Recently, the Kolmogorov-Arnold Network (KAN) has been proposed, significantly outperforming MLP in terms of interpretability and symbolic representation. In practice, KANs are required to fit data to extremely high precision. For instance, in typical applications of KAN like inferring precise equations from data, high accuracy is an intrinsic requirement. In the current architecture of KAN, cubic B-spline basis functions were selected as the approximate tools. However, the fixed degree and knots in B-splines limit the activation function’s adaptability to complex signals. Due to these inherent limitations of B-spline functions, especially low-order and computational complexity, KAN still has room for improvement in accuracy. In this paper, we propose the Legendre-KAN that can enhance the degrees of freedom of the basis functions in the KAN. Compared to the traditional Spline-KAN, Legendre-KAN utilizes parameterized Legendre basis functions and normalization layers at the edges of the KAN. Benefiting from higher-order orthogonal polynomials, Legendre-KAN achieves tighter theoretical error bounds compared to Spline-KAN. Extensive experiments show that Legendre-KAN achieves higher accuracy and parameter efficiency, achieving up to 10 times higher accuracy than Spline-KAN in some cases. For those functions which can be symbolized, this leads to more correct results as opposed to Spline-KAN.

The key to semi-supervised medical image segmentation lies in how to leverage a large amount of unlabeled data to improve model performance. Previous methods typically set a threshold based on confidence or uncertainty to select reliable pixels from unlabeled data in order to improve performance. However, these methods often overlook the fact that reliable pixels may contain numerous incorrect labels and that potentially valuable unreliable pixels are ignored. This paper proposes a simple and efficient semi-supervised medical image segmentation method, named MCPG. We design two modules: Multiple-constraints Exploration (ME) and Prototype-guided Noise Identification (PNI). The ME module applies multiple constraints to select reliable pixels. Subsequently, the PNI module uses these reliable pixels as class prototypes, calculates the distance between unreliable pixels and these prototypes, and identifies the unreliable pixels with correct labels. Based on two modules, MCPG effectively utilizes a large number of reliably predicted correct pixels and unreliable pixels, while being unaffected by threshold settings. Experiments conduct on Left Atrium and Pancreas datasets demonstrate that MCPG achieves superior results compared to other state-of-the-art methods. Our code is available at https://github.com/HengfanLi/MCPG.

There are significant scale and morphological differences between fire and smoke features in forest fire detection. This paper proposes a detection method based on dual-branch multi-scale adaptive feature fusion network (DMAFNet). In this method, convolutional neural network (CNN) and transformer are used to form a dual-branch backbone network to extract local texture and global context information, respectively. In order to overcome the difference in feature distribution and response scale between the two branches, a feature correction module (FCM) is designed. Through space and channel correction mechanisms, the adaptive alignment of two branch features is realized. The Fusion Feature Module (FFM) is further introduced to fully integrate dual-branch features based on the two-way cross-attention mechanism and effectively suppress redundant information. Finally, the Multi-Scale Fusion Attention Unit (MSFAU) is designed to enhance the multi-scale detection capability of fire targets. Experimental results show that the proposed DMAFNet has significantly improved in mAP (mean average precision) indicators compared with existing mainstream detection methods.

B-spline functions necessitate piecewise interval calculations under De Boor algorithm, while Legendre polynomials enable global domain operations without segmentation. This characteristic renders Legendre polynomials computationally more efficient. Motivated by this efficiency, we introduce Legendre-KAN, a reformulated version of Kolmogorov-Arnold Networks (KAN) that substitutes B-spline basis functions with Legendre polynomials. Our approach reparameterizes the network's weights using Legendre polynomials, which not only substantially reduces training time but also preserves competitive approximation accuracy, albeit with marginally higher root-mean-square error (RMSE) in certain tasks. Experimental results confirm that Legendre-KAN provides a computationally efficient alternative to traditional KAN architectures and surpasses MLP modules in terms of training accuracy. This work offers a promising avenue for addressing the inherent time-cost challenges in KAN.

In recent years, multi-agent reinforcement learning techniques have been widely used and evolved in the field of artificial intelligence. However, traditional reinforcement learning methods have limitations such as long training time, large sample data requirements, and highly delayed rewards. Therefore, this paper systematically and specifically studies the MARL algorithm. Firstly, this paper uses Citespace software to visually analyze the existing literature on multi-agent reinforcement learning and briefly indicates the research hotspots and key research directions in this field. Secondly, the applications of traditional reinforcement learning algorithms under two task objects, namely single-agent and multi-agent systems, are described in detail. Then, the paper highlights the diverse applications, challenges, and corresponding solutions of MARL algorithmic techniques in the field of MAS. Finally, the paper points out future research directions based on the existing limitations of the algorithm. Through this paper, readers will gain a systematic and in-depth understanding of MARL algorithms and how they can be utilized to better address the various challenges posed by MAS.

