2025年 05期
动态场景下基于激光雷达的SLAM算法
LiDAR-based SLAM Algorithm for Dynamic Environments
冯迎宾;李亚玮;王天龙;曾城锋; FENG Yingbin;LI Yawei;WANG Tianlong;ZENG Chengfeng;Shenyang Ligong University;使用激光雷达在动态场景下实现精确的位姿估计与地图映射是同时定位与建图(simultaneous localization and mapping, SLAM)研究领域的重要内容之一，但动态环境中物体移动会导致SLAM算法精度下降，为此提出一种低成本且可有效剔除动态影响的激光雷达SLAM算法。首先引入深度图投影，通过检测相邻时刻深度图之间的像素值波动，筛选并去除动态点云；然后进行地面点云分割，利用非地面点云的特征实现位姿估计和地图映射，利用地面点云的特征施加地面约束，限制高度漂移；最后引入回环检测矫正全局姿态。实验结果表明，与LOAM、LeGO-LOAM和SuMa算法相比，本文算法可更有效剔除动态目标，提供更优秀的定位建图效果和鲁棒性能。
To achieve LiDAR-based precise pose estimation and mapping in dynamic environments is a big challenge in the field of simultaneous localization and mapping(SLAM).Dynamic objects can significantly interfere with a robot's environmental perception and autonomous navigation.To address the decline in SLAM accuracy caused by moving objects in dynamic environments, this study proposes a low-cost yet effective LiDAR SLAM algorithm to eliminate dynamic interference.The algorithm introduces depth map projection to identify and remove dynamic point clouds by detecting pixel value fluctuations between consecutive depth maps.Ground point cloud segmentation is then performed, leveraging non-ground point cloud features for pose estimation and mapping.Ground point cloud features are utilized to impose ground constraints, mitigating height drift.Additionally, loop closure detection is incorporated to correct global pose drift.Experimental results demonstrate that, compared to LOAM,LeGO-LOAM,and SuMa, the proposed algorithm more effectively filters out dynamic objects, delivering superior localization and mapping performance with enhanced robustness.
基于改进YOLOv8的脑肿瘤检测模型
Brain Tumor Detection Model Based on Improved YOLOv8
张傲;刘微;刘阳;杨思瑶;管勇;李波;刘芳菲; ZHANG Ao;LIU Wei;LIU Yang;YANG Siyao;GUAN Yong;LI Bo;LIU Fangfei;为应对磁共振成像(MRI)中形态复杂和边界不规则的脑肿瘤检测，提出一种改进的YOLOv8模型TumorNet-YOLO。该模型通过三项创新模块提升检测性能：自适应感受野卷积模块增强多尺度肿瘤特征的提取能力，降低漏检率；分割融合卷积模块通过多尺度特征融合，增强浅层和深层特征的协同作用；可变形融合模块优化对不规则肿瘤区域的检测，提升模型在复杂MRI背景下的鲁棒性。实验结果表明，TumorNet-YOLO在脑肿瘤检测数据集Br35H中表现优异，平均精度均值mAP@0.5为96.6%,mAP@0.5∶0.95为73.8%。此外，模型计算量(GFLOPs)为8.6,显著优于现有方法。为了验证模型的泛化能力，在BCCD和BTOD数据集上进行了对比实验，结果显示TumorNet-YOLO在mAP@0.5和mAP@0.5∶0.95等多个指标上超越了YOLOv8n,表明TumorNet-YOLO可为脑肿瘤检测和医学图像分析提供更为有效的解决方案。
In order to address the challenges of detecting brain tumors with complex shapes and irregular boundaries in Magnetic Resonance Imaging(MRI),an improved YOLOv8 model, TumorNet-YOLO,is proposed.This model enhances detection performance through three innovative modules: the Adaptive Receptive Field Convolution Module improves the ability to extract multi-scale tumor features and reduce false negatives; the Segmentation Fusion Convolution Module strengthens the synergy between shallow and deep features through multi-scale feature fusion; the Deformable Fusion Module optimizes the detection of irregular tumor regions and improves the model's robustness in complex MRI backgrounds.Experimental results show that TumorNet-YOLO performs excellently on the brain tumor detection dataset Br35H,with a mean Average Precision mAP@0.5 of 96.6% and mAP@0.5∶0.95 of 73.8%.Furthermore, the model's computational cost is 8.6 GFLOPs, which significantly outperforms existing methods.To validate the model's generalization ability, comparative experiments were conducted on the BCCD and BTOD datasets.The results demonstrate that TumorNet-YOLO outperforms YOLOv8n in multiple metrics, including mAP@0.5 and mAP@0.5∶0.95,indicating that TumorNet-YOLO provides an effective solution for brain tumor detection and medical image analysis.
