| 1 | 0 | 88 |
| 下载次数 | 被引频次 | 阅读次数 |
针对人工盘煤成本高昂与激光测量方法精度受限等问题，提出基于PointCNN网络的煤场煤堆点云识别与体积计算方法。首先，利用欧式距离对毫米波雷达获取的煤堆原始点云数据进行分割；其次，采用PointCNN网络精确识别目标煤堆点云数据，并采用Delaunay三角剖分算法及投影法实现煤堆点云数据的三维曲面重建和煤堆的体积计算；最后，以某燃煤电站煤场为研究对象，对所提方法进行验证。结果表明，相较于传统测量方法，本文所提方法精度更高，相对误差低于5%，能够满足燃煤电站对煤场煤堆的体积测量要求。
Abstract：Aiming at the high cost of manual coal tray and the limited accuracy of laser measurement method, a point cloud recognition and volume calculation method of coal pile based on PointCNN network is proposed. Firstly, Euclidean distance is used to segment the original point cloud data of coal pile obtained by millimeter wave radar. Secondly, PointCNN network is used to accurately identify the point cloud data of target coal pile, and Delaunay triangulation algorithm and projection method are used to realize the three-dimensional surface reconstruction of coal pile point cloud data and volume calculation of coal pile. Finally, taking a coal yard of a coal-fired power station as the research object, the proposed method is validated. The results show that compared with traditional measurement methods, the proposed method has higher accuracy, and the relative error is less than 5%, which can meet the volume measurement demand of coal pile in coal yard of coal-fired power station.
[1] 周智宾，李伟伟，王凯，等.数字化煤场及智慧管理平台一体化建设方案设计[J].东北电力技术，2023，44（7）：58-62. ZHOU Zhibin, LI Weiwei, WANG Kai, et al. Design of Construction Scheme of Digital Coal Yard and Intelligent Management Platform Integration[J]. Northeast Electric Power Technology, 2023, 44(7): 58-62.
[2] 尹雨枫.基于RGB-D深度相机的料堆体积快速测量方法研究[D]. 长沙：湖南大学，2022.
[3] 崔峥.基于深度学习点云分割的散料堆体积测量方法的研究[D]. 济南：山东大学，2022.
[4] 孟坦.基于三维重建技术的散料堆体积测量方法研究[D].长春：长春工业大学，2023.
[5] 唐晓彬，沈童.深度学习框架发展综述[J].调研世界，2023（4）：83-88. TANG Xiaobin, SHEN Tong. Overview of the Development of Deep Learning Frameworks[J]. Research World, 2023(4): 83-88.
[6] 乐英，杨冰雁.基于PointNet++的煤场点云分割与识别方法[J].中国工程机械学报，2023，21（3）：199-203. LE Ying, YANG Bingyan. Coal Yard Point Cloud Segmentation and Recognition Method Based on PointNet ++[J]. Chinese Journal of Mechanical Engineering, 2023, 21(3): 199-203.
[7] 高远，罗皓轩，石田，等.基于改进注意力机制的燃煤电厂电煤库存校核研究[J].国外电子测量技术，2023，42（11）：106-112. GAO Yuan, LUO Haoxuan, SHI Tian, et al. Research on Coal Inventory Verification in Coal-fired Power Plants Based on Improved Attention Mechanism[J]. Foreign Electronic Measurement Technology, 2023, 42(11): 106-112.
[8] 朱天晓，闫丰亭.结合偏移自注意力机制和残差连接的点云分类[J].激光与红外，2023，53（8）：1177-1185. ZHU Tianxiao, YAN Fengting. Point Cloud Classification Combining Offset Self-Attention Mechanism and Residual Connection[J]. Laser & Infrared, 2023, 53(8): 1177-1185.
[9] 朱均超，卞永鑫，韩芳芳，等.基于改进欧式聚类算法的双目主动视觉点云目标检测研究[J].光电子·激光，2023，34（12）：1288-1297. ZHU Junchao, BIAN Yongxin, HAN Fangfang, et al. Research on Point Cloud Object Detection in Binocular Active Vision Based on Improved Euclidean Clustering Algorithm[J]. Journal of Optoelectronics·Laser, 2023, 34(12): 1288-1297.
