To address challenges in the detection of wind turbine blade damage images, characterized by complex backgrounds and multiscale feature distribution, we propose a method based on an enhanced YOLOV8 model. Our approach focuses on three key aspects: First, we enhance the extraction of small target features by integrating the CBAM attention mechanism into the backbone network. Second, the feature fusion process is refined using the Weighted Bidirectional Feature Pyramid Network (BiFPN) to replace the path aggregation network (PANet). This modification prioritizes small target features within the deep features and facilitates the fusion of multiscale features. Lastly, we improve the loss function from CIoU to EIoU, enhancing sensitivity to small targets and the perturbation resistance of bounding boxes, thereby reducing the gap between computed predictions and real values. Experimental results demonstrate that compared with the YOLOV8 model, the CBAM-BiFPN-YOLOV8 model exhibits improvements of 1.6%, 1.0%, 1.4%, and 1.1% in precision rate, recall rate, mAP@0.5, and mAP@0.5:.95, respectively. This enhanced model achieves substantial performance improvements comprehensively, demonstrating the feasibility and effectiveness of our proposed enhancements at a lower computational cost.

To improve the wear monitoring accuracy of steel tribo-pairs, a ferromagnetic particle sensor that consists of a magnetic loop with two cylindrical magnetic poles and one magnetic housing is proposed. The induction coil is placed inside the magnetic poles to shield against external electromagnetic interference. The air gap is designed to 1 mm between the two magnetic poles, where a rotational symmetric high-gradient static magnetic field will be generated under constant direct current (dc) drive. The sine-wave-like signal of one period will be induced when one ferromagnetic particle flows through the sensor. The working principle of the sensor and the internal magnetic field distribution is explained by establishing a mathematical model and finite element analysis (FEA). Furthermore, the results about the effects of coil length and position show that the output voltage is not sensitive to the axial mounting positions of the induction coil, and a shorter inductive coil produces a larger output voltage. By the oil dynamic monitoring experiment, the sensor can distinguish 13- $\mu \text{m}$ ferromagnetic particles in an oil tube of 10-mm inner diameter. With its high sensitivity to ferromagnetic particles, the sensor is suitable for online wear monitoring of steel tribo-pairs.

Inductive wear particle sensors have been widely studied due to their ability to monitor the wear status of equipment in real time. To detect wear particles more precisely, a wear particle sensor based on multiple inductive coils under a toroidal magnetic field is proposed in this paper. With inductive coils located near a stronger magnetic field, the sensor can not only ensure the flow rate, but also enhance wear particle detection accuracy. The proportional relationship between the signal and the wear particle velocity and the excitation current is discussed in this paper. The crosstalk phenomenon among different inductive coils was also investigated by single particle experiments and lubricating oil dynamic experiments. In the dynamic oil experiment, the sensor detected $13~\mu \text{m}$ particles under a flow rate of 570 mL/min, which satisfied most industrial online monitoring demands.

Wear debris sensing is an important approach for machine condition monitoring. To date, different types of debris sensors have been developed, among which the magnetostatic inductive debris sensor is preferable for monitoring ferromagnetic debris, but the existing induced voltage model cannot be used to quantify wear debris in lube oil. Accordingly, a modeling method was proposed for the debris sensor based on a high-gradient magnetostatic field. First, a Gaussian curve is used to fit the magnetic field along the axis of the sensor. Then, time-harmonic electromagnetic field analysis and Fourier series decomposition are combined to establish an induced voltage model. Furthermore, the influence of the relative permeability, size, and velocity of wear debris and induction coil parameters on the output voltage is analyzed. The results show a linear relationship between the theoretical calculations and experimental results. Therefore, the proposed model can be used to realize quantitative analysis of wear debris, laying a foundation for optimizing the sensor structure, and extracting features from output voltage signals.

Accurate and reliable fault diagnosis is one of the key and difficult issues in mechanical condition monitoring. In recent years, Convolutional Neural Network (CNN) has been widely used in mechanical condition monitoring, which is also a great breakthrough in the field of bearing fault diagnosis. However, CNN can only extract local features of signals. The model accuracy and generalization of the original vibration signals are very low in the process of vibration signal processing only by CNN. Based on the above problems, this paper improves the traditional convolution layer of CNN, and builds the learning module (local feature learning block, LFLB) of the local characteristics. At the same time, the Long Short-Term Memory (LSTM) is introduced into the network, which is used to extract the global features. This paper proposes the new neural network—improved CNN-LSTM network. The extracted deep feature is used for fault classification. The improved CNN-LSTM network is applied to the processing of the vibration signal of the faulty bearing collected by the bearing failure laboratory of Inner Mongolia University of science and technology. The results show that the accuracy of the improved CNN-LSTM network on the same batch test set is 98.75%, which is about 24% higher than that of the traditional CNN. The proposed network is applied to the bearing data collection of Western Reserve University under the condition that the network parameters remain unchanged. The experiment shows that the improved CNN-LSTM network has better generalization than the traditional CNN.