基于超螺旋算法的T-S模糊广义系统动态滑模控制器的设计
Super-Twisting Algorithm-based Dynamic Sliding Mode Controller for T-S Fuzzy Descriptor System
李翔宇;孙鹏然;高宪文;袁春华; LI Xiangyu;SUN Pengran;GAO Xianwen;YUAN Chunhua;Northeastern University;为消除滑模控制中的振颤问题，设计一种基于超螺旋算法的T-S模糊广义系统动态滑模控制器。首先，构造一个新型的动态滑模面，利用广义系统描述滑模运动的动态过程，同时使用基于输出反馈动态滑模的控制方法进行系统设计，并利用凸优化与状态输入增强相结合的方法分析出系统渐近稳定的充分条件；其次，采用多变量超螺旋算法设计二阶且连续控制器，可有效减缓滑模控制系统中存在的振颤问题。实验结果表明，所提出的控制器可以使系统状态在有限时间内到达滑模面，并保持稳定的滑动模态，同时也有效消除系统中的振颤问题，提升系统的控制性能。
To address the chattering problem in sliding mode control, a dynamic sliding mode controller for T-S fuzzy descriptor system based on the super-twisting algorithm is designed.Firstly, a novel dynamic sliding surface is constructed.The process of dynamic sliding mode motion is described using a descriptor system.Additionally, a control method based on output feedback of dynamic sliding mode is employed for system design.The algorithm combines convex optimization with status-input augmentation to analyze sufficient conditions for asymptotic stability of the system.Secondly, a second-order and continuous controller is designed using a multivariable super-twisting algorithm, which effectively alleviates the chattering inherent in sliding mode control systems.Experimental results demonstrate that the proposed controller enables the system state to reach the sliding surface within a finite time and maintain in a stable sliding mode, effectively eliminating the chattering and enhancing the control performance of the system.
基于HDBO-LSTM的短期电力负荷预测方法
Short-term Power Load Forecasting Method Based on HDBO-LSTM
蔡春雷;刘微;任腾腾; CAI Chunlei;LIU Wei;REN Tengteng;为提高电力系统短期电力负荷预测的准确性，提出一种基于混合蜣螂优化(hybrid dung beetle optimization, HDBO)算法优化长短期记忆(long short-term memory, LSTM)网络的预测模型(HDBO-LSTM)。首先，为克服原始蜣螂优化(dung beetle optimization, DBO)算法全局搜索能力较弱且易陷入局部最优的问题，在原始DBO算法的基础上引入随机对立学习策略、哈里斯鹰优化算法、逐维高斯变异策略和动态处理机制，以此形成HDBO算法，增强算法的搜索能力和收敛速度，并通过10个基准函数的对比实验验证HDBO算法的搜索性能；其次，采用HDBO算法优化LSTM网络的超参数，以减小随机超参数对负荷预测精度的影响。使用电工数学建模竞赛的电力负荷数据集对模型进行评估，结果显示，在数据集内随机选取的七天预测任务中，HDBO-LSTM模型的均方根误差(RMSE)、平均绝对百分比误差(MAPE)、平均绝对误差(MAE)和变异系数(COV)较原始LSTM模型分别降低了22.38%、16.33%、19.16%和22.52%,该模型能够有效提升短期电力负荷预测的准确性。
To improve the accuracy of short-term power load forecasting in the power system, a prediction model(HDBO-LSTM) based on hybrid dung beetle optimization(HDBO) algorithm to optimize long short-term memory(LSTM) network is proposed.Firstly, to solve the problem of weak global search ability and susceptibility to local optima in the original dung beetle optimization(DBO) algorithm, the HDBO algorithm is formed by introducing random opposition learning strategy, Harris hawks optimization algorithm, dimension-by-dimension Gaussian variation strategy, and dynamic boundary handling mechanism based on the original DBO algorithm.This enhances the search ability