Abstract Path planning is the key part in the process of transportation conducted by logistics robots, and there often exist some problems with it. The path designed is not always smooth enough and its search efficiency is low, for example. As a common global path planning algorithm, A‐star is based on the traditional algorithm, which is unable to solve the problem of uneven path in the movement of logistics robots. Through improving the heuristic function of the traditional A‐star algorithm, weighing the heuristic function dynamically, removing the redundant points of the traditional star algorithm path with Floyd algorithm, and setting a safe distance to prevent the logistics robot from collision at the same time, the path is finally curved to be more appropriate to the movement path of the logistics robot. The MATLAB simulation of A‐star algorithm before and after the improvement shows that the turning points of the advanced A‐star algorithm reduced 61.5% on average compared to the traditional algorithm. The path length decreased 2.4% and the traversing points reduced 58.5%. At the same time, the DWA algorithm introduces dynamic weight coefficients, which can dynamically adjust the weight coefficients when encountering obstacles, so as to safely reach the target point.

Abstract Transformerless grid‐connected inverters (GCIs) are universally utilized in the PV system. However, they have the shoot‐through issue and common mode leakage current (CMLC). To address the above‐mentioned issues, a transformerless common‐ground GCI without shoot‐through issue is proposed. The proposed topology has one common ground between the PV array and the grid that eliminates the CMLC. The topology is the combination of two buck‐boost converters, so it has the boost capability. Each buck‐boost converter operates in half‐line period, so the proposed inverter is a single‐stage system. Only two switches operate in high frequency in positive half‐line period and only one switch operates in high frequency in negative half‐line period. Therefore, efficiency of the system can be improved. There is no current flowing through the body diode of the MOSFET. Additionally, the presented inverter has no shoot‐through problem. The operating principle is elaborated. At last, simulation and experimental results of a 250 W proposed inverter validate the theoretical analysis. The maximum efficiency of the system is 96.2%.

Abstract Pattern recognition of optical label is utilized not only in packet switches but also in the field of data security due to its wide range of applications. In particular, pattern recognition using waveguide devices is highly valued because it eliminates the need for encoder processing. However, existing methods suffer from the problem of degraded contrast ratio because half of the input power are emitted from non‐target waveguides. In this article, a method is proposed to improve the contrast ratio employing a 44 multi‐mode interference (MMI) coupler with multiple input signals. The theoretical results reveal that the proposed method improves the contrast ratio from 3.0 dB to 9.5 dB and dB for the three‐input and four‐input cases, respectively, for quadrature phase‐shift keying (QPSK)‐modulated signal. Furthermore, numerical simulation was performed through the three dimension finite‐difference time‐domain (3D‐FDTD) method, and the proposed scheme successfully discriminated each QPSK pattern. The power value at all ports closely matches the theoretical value, and the ratios of 9.5 dB and 30.0 dB are numerically obtained.

Abstract Planar object tracking (POT) is crucial in vision‐based robotic applications. Despite significant advancements, tracking planar objects under real‐world conditions remains a challenge owing to various factors. The method proposed by the authors leverages SIFT features and generalized Hough transform, which is enhanced by spatial locality, to mitigate background clutter and false matching. The experimental results demonstrate that this method significantly outperforms existing approaches, achieving higher precision across various alignment error thresholds.

Abstract The data scale keeps growing by leaps and the majority of it is high‐dimensional imbalanced data, which is hard to classify. Data missing often happens in software which further aggravates the difficulty of classifying the data. In order to resolve high‐dimensional imbalanced mixed‐variables missing data classification problem, a novel method based on particle swarm optimization is developed. It has three original components including multiple feature selection, mixed attribute imputation, and quantum oversampling. Multiple feature selection uses a two‐stage strategy to obtain stable relevant features. Mixed attribute imputation separates features into continuous and discrete features and fills missing values with different models. Quantum oversampling chooses instances to balance data based on the quantum operator. Furthermore, particle swarm optimization is employed to optimize the parameters of the components to obtain preferable classification results. Six representative classification datasets, six typical algorithms, and four measures are taken to conduct exhaust experiments, and results indicate that the proposed method is superior to the comparison algorithms.