Multi-focus image fusion is an effective image fusion technology, which aims to combine source images from different focal planes of the same scene to obtain a good fusion result. This means that the fused image will focus on all focal planes, that is, it contains more abundant scene information. The development of deep learning promotes the great progress of image fusion, and the powerful feature extraction and reconstruction ability of neural network makes the fusion result promising. In recent years, more and more multi-focus image fusion methods based on deep learning have been proposed, such as convolutional neural network (CNN), generative adversarial network (GAN) and automatic encoder, etc. In order to provide effective reference for relevant researchers and technicians, firstly, this paper introduces the concept of multi-focus image fusion and some evaluation indicators. Then, it analyzes more than ten advanced methods of multi-focus image fusion based on deep learning in recent years, discusses the characteristics and innovation of various methods, and summarizes their advantages and disadvantages. In addition, it reviews the application of multi-focus image fusion technology in various scenes, including photographic visualization, medical diagnosis, remote sensing detection and other fields. Finally, it proposes some challenges faced by current multi-focus image fusion related fields and looks forward to future possible research trends.

Traditional medical image sensors face multiple challenges. First, these sensors typically rely on large amounts of labeled data, which are time-consuming and costly to obtain. Second, when the data volume and image size are large, traditional sensors have limited computational power, making it difficult to effectively train and infer models. Additionally, traditional sensors have poor generalization ability and struggle to adapt to datasets with different modalities. This paper devises a novel framework, named LSDSL, and deploys it in the sensor. LSDSL utilizes low-quality sensor data for semi-supervised learning in medical image segmentation. in supervised learning, we devise the hard region exploration (hre) module to enhance the model’s comprehension of low-quality pixels in hard regions. in unsupervised learning, we introduce a pseudo-label sharing (ps) module, which allows low-quality pixels in one network to learn from the high-quality pixels in the other networks. our model outperforms other semi-supervised methods on the datasets of two different modalities (CT and MRI) in medical image sensors, achieving superior inference speed and segmentation accuracy.

The significance of low-quality data in unlabeled medical images is always underestimated. We believe that these underestimated data contain valuable information that remains largely unexplored. We present a novel uncertainty-guided different levels of pseudolabels (UDLP) framework to explore the underestimated data in medical images. The framework consists of a student-teacher model that uses uncertainty to classify the pseudolabels predicted by the teacher model into three levels: high confidence, low confidence, and unreliability. The student model learns directly from high-confidence pseudolabels. By using the confident learning method in low-confidence pseudolabels, the teacher model corrects the noisy labels in low-confidence voxels to provide positive feature information for the student model. We design a method for removing unreliable pseudolabels, to further enhance model's generalizability. The proposed framework UDLP is evaluated on two datasets and demonstrates superior performance compared to other state-of-the-art methods.

In recent years, in the field of Multi-Agent Systems (MAS), significant progress has been made in the research of algorithms that combine Reinforcement Learning (RL) with Attention Mechanism (AM). However, there is a lack of comprehensive reviews in this field. Based on this, this paper does the following work. Firstly, it reviews the classical algorithms of RL and AM; Secondly, it systematically introduces the combination of RL and AM; Thirdly, it sorts out their application in the field of single-agent and multi-agent, and pays attention to and looks forward to the challenges and future research prospects in this field; The last part offers a comprehensive analysis of the challenges encountered by research in this field and anticipates future research paths. The research offers a conceptual understanding and theoretical foundation for future applications of RL using AM and facilitates further in-depth study in this field for researchers.

Temporal Action Detection (TAD) aims to accurately capture each action interval in an untrimmed video and to understand human actions. This paper comprehensively surveys the state-of-the-art techniques and models used for TAD task. Firstly, it conducts comprehensive research on this field through Citespace and comprehensively introduce relevant dataset. Secondly, it summarizes three types of methods, i.e., anchor-based, boundary-based, and query-based, from the design method level. Thirdly, it summarizes three types of supervised learning methods from the level of learning methods, i.e., fully supervised, weakly supervised, and unsupervised. Finally, this paper explores the current problems, and proposes prospects in TAD task.