and convergence speed of the algorithm, and the search performance of the HDBO algorithm is verified through comparative experiments with 10 benchmark functions.Secondly, the HDBO algorithm is used to optimize the hyperparameters of the LSTM network, in order to reduce the impact of random hyperparameter selection on the accuracy of load forecasting.Finally, the model is evaluated by using the power load dataset from the electrical mathematics modeling competition.The results show that compared to the original LSTM model, the HDBO-LSTM model reduces root mean square error(RMSE),mean absolute percentage error(MAPE),mean absolute error(MAE),and coefficient of variation(COV) by 22.38%,16.33%,19.16%,and 22.52%,respectively, in the randomly selected seven day prediction task within the dataset.This model can effectively improve the accuracy of short-term power load forecasting.
基于MSEF-YOLO的儿童腕部骨折检测算法
Pediatric Wrist Fracture Detection Algorithm Based on MSEF-YOLO
宫硕;蒋强;李婷雪; GONG Shuo;JIANG Qiang;LI Tingxue;针对儿童腕部X光图像中细微骨折检测精度较低的问题，提出一种基于YOLOv11n改进的MSEF-YOLO(multi-scale efficient fusion network-YOLO)目标检测算法。首先，将空间和通道重构卷积(SCConv)模块与C3k2模块融合，通过空间重构单元(SRU)和通道重构单元(CRU)并行处理空间与通道的冗余，增强对小目标的感知能力；其次，引入多尺度扩张注意力(MSDA)机制提高特征提取能力，进而提高模型检测精度与泛化性，有效减少漏检和误检；最后，优化尺度序列特征融合(SSFF)模块并设计SSFF-X模块，通过3D卷积增强多尺度特征融合能力，进一步提升对细微骨折的检测效果。实验结果表明，相较于原YOLOv11n算法，MSEF-YOLO算法的精确率、召回率、mAP@0.5和mAP@0.5～0.95分别提高了3.1%、3.8%、3.0%和3.5%。MSEF-YOLO算法能够有效协助放射科医生检测儿童腕部骨折，为医学图像的诊断提供技术支持。
To address the issue of low detection accuracy for subtle fractures in pediatric wrist X-ray images, an improved YOLOv11n-based object detection algorithm called MSEF-YOLO(multi-scale efficient fusion network-YOLO)is proposed.First, the spatial and channel reconstruction convolution(SCConv)module is integrated with the C3k2 module, utilizing the spatial reconstruction unit(SRU)and channel reconstruction unit(CRU)to process spatial and channel redundancy in parallel, thereby enhancing the perception of small objects.Second, the multi-scale dilated attention(MSDA)mechanism is introduced to improve feature extraction capability, thereby enhancing detection accuracy and generalization ability while effectively reducing missed and false detections.Finally, the scale sequence feature fusion(SSFF)module is optimized and the SSFF-X module is designed, leveraging 3D convolution to enhance multi-scale feature fusion, further improving the detection performance for subtle fractures.Experimental results demonstrate that, compared to the original YOLOv11n algorithm, the MSEF-YOLO algorithm improves precision, recall, mAP@0.5,and mAP@0.5～0.95 by 3.1%,3.8%,3.0%,and 3.5%,respectively.The MSEF-YOLO algorithm effectively assists radiologists in detecting pediatric wrist fractures, providing technical support for medical image diagnosis.