Abstract Semantic labelling of remote sensing images is crucial for various remote sensing applications. However, the dense distribution of man‐made and natural objects with similar colours and geographic proximity poses challenges for achieving consistent and accurate labelling results. To address this issue, a novel deep learning model incorporating an octave convolutional neural networks (CNNs) within an end‐to‐end U‐shaped architecture is presented. The approach differs from conventional CNNs in that it employs octave convolutions instead of standard convolutions. This strategy serves to minimize low‐frequency information redundancy while maintaining segmentation accuracy. Furthermore, coordination attention is introduced in the encoder module to enhance the network's ability to extract useful features, focusing on spatial and channel dependencies within the feature maps. This attention mechanism enables the network to better capture channel, direction, and location information. In conclusion, the U‐shaped network is engineered with a completely symmetric structure that employs skip connections to merge low‐resolution information, used for object class recognition, with high‐resolution information to enable precise localization. This configuration ultimately improves segmentation accuracy. Experimental results on two public datasets demonstrate that our U‐ONet achieves state‐of‐the‐art performance, making it a compelling choice for remote sensing image semantic labelling applications.

Abstract With the accelerated deployment of 5G massive multi‐input multi‐output (MIMO)‐structured base stations (BSs), accurate and efficient testing of massive MIMO arrays is essential to ensure that the 5G massive MIMO antenna array can perform as expected. The increasing trend of both antenna element and array physical size makes massive MIMO antenna pattern measurement in the direct‐far‐field OTA environment unrealistic due to the signal attenuation, long measurement time, and ultra‐high test system construction cost. Therefore, the necessity of a compact, efficient, low‐complex, yet accurate OTA test method is evident, especially for 5G massive MIMO BS arrays. This paper proposes a novel multi‐probe‐enabled midfield (MF) OTA test method for 5G massive MIMO devices in compact measurement settings where antenna pattern measurement can be efficiently performed. The detailed theoretical analysis for the 5G massive MIMO array MF antenna pattern reconstruction method is presented, and the simulated validation results based on the typical 5G BS antenna arrays are provided and analysed, which demonstrate the effectiveness of the proposed MF OTA test method and can provide insights into the radio frequency parametric measurement for 5G massive MIMO devices.

Abstract Visual defect recognition techniques based on deep learning models are crucial for modern industrial quality inspection. The backbone, serving as the primary feature extraction component of the defect recognition model, has not been thoroughly exploited. High‐performance vision transformer (ViT) is less adopted due to high computational complexity and limitations of computational resources and storage hardware in industrial scenarios. This paper presents LSA‐Former, a lightweight transformer architectural backbone that integrates the benefits of convolution and ViT. LSA‐Former proposes a novel self‐attention with linear computational complexity, enabling it to capture local and global semantic features with fewer parameters. LSA‐Former is pre‐trained on ImageNet‐1K and surpasses state‐of‐the‐art methods. LSA‐Former is employed as the backbone for various detectors, evaluated specifically on the PCB defect detection task. The proposed method reduces at least 18M parameters and exceeds the baseline by more than 2.2 mAP.

Abstract Promising synaptic behaviour has been exhibited by memristors based on natural organic materials. Such memristor‐based neuromorphic systems offer notable benefits, including environmental sustainability, low production and disposal costs, non‐volatile storage capability, and bio/Complementary Metal‐Oxide‐Semiconductor (CMOS) compatibility. Here, a 256‐level honey memristor‐based neuromorphic system is experimentally evaluated for image recognition. In detail, first, 256‐level honey memristors are manufactured and tested based on in‐house technology; next, the non‐linear characteristics and inherent variation of honey memristor devices, which lead to imprecise weight updates and limit the inference accuracy, are investigated. Experimental results indicate that the inference accuracy of the 256‐level honey memristor‐based neuromorphic system is greater than 88% without cycle‐to‐cycle variations and 87% with cycle‐to‐cycle variations for different optimization algorithms. The overall performance of optimization algorithms with and without variation is compared in terms of energy and latency, where the momentum algorithm consistently outperforms the rest of the algorithms. This 256‐level honey memristor is a promising alternative enabling sustainable neuromorphic systems, encouraging further research into natural organic materials for neuromorphic computing.