一种低压电弧故障检测的新方法

利用二进小波变换法分解低压配电线路电弧故障时的电流波形, 获得电流波形各尺度小波变换的高频系数. 电弧故障时的电流波形小波变换高频系数与负载正常启动和运行状态时相比, 有其明显的特征, 可有效地应用于电弧故障检测. 利用LabVIEW和MATLAB软件平台实现了电流在线采集和分析, 实验结果表明该方法能准确地辨识低压配电线路中的电弧故障.     

改进YOLOX-nano的火灾火焰烟雾检测

针对在检测火焰和烟雾的火灾检测过程中存在火灾初期小目标难以检测的情况, 本文提出了一种基于自然指数损失(eCIoU)的改进YOLOX-nano (ASe-YOLOX-nano)目标检测算法. 首先, 提出一种新的目标检测函数eIoU损失函数来替代传统IoU损失, 解决在检测小目标时预测框和真实框易出现无交集的情况, 及无法反应宽高影响等问题. 其次, 在网络模型中引入注意力模块, 在网络初期模糊定位目标位置, 提高网络后期对目标尤其是小目标检测的准确性. 此外, 本文还采用软池化空间金字塔池化结构提取不同尺寸的空间特征信息, 可以提升模型对于空间布局和物体变性的鲁棒性, 因此目标较小时也可以提取充足的特征, 采用Mosaic增强技术预处理数据集, 提升模型的泛化能力, 以此进一步提高网络性能. 通过目标数据集进行对比验证, 其结果显示, mAP指标达到70.07%, 比原模型提高了3.46%, 火焰的准确率达到84.66%, 烟雾的达到74.56%, FPS能够稳定在73, 相对于传统YOLOX-nano算法拥有更好的火灾检测能力.     

基于机器视觉的螺纹钢表面缺陷检测方法

螺纹钢是一种广泛应用的建筑材料,在轧制过程中如果不能及时发现其尺寸和表面缺陷,就会生产出大量废品,给企业带来损失.本文设计了一种基于视觉的螺纹钢表面缺陷检测方法.先利用仿射变换对图像中歪斜的螺纹钢进行校正,然后基于霍夫变换检测纵肋边缘直线位置的方法对螺纹钢正面、侧面图像进行区分.最后针对正面、侧面图像分别进行缺陷检测,快速准确地判别表面是否存在缺陷.实验表明所设计的方法具有较好的稳定性和实用性,能有效地解决人工检测过程中效率低、误检率高等问题.     

基于深度学习的学生课堂注意力评价

针对学生课堂行为状态识别准确率较低的问题, 提出一种基于YOLOv4的改进模型. 通过建立学生课堂行为状态数据集, 调整YOLOv4算法训练模型的参数, 修改卷积块激活函数为ELU函数以优化模型, 同时提出将DIoU-Soft-NMS作为非极大值抑制机制, 识别分析教室中学生课堂行为状态; 根据各状态持续时长及状态变化频率计算学生听课有效时长, 并参考山东高考赋分原则, 建立学生课堂注意力量化评价准则, 同时建立教师课堂授课效果量化评价标准. 实验结果表明, 以同一评价指标衡量模型, 该模型在学生课堂行为检测速率不变的情况下, 平均精度均值(mAP)达到98.8%, 比原YOLOv4模型提升了3.53%, 学生服课堂注意力量化评价准则, 有较高的契合度.     

Fast and scalable lock methods for video coding on many-core architecture

Many-core processors are good candidates for speeding up video coding because the parallelism of these applications can be exploited more efficiently by the many-core architecture. Lock methods are important for many-core architecture to ensure correct execution of the program and communication between threads on chip. The efficiency of lock method is critical to overall performance of chipped many-core processor. In this paper, we propose two types of hardware locks for on-chip many-core architecture, a centralized lock and a distributed lock. First, we design the architectures of centralized lock and distributed lock to implement the two hardware lock methods. Then, we evaluate the performance of the two hardware locks and a software lock by quantitative evaluation micro-benchmarks on a many-core processor simulator Godson-T. The experimental results show that the locks with dedicated hardware support have higher performance than the software lock, and the distributed hardware lock is more scalable than the centralized hardware lock.     

改进RRT的移动机器人路径规划算法

针对快速扩展随树算法(RRT)在移动机器人路径规划中随机性强、转折点多、存在大量冗余点的缺点,提出了一种基于单元分解法的改进RRT算法。首先在算法初始阶段,利用单元分解法将地图划分为可行区域与障碍区域;后根据区域与区域之间邻接关系,将随机采样点的选择固定在邻居区域内,直至扩展到目标点所在区域;并对搜索到的最终路径进行优化处理,改善路径转折点过多的问题。仿真结果表明,改进RRT算法生成的路径长度更短、消耗时间更少,最后将改进RRT算法应用在实际移动机器人中,进一步证明改进算法的实用性和有效性。     

Influence of Sb-doping on dielectric properties of NaCu3Ti3TaO12 ceramics and relevant mechanism(s)

A series of NaCu₃Ti₃Ta_1−xSb_xO₁₂ ceramics were prepared by the conventional solid-state reaction technique, and their dielectric properties, crystalline structures, microstructures and complex impedance were investigated systematically. All the ceramics show the main phases of perovskite-related crystallographic structure, and their dielectric properties change significantly with the increasing Sb-doping. All these ceramics exhibit giant dielectric-permittivity properties, and impedance spectroscopy analysis reveals that NaCu₃Ti₃Ta_1−xSb_xO₁₂ ceramics are electrically heterogeneous and composed of insulating grain boundaries and semiconducting grains. Moreover, CuO secondary phase and Cu²⁺/Cu¹⁺, Ti⁴⁺/Ti³⁺, Sb⁵⁺/Sb³⁺ and Ta⁵⁺/Ta³⁺ aliovalences are found to exist in NaCu₃Ti₃Ta_1−xSb_xO₁₂ ceramics through XRD, EDS and XPS analysis. Therefore, CuO segregation and aliovalences of metal ions were suggested to contribute greatly to the internal barrier layer capacitance effect formation in NaCu₃Ti₃Ta_1−xSb_xO₁₂ ceramics. Furthermore, Sb-doping could decrease the tanδ of NaCu₃Ti₃Ta_1−xSb_xO₁₂ ceramics at low frequencies, and the reason was discussed.