基于脉冲位置调制的帧变换差分混沌移位键控系统
Frame-transform Differential Chaotic Shift Keying System Based on Pulse Position Modulation
隋涛;韩佳依;杜智豪;于其豪; SUI Tao;HAN Jiayi;DU Zhihao;YU Qihao;为解决传统差分混沌移位键控(differential chaotic shift keying, DCSK)系统安全性较差及误码性能受限的问题，提出一种结合脉冲位置调制与动态帧变换技术的DCSK(frame-transform DCSK system based on pulse position modulation, FT-PPM-DCSK)系统，实现安全性与误码性能的协同优化。通过推导系统在加性高斯白噪声(additive white Gaussian noise, AWGN)信道下的误比特率(BER)表达式，基于蒙特卡洛仿真验证其性能。实验结果表明，与基准DCSK系统相比，FT-PPM-DCSK系统提高了误码性能且具有更高的安全性。
To solve the problems of poor security and limited bit error performance of traditional differential chaotic shift keying system(DCSK),a differential chaotic shift keying system combining pulse position modulation and dynamic frame transform technology(FT-PPM-DCSK)is proposed.Based on the hybrid modulation strategy of frame transformation, the system uses pulse position modulation(PPM)to modulate part of the information bits to achieve the collaborative optimization of security and bit error performance.The bit error rate(BER)expression of the system in additive white Gaussian noise(AWGN)channel is derived, and its performance is verified by Monte Carlo simulation.The experimental results show that, compared with the benchmark DCSK system, the FT-PPM-DCSK system improves the BER performance and has higher security.
基于电容层析成像的包装大米水分检测传感器研究
Research on Capacitance Tomography Sensor for Moisture Detection of Packaged Rice
蒋莹莹;邓梦瑶;杜卓昕;周越;石天玉;杨东; JIANG Yingying;DENG Mengyao;DU Zhuoxin;ZHOU Yue;SHI Tianyu;YANG Dong;Academy of National Food and Strategic Reserves Administration;National Engineering Research Center of Grain Storage and Logistics;包装大米水分快速无损检测是确保其在储运环节质量安全的有效手段。首先，为了探究包装大米水分的非破坏性检测方法，开展基于电容层析成像(ECT)技术的研究，通过数值模拟分析，构建一套适用于包装粮实际尺寸的三维立体式ECT传感器仿真模型(40 cm×12 cm×60 cm),并完成其正问题求解。其次，进一步对阵列电极传感器的性能参数进行优化研究，综合分析传感器极板的宽度、高度、极板上下间距、屏蔽层厚度等结构参数对测量空间的灵敏度分布均匀性、电容的动态测量范围及重构图像相对误差三项评价指标的影响规律。通过设计正交试验，对比分析各个结构参数对三项指标的评价结果，最终确定了一组性能最优的ECT传感器结构参数，即极板高为7 cm、极板宽为5 cm、极板上下间距为6.5 cm、屏蔽层厚度为6.5 cm。最后，采用Landweber算法进行图像重构并完成效果验证，结果表明重构图像能够基本映射出高介电常数区域(高含水率大米)分布位置，但交界处图像质量还有待提升。研究结果有望为包装成品粮水分ECT检测系统的研发提供理论依据。
Rapid nondestructive detection of moisture in packaged rice is an effective means to ensure its quality and safety in storage and transportation.Firstly, in order to explore the non-destructive moisture detection method of packaged rice, research based on capacitance tomography(ECT) technology is carried out, and a set of three-dimensional stereo ECT imaging sensor simulation model(40 cm×12 cm×60 cm) applicable to the actual size of packaged grains is constructed through numerical simulation analysis and its forward problem solving is completed.Secondly, the performance parameters of the array electrode sensor are further optimized to comprehensively analyze the influence laws of the structural parameters such as the width and height of the sensor's pole plate, the spacing between the top and bottom of the pole plate, and the thickness of the shielding layer on the three evaluation indexes, namely, the uniformity of the sensitivity distribution in the measurement space, the dynamic measurement range of the capacitance, and the relative error of the reconstructed image.By designing orthogonal tests and comparing and analyzing the evaluation results of each structural parameter on the three indexes, a set of structural parameters of the ECT sensor with optimal performance is finally determined, i.e.,the height of the pole plate is 7 cm, the width of the pole plate 5 cm, the spacing between the top and bottom of the pole plate 6.5 cm, and the thickness of the shielding layer 6.5 cm.Finally, the image reconstruction is carried out using the Landweber algorithm to show that the reconstructed image can be basically reconstructed and the effect of reconstruction is verified.The results show that the reconstructed image can basically map out the distribution location of the high dielectric constant region(rice with high moisture content),but the image quality of the junction needs to be improved.The results of this study are expected to provide a theoretical basis for the development of moisture ECT imaging detection system for packaged grain.