To solve the problems of wear and corrosion failure in cold-working die steel during long-term service, Ni60 coatings with different WC contents are prepared on 12CrMoV substrate using laser cladding technology. The forming characteristics, microstructure, mechanical properties, and corrosion resistance of the coatings are characterized using X-ray diffraction, scanning electron microscopy, energy dispersive spectroscopy, friction and wear testing machines, and electrochemical workstations. The results indicate that the additive amount of the WC can change the solidification characteristics of the melt pool, thereby reducing the width of the coatings and increasing the depth of the coatings. When the additive amount of the WC is 30% (mass fraction), the hardness of the coating is 853 HV0. 3, and the average friction coefficient is only 0. 467. It has excellent mechanical properties, but the toughness of the coating decreases, leading to the formation of pores during the solidification process, which reduces the corrosion resistance of the coating. When the additive amount of the WC is 10% and 20% respectively, the size of the grain structure is refined, the grain boundary area is reduced, and it has excellent corrosion resistance. When the additive amount of the WC is 20%, the corrosion current density of the coating is only 3. 65 x 10(-5) A/cm(2), and the mechanical properties are only slightly lower than the coating obtained with 30% WC additions. Therefore, this study explored suitable WC/Ni60 coatings for wear and corrosion environments, determined the optimal additive amount of the WC, expanded the application range of WC/Ni60 coatings in mold steel, and has guiding significance for engineering practice.

In recent years,the combination of reinforcement learning and attention mechanisms has attracted an increasing attention in algorithmic research field.Attention mechanisms play an important role in improving the performance of algorithms in reinforcement learning.This paper mainly focuses on the development of attention mechanisms in deep reinforcement learning and examining their applications in the multi-agent reinforcement learning domain.Relevant researches are conducted accordingly.Firstly,the background and development of attention mechanisms and reinforcement learning are introduced,and relevant experimental platforms in this field are also presented.Secondly,classical algorithms of reinforcement learning and attention mechanisms are reviewed and attention mechanism is categorized from different perspectives.Thirdly,practical applications of attention mechanisms in the reinforcement field are sorted out based on three types of tasks including fully cooperative,fully competitive and mixed,with focus on the application in the field of multi-agent.Finally,the improvement of attention mechanisms on reinforcement learning algorithms is summarized.The challenges and future prospects in this field are discussed.

Aluminum–air batteries are characterized as “green energy for the 21st century” due to their clear advantages in terms of high current discharge, high specific energy, low cost, and easy-to-obtain electrode materials. This study develops the SLM aluminum anode quality prediction model and evaluates its learning and training results using the BP neural network architecture. By altering the network topology of the SLM aluminum anode quality prediction model, we create a process parameter backpropagation model that takes advantage of the extremely adaptable capabilities of artificial neural networks. The quick and exact selection of process parameters meets the goals of density, self-corrosion current, and anode usage, hence improving the forming quality and processing efficiency of SLM aluminum anodes. The experimental results show that the process parameter backpropagation model’s parameter configurations match to the real densities and self-corrosion currents, which are somewhat higher than the specified target values. The maximum error rate for the aluminum anode forming quality prediction model is 8.23%. Furthermore, the actual anode utilization rate is somewhat lower than the projected target value, indicating that the backpropagation model can satisfy actual production needs.

In this paper, a novel robust iterative learning control strategy for linear discrete systems in a finite frequency domain is proposed. The strategy first converts the iterative learning control process into a discrete linear repeating system model by means of the equivalence transformation technique, and employs a convex bounded uncertainty domain to define the uncertainty range of the system. Subsequently, the control law design problem in discrete linear repetitive systems is solved by utilizing the generalized KYP lemma, which is transformed into the problem of solving linear matrix inequalities (LMIs). The designed control law is able to satisfy the robustness requirement in the direction of the iteration axis and ensure the monotonic convergence of the dynamic error between each experiment, so as to realize the expected control objective step by step. Finally, the advantages and practicality of the method are experimentally verified, and its potential limitations and constraints are analyzed and summarized.

