As a promising and non-contact detection technique, machine vision has been widely used in fault diagnosis of substation equipment. The rapid and accurate detection of substation equipment in infrared images is one of the key steps for automatic fault diagnosis. However, the complexity of image background, the low contrast of infrared images and the rotational targets in infrared images pose a great challenge to detection task. This study aims to improve the detection accuracy of the model while having real-time detection speed, and propose a lightweight power equipment detection network (PEDNet) based on YOLOv4-tiny. First, a novel global information aggregation module is constructed to guide the network to focus on the salient regions where the target equipment is located. Second, an improved spatial transformer network is introduced to reduce the impact of rotational targets on detection accuracy. Finally, a feature enhanced fusion network is designed through the use of a multi-scale feature cross-fusion structure. It can fully fuse the feature information of the salient region, the rotational targets and the strong semantic information. The experimental results show that the proposed PEDNet can reach 92.66% detection accuracy and 107.07 FPS real-time detection speed on testing datasets. Compared with YOLOv4-tiny, there is a small sacrifice in detection speed, but the detection accuracy is improved and significantly higher than the existing state of the art (SOTA) object detection models.

Regular inspection of photovoltaic (PV) modules is a crucial measure to ensure the safety of solar power generation. To address the challenges posed by small, multi-scale infrared defects in PV modules, this paper proposes a method called FAA-YOLO (Filter and Adaptive assisted YOLO) for detecting small infrared defects in PV modules. Firstly, we design a MPED (Multi-scale Parameter-free Edge Detection) module to enhance shallow features such as texture and edges, thereby improving the network’s capability to extract features from small defect targets. Secondly, we introduce an improved feature fusion network named SlimNet, which mitigates the challenges of small infrared defects in PV modules and the tendency of such features to be overlooked by fusing information from different scales. Lastly, we propose an Adaptive Auxiliary Detection Layer (AAD-Layer). The multi-scale adaptive auxiliary detection layer enables feature reuse and enriches defect feature representation by cascading features of different scales across layers and combining deep and shallow features. Experimental results show that the proposed FAA-YOLO model achieves a detection accuracy of 87.5%. Compared to others methods, it also strikes a good balance between accuracy and computational complexity.

The reagent dosages can directly affect the concentrate grade and recovery in the mineral flotation process. For many years, the dosage has mainly been controlled by human observation of froth image features, especially the bubble size. However, the reagent dosage control based on human experience can easily cause fluctuation of the performance index, which may result in wasted ore resources and chemical reagents. In this paper, a control method of reagent dosages based on bubble size characteristics is proposed. By combining the distribution features of the bubble size, the estimation method of the probability density function of the bubble size was introduced, as was the method of determining the optimal width of the kernel function based on the maximum entropy. The error of the output bubble size PDF and the optimal bubble size PDF was used as the performance indicator to transform the control of the reagent dosages into optimization of seeking the minimum performance indicator. In the process of selecting the best individual in every generation of the DE algorithm, the constraint of the reagent cost is taken into consideration. The experimental results showed the effectiveness of the proposed method.

Nickel-plated punched steel strip is a product obtained by punching holes on the surface of cold-rolled white sheet steel strip and then electrochemical nickel plating. It is necessary to make accurate and fast detection of punching circle parameters, since it is of crucial importance to ensuring the quality of nickel-plated punched steel strips. Accordingly, in this article, an improved circle fitting algorithm of nickel-plated punched steel strip is proposed. Firstly, the least squares fitting is performed to obtain the circle center and radius dataset by iterative algorithm with different values for the initial point positions and intervals. Then, the mean shift algorithm is used to optimize the results after iteration, and the segmented fitted circle centers are all concentrated around the true circle center to obtain the best radius and center coordinates. Finally, comparison experiments with different numbers of circular holes and verification experiments with nickel-plated punched steel strips are carried out. As the results show, the algorithm proposed in this article is more robust than the least squares algorithm in detecting multiple circles and has better real-time performance than the Hough transform. Therefore, it can meet the industrial production needs with high accuracy and real-time requirements, such as nickel-plated punched steel strips.

This paper introduces a novel hybrid control strategy for quadrotor UAVs inspired by neural dynamics. Our approach effectively addresses two common issues: the velocity jump problem in traditional backstepping control and the control signal chattering in conventional sliding mode control. The proposed system combines an outer-loop bioinspired backstepping controller with an inner-loop bioinspired sliding mode controller, ensuring smooth trajectory tracking even under external disturbances. We rigorously analyzed the system’s stability using Lyapunov stability theory. To validate our algorithm’s effectiveness, we conducted trajectory tracking experiments in both disturbance-free and step-disturbance conditions, comparing it with the traditional backstepping control, conventional sliding mode control, and saturated sliding mode control. The results demonstrate that our algorithm not only tracks trajectories more effectively but also significantly outperforms these methods in suppressing velocity jumps and signal chattering.