultrasonic flaw detection and classification in the presence of high scattering microstructure noise (i. e. clutter echoes), plays an important part in the nondestructive testing and evaluation of the materials, and it's the significant and challenging problem in this field [1—14]. The discrimination between grain and flaw echoes often requires additional processing or scanning to distinguish them. Several approaches have been developed in earlier works to exploit ultrasonic signal signature differences between the flaw and grain scatters which have motivated time and frequency diversity techniques [1—7, 15—24]. Clutter echoes exhibit randomness and are sensitive to frequency bands. However, flaw echoes are less vulnerable to frequency variations. Therefore, frequency diverse signal decomposition can be advantageous in differentiating the flaw information from the clutter echoes. It is the objective of this work.
This paper presents a novel signal processing approach called Adaptive Split Spectrum Processing (A-SSP) for enhancing defect detection in materials consisting of grain-like microstructures and materials characterization. This novel technique combines the conventional split spectrum processing with the empirical mode decomposition for characterizing ultrasonic signal signature in NDT of materials using the pulse echo test [21—31]. In A-SSP, it is also proposed a new scheme for recombining the outputs of the using filters by linear and nonlinear operations.
In this paper the mortar with artificial defect and paste cement specimens are used to show the relative performances of the proposed A-SSP and to present a comparative studies looking to the conventional SSP.
This paper is organized as follows: Section 1 gives a brief introduction of ultrasonic testing and flaw detection problem. Section 2 gives a brief description of previous works related to SSP that impacted its development. Section 3 presents EMD and its principles. Section 4 presents the implementation details of the proposed A-SSP technique. Section 5 describes the experimental set-up and simulation results demonstrating relative performances between the proposed technique (A-SSP) and the conventional one (SSP). Section 6 gives the main conclusions of this paper.
2. SPLIT SPECTRUM PROCESSING
The SSP first published in 1979 and it is considered as sub-optimal to the problem of the ultrasonic pulse in a structural noise. This technique splits the received wideband signal into a group of frequency-diverse narrowband through a filter bank, exhibiting different signal-to-noise ratios (SNR), and subsequently recombines them nonlinearly to enhance flaw visibility. This approach is limited to the detection of single targets or multiple targets having similar spectral characteristics. The frequency diversity of this technique decorrelated the grain echo amplitudes, while leaving the target echo amplitudes highly correlated over the frequency bands, using various compounding algorithms that involved a nonlinear operation. Although significant flaw visibility enhancement can be achieved using SSP with nonlinear operations, their effectiveness is mainly limited. Previous works have shown that the SSP technique is fairly sensitive to frequency region used to obtain the narrowband signals. Consequently, the performance of SSP is strongly influenced by the utilized spectral region for processing [1—15, 17, 23].
3. EMPIRICAL MODE DECOMPOSITION (EMD)
The empirical mode decomposition method is developed from the simple assumption that any signal consists of different simple intrinsic mode oscillations. This method of signal decomposition was introduced for analyzing nonlinear and non-stationary signals. It was initially proposed for the study of ocean waves, and
found immediate applications in biomedical engineering. The major advantage of EMD is that the basis functions are derived directly from the signal itself. Hence, the analysis is adaptive, in contrast to Fourier analysis, where the basis functions are linear combinations of fixed sinusoids [16, 17, 19—22, 23—31]. This nonlinear technique has been pioneered by N.E. Huang et al. for adaptively representing non-stationary signals as sums of zero-mean amplitude modulated and frequency modulated (AM-FM) components [26]. It has been successfully applied to many problems in signal analysis that were not satisfactorily addressed by Fourier analysis. Despite these practical successes it has lacked a firm theoretical foundation.
The principle of EMD is to decompose a signal into a sum of oscillatory functions, namely intrinsic mode functions (IMFs), that: 1) have the same numbers of extrema and zero-crossings or differ at most by one; and 2) are symmetric with respect to local zero mean. With these two requirements, the meaningfully instantaneous frequency of an IMF can be well defined. Otherwise, if blindly applied to any signal, the instantaneous frequency may result in a few paradoxes [24—31]: it may go beyond the band for band limited signal or it may not represent one of the frequencies in the Fourier spectrum in the global sense. To make use of EMD, the signal must have at least two extrema one maximum and one minimum to be successfully decomposed into IMFs.
So the starting point of the EMD is to consider oscillations in signals at a very local level. In fact, if we look at the evolution of a signal x(t) between two consecutive extrema, we can heuristically define a (local) high-frequency part, or local detail, which corresponds to the oscillation terminating at the two minima and passing through the maximum which necessarily exists in between them. For the picture to be complete, one still has to identify the corresponding (local) low-frequency part, or local trend. The decomposition method involves the following steps:
Step 1. Given a signal x(n), finding the local maxima and minima of the
Для дальнейшего прочтения статьи необходимо приобрести полный текст. Статьи высылаются в формате PDF на указанную при оплате почту. Время доставки составляет менее 10 минут. Стоимость одной статьи — 150 рублей.