Fire incidents pose a significant threat to human life and property security. Accurate fire detection plays a crucial role in promptly responding to fire outbreaks and ensuring the smooth execution of subsequent firefighting efforts. Fixed-size convolutions struggle to capture the irregular variations in smoke and flames that occur during fire incidents. In this paper, we introduce FireViT, an adaptive lightweight backbone network that combines a convolutional neural network (CNN) and transformer for fire detection. The FireViT we propose is an improved backbone network based on MobileViT. We name the lightweight module that combines deformable convolution with a transformer as th DeformViT block and compare multiple builds of this module. We introduce deformable convolution in order to better adapt to the irregularly varying smoke and flame in fire scenarios. In addition, we introduce an improved adaptive GELU activation function, AdaptGELU, to further enhance the performance of the network model. FireViT is compared with mainstream lightweight backbone networks in fire detection experiments on our self-made labeled fire natural light dataset and fire infrared dataset, and the experimental results show the advantages of FireViT as a backbone network for fire detection. On the fire natural light dataset, FireViT outperforms the PP-LCNet lightweight network backbone for fire target detection, with a 1.85% increase in mean Average Precision (mAP) and a 0.9 M reduction in the number of parameters. Additionally, compared to the lightweight network backbone MobileViT-XS, which similarly combines a CNN and transformer, FireViT achieves a 1.2% higher mAP while reducing the Giga-Floating Point Operations (GFLOPs) by 1.3. FireViT additionally demonstrates strong detection performance on the fire infrared dataset.

针对无人机在复杂环境下受到多种威胁时的航迹规划问题,提出一种改进的基于球面向量的粒子群优化算法(ISPSO).利用融合压缩因子和异步变化学习因子的ISPSO算法,通过粒子位置和速度同无人机转角和爬升角的对应关系,高效地搜索无人机的构形空间,找到成本函数最小的最优路径.为了评估ISPSO的性能,从真实的数字高程模型地图中生成2个基准场景,仿真结果表明,该算法优于基于球面向量的粒子群算法.

The practical cross-regional, rapid, and large-scale deployment of unmanned aerial vehicle (UAV) clusters can effectively be improved by using large UAVs to carry several micro-UAVs for launching operations.In contrast to a UAV taking off from the ground, when multiple micro-UAVs are launched into motion, the UAV body has a nonzero velocity and nonzero attitude, that is, a nonstationary initial state.Combining existing methods for UAV clusters cannot solve the dynamic optimization problem of the group trajectory in a nonstationary initial state.A consensus artificial potential field (APF)-based cooperative method is proposed.This method requires only the position and attitude data of adjacent UAVs, which can reduce sensor data as much as possible.First, a geometric control method and dynamic constraints are used to realize the single-body pose stabilization control in a nonstationary initial state.Second, the consensus control and APF method are combined on the basis of the single-body stabilization control to realize UAV group formation and dynamic collision avoidance with minimum time and position errors.Finally, in the nonstationary initial state, on the basis of the proposed dynamic model and control method, a whole-process simulation of a large UAV carrying 40 micro-UAVs is designed to verify the effectiveness of the proposed method.

针对江苏省应用型本科院校教师继续教育存在的问题,对一流应用型本科院校教师继续教育体系做创新性探索,提出落实立德树人根本任务,将文化传承创新培训纳入教师教育培训内容,科学厘定应用型本科院校教师继续教育理念及目标、改革继续教育课程、构建多元的教师继续教育模式等举措.根据江苏省对地方普通高校高质量考核实施方案,制定应用型本科院校教师继续教育评价指标,健全继续教育效果评估体系,加强继续教育保障体系建设,以进一步提高应用型本科院校教师的综合素质,助力江苏省应用型本科院校高质量发展.

Increasing the proportion of clean energy is an inevitable trend of high-quality energy development. While enjoying the benefits of clean energy, traditional fossil energy enterprises face great challenges. In this study, coal power and photovoltaic power are selected as examples for analysis. Considering coal consumption, carbon emission, unit ramp conditions and photovoltaic abandonment, a comprehensive income model containing thermal power and photovoltaic power generation is proposed. The objective function of maximizing earnings is established and solved by genetic algorithm. The results show that the overall benefit will increase if the proportion of photovoltaic does not increase blindly. The unreasonable ratio of thermal and photovoltaic power and the imbalance of supply and consumption will lead to the decline of economic benefits and the increase of coal consumption and carbon emission. In the case of a high proportion of clean energy penetration, on the one hand, traditional power plants should be responsible for ensuring the instability of electricity output when the photovoltaic is weak, on the other hand, they need to give up some of the supply proportion when clean power is at a high output state. Therefore, while increasing the proportion of clean energy, it is necessary to improve the speed of load adjustment, the depth of peak regulation and the efficiency of low load work in traditional power plants to ensure the healthy development of new energy. (c) 2023 The Author(s). Published by Elsevier Ltd. This is an open access article under theCCBY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).