[10] Li Y Y, Bu R, Sun M C, et al. PointCNN: convolution on X-transformed points[J]. Advances in Neural Information Processing Systems, 2018, 31: 820-830.
[11] 许安宁.基于深度学习的点云分类及分割算法研究[D].宜昌：三峡大学，2021.
[12] 孙国强，庞岩.基于PointCNN的车载行道树点云分割方法[J].北京测绘，2023，37（1）：61-66. SUN Guoqiang, PANG Yan. Point Cloud Segmentation Method for Roadside Trees Based on PointCNN[J]. Beijing Surveying and Mapping, 2023, 37(1): 61-66.
[13] 刘怡然，李丽君，杜月云.Delaunay三角剖分的汽车螺旋锥齿轮磨损检测[J].机械设计与制造，2024（1）：285-288，293. LIU Yiran, LI Lijun, DU Yueyun. Wear Detection of Automotive Spiral Bevel Gears Based on Delaunay Triangulation[J]. Mechinery Design and Manufacture, 2024(1): 285-288, 293.
[14] 方炳坤，楚瀛.基于Voronoi图与条件随机场的自然场景文本检测方法[J].计算机应用与软件，2024，41（1）：119-125. FANG Bingkun, CHU Ying. Natural Scene Text Detection Method Based on Voronoi Diagram and Conditional Random Field[J]. Computer Applications and Software, 2024, 41(1): 119-125.
[15] 张云赫，苏立晨，董云帆，等.基于Voronoi图最近邻协商的多机协同追捕方法[J].哈尔滨工程大学学报，2023，44（2）：284-291. ZHANG Yunhe, SU Lichen, DONG Yunfan, et al. Cooperative pursuit of multiple UAVs based on Voronoi partition nearest neighbor negotiation[J]. Journal of Harbin Engineering University, 2023, 44(2): 284-291.
[16] 林福根，温祥西，吴明功，等.基于Voronoi图和改进K-means的扇区优化研究[J].西北工业大学学报，2023，41（1）：170-179. LIN Fugen, WEN Xiangxi, WU Minggong, et al. Research on airspace sector optimization based on Voronoi diagram and improved K-means algorithm[J]. Journal of Northwestern Polytechnical University, 2023, 41(1): 170-179.
[17] 熊鹏文，周晓芸，熊宏锦，等.面向大规模交互数据空间划分的Voronoi图生成算法及应用[J].国防科技大学学报，2022，44（1）： 129-136. XIONG Pengwen, ZHOU Xiaoyun, XIONG Hongjin, et al. Voronoi diagram generation algorithm and application for large-scale interactive data space partition[J]. Journal of National University of Defense Technology, 2022, 44(1): 129-136.
[18] 王小双，张良奇，肖姚，等.红黑着色的相场两相流并行投影算法[J].重庆大学学报，2024，47（6）：75-85. WANG Xiaoshuang, ZHANG Liangqi, XIAO Yao, et al. A novel red-black coloring parallel projection algorithm for two-phase flow using the phase field method[J]. Journal of Chongqing University, 2024, 47(6): 75-85.
[19] 彭蹦，杨耀权，江鹏宇.改进ICP算法在无人机煤堆体积测量的应用研究[J].自动化与仪表，2019，34（7）：48-53. PENG Beng, YANG Yaoquan, JIANG Pengyu. Application Research of Improved ICP Algorithm in Coal Pile Volume Measurement by UAV[J]. Automation & Instrumentation, 2019, 34(7): 48-53.
[20] 张森，孔明，王道档，等.圆形煤场煤堆体积快速测量方法[J].中国计量大学学报，2018，29（3）：287-292. ZHANG Sen, KONG Ming, WANG Daodang, et al. A fast measurement method of coal stack volume in circular bunkers [J]. Journal of China Jiliang University, 2018, 29(3): 287-292.
基本信息:
DOI：10.19929/j.cnki.nmgdljs.2025.0053
中图分类号:
引用信息:
[1]费亦凡,张豪庆,俞更喜.基于PointCNN的煤场煤堆点云识别与体积计算[J],2025,43(4):95-100.DOI:10.19929/j.cnki.nmgdljs.2025.0053.
基金信息:
中国华能集团有限公司科技项目“5G智慧电厂料场煤堆实时三维建模”（HNKJ21-HF86）