基于改进YOLOv8的遥感图像目标检测算法
Remote Sensing Image Target Detection Algorithm Based on Improved YOLOv8
郑杰;宁佳绪;刘子怡; ZHENG Jie;NING Jiaxu;LIU Ziyi;针对遥感图像背景复杂、小目标数量多、目标尺度各异的特点，提出一种基于改进YOLOv8的遥感图像目标检测算法。首先，在主干网络引入深度可分离卷积(DSConv),减少模型的计算量和参数量，提升计算效率；其次，为提升对遥感图像的目标检测能力，在颈部(Neck)添加一种高效通道空间注意力模块(ECSA),通过对通道和空间特征的融合提高模型的识别能力；最后，增加检测尺度为160×160的小目标检测层，提升图像中小目标检测能力，并将边界框损失函数替换为SIoU,关注预测边界框与真实边界框的角度信息，提升检测精度。在遥感图像数据集SIMD上的实验结果表明：改进YOLOv8算法具备更强的特征提取能力，与YOLOv8算法相比，平均精度均值提升了2.63%,模型参数量减少了19.43%,模型计算量减少了28.31%,证明了改进YOLOv8算法的有效性。
Aiming at the characteristics of remote sensing image with complex background, large number of small targets and different target scales, a remote sensing image target detection algorithm based on improved YOLOv8 is proposed.Firstly, the depth separable convolution(DSConv) is introduced to construct the backbone network, which reduces the computational volume and number of parameters of the model and improves the computational efficiency.Secondly, in order to improve the detectability for remote sensing images of the target, an efficient channel spatial attention module(ECSA) proposed is added in the Neck section to improve the recognition ability of the model by fusing the channel and spatial features.Finally, a small target detection layer with a detection scale of 160×160 is added to improve the detectability of small targets in the image; the bounding box loss function is replaced with SIoU,which focuses on the angle information between the predicted bounding box and the real bounding box to improve the detection accuracy.The experimental results on the remote sensing image dataset SIMD show that The optimized YOLOv8 algorithm demonstrates significantly improved feature extraction capacity.Compared with the YOLOv8 algorithm, the average accuracy is improved by 2.63%,the model parameter quantity is reduced by 19.43%,and the model computation is reduced by 28.31%,which proves the effectiveness of the improved YOLOv8 algorithm.
