Detection of Knocking Combustion Using the Continuous Wavelet Transformation and a Convolutional Neural Network

The phenomenon of knock is an abnormal combustion occurring in spark-ignition (SI) engines and forms a barrier that prevents an increase in thermal efficiency while simultaneously reducing CO2 emissions. Since knocking combustion is highly stochastic, a cyclic analysis of in-cylinder pressure is necessary. In this study we propose an approach for efficient and robust detection and identification of knocking combustion in three different internal combustion engines. The proposed methodology includes a signal processing technique, called continuous wavelet transformation (CWT), which provides a simultaneous analysis of the in-cylinder pressure traces in the time and frequency domains with coefficients. These coefficients serve as input for a convolutional neural network (CNN) which extracts distinctive features and performs an image recognition task in order to distinguish between non-knock and knock. The results revealed the following: (i) The CWT delivered a stable and effective feature space with the coefficients that represents the unique time-frequency pattern of each individual in-cylinder pressure cycle; (ii) the proposed approach was superior to the state-of-the-art threshold value exceeded (TVE) method with a maximum amplitude pressure oscillation (MAPO) criterion improving the overall accuracy by 6.15 percentage points (up to 92.62%); and (iii) The CWT + CNN method does not require calibrating threshold values for different engines or operating conditions as long as enough and diverse data is used to train the neural network.

Damage Detection Method Based on Continuous Wavelet Transformation of Lamb Wave Signals

A damage estimation method based on continuous wavelet transformation (CWT) of the normalized Lamb wave signals is proposed here. Lamb waves are actuated and sensed using piezoelectric (lead zirconate titanate, PZT) transducers arranged in the form of square detection cells on a plate-like structure. Excitation sequences based on pitch–catch and pulse–echo configurations are tested for the same arrangement of the transducers. The possibilities of the existence of damage for each actuator–senor pair are formed by using the normalized coefficient of CWT. The size of the possible damage region is directly controlled through envelopes defined by the coefficients of CWT, and no additional parameter is required to define its size. The aggregate damage image is constructed by the fusion of damage possibilities from all actuator–sensor pairs using damage indices based on conjunctive and compromised fusion schemes. The results indicate that the proposed method can estimate the location and severity of multiple damage with signals directly from the damaged plate, without the need of baseline signals from the undamaged plate, and the time-compensated signals provide better damage imaging than the raw signals. The most accurate and computationally inexpensive combination is the pulse–echo configuration with damage index based on conjunctive image fusion scheme. The method is computationally inexpensive and can be applied for multiple damage estimations in large structures to reduce the evaluation cost and inspection time during on-line structural health monitoring.