渗铝涂层与高温合金在1100℃互扩散行为研究
Interdiffusion Behavior of Aluminized Coating and Superalloy at 1 100 ℃
崔道融;刘贺;车世凯;周金涛;刘晴晴;徐雪磊; CUI Daorong;LIU He;CHE Shikai;ZHOU Jintao;LIU Qingqing;XU Xuelei;Shenyang Ligong University;渗铝涂层与高温合金在高温氧化过程中会发生元素互扩散行为，为探究元素互扩散对单晶高温合金基体微观结构和相变的影响，通过高温化学气相渗铝的方法在样品表面上沉积一层铝化物涂层，并使用扫描电镜(SEM)观察沉积后涂层的微观结构，发现涂层与基体之间发生了元素互扩散行为，形成了互扩散区(IDZ)。对沉积后的样品在1 100℃条件下进行500 h的氧化处理，结果表明：IDZ随氧化时间增长逐渐变薄，最终导致拓扑密堆(TCP)相析出，并形成二次反应区(SRZ);难熔元素随氧化时间的延长不断增加，不稳定的γ′-Ni3Al相也随氧化时间的延长发生转化，最终与难熔元素一起在基体中形成TCP相，影响合金的使用寿命。
Aiming at the fact that element interdiffusion has influence on the microstructure and phase transformation of single crystal superalloy matrix in the high-temperature oxidation process.To investigate this influence, a layer of aluminide coating was deposited on the surface of the sample by high-temperature chemical vapor aluminization, and the microstructure of the deposition coating was observed by scanning electron microscopy(SEM).It was found that the element interdiffusion behavior occurred between the coating and the matrix, forming an interdiffusion zone(IDZ).The samples were oxidized at 1 100 ℃ for 500 h, and the results showed that the IDZ gradually thinned with the oxidation time, which eventually led to the precipitation of the TCP phase and the formation of the secondary reaction zone(SRZ).The refractory elements increase with the oxidation time, and the unstable γ′-Ni3Al phase also transforms with the oxidation time, and finally forms the TCP phase in the matrix together with the refractory elements, which affects the service life of the alloy.
羧基改性纤维素气凝胶对Ni(II)的去除研究
Study on the Removal of Ni(II)by Carboxyl-modified Cellulose Aerogels
陈梁心铭;张丽芳;宋颖韬;许代兵; CHEN Liangxinming;ZHANG Lifang;SONG Yingtao;XU Daibing;为提高纤维素对Ni(II)的吸附效果，通过环氧氯丙烷交联和苹果酸羧基化改性，制备具有良好吸附性能的羧基改性纤维素气凝胶。采用扫描电子显微镜、傅里叶变换红外光谱仪和X射线衍射仪系统表征材料羧基化前后的变化，结果表明，改性过程实现了纤维素表面羟基与苹果酸中羧基的有效酯化，所得气凝胶呈现典型的三维蜂窝状多孔结构。通过条件优化实验发现，当苹果酸与纤维素质量比为1.75∶1、磷酸二氢钠用量为0.5 g/g、130℃反应210 min时，制备的改性纤维素气凝胶在pH为7.0、吸附时间为120 min条件下，对Ni(II)的去除率可达92.61%。动力学及热力学分析结果表明，苹果酸改性纤维素气凝胶对Ni(II)的吸附过程可由准二级动力学模型描述，符合Langmuir等温吸附模型，且为自发进行的吸热过程。
To improve the adsorption effect of cellulose on Ni(II),carboxyl-modified cellulose aerogels with good adsorption properties were prepared by epichlorohydrin cross-linking and carboxylation modification of malate.Scanning electron microscopy, Fourier transform infrared spectroscopy and X-ray diffraction were used to characterize the changes before and after carboxylation.Through the condition optimization experiment, it was found that, when the mass ratio of malic acid to cellulose was 1.75∶1,the amount of sodium dihydrogen phosphate was 0.5 g/g, and the reaction was 130 ℃ for 210 min, the removal rate of Ni(II)in the prepared modified cellulose aerogel could reach 92.61% under the conditions of pH=7.0 and adsorption time of 120 min.The results of kinetic and thermodynamic analysis showed that the adsorption process of Ni(II)by malic acid-modified cellulose aerogel could be described by the quasi-second-order kinetic model, which was in line with the Langmuir isothermal adsorption model and was a spontaneous endothermic process.