基于计算机视觉的混凝土结构裂缝检测研究现状与展望

谢明志; 樊丁萌; 蒋志鹏; 邓飞; 王坤; 韩晨; 杨永清

doi:10.3969/j.issn.0258-2724.20240115

基于计算机视觉的混凝土结构裂缝检测研究现状与展望

doi: 10.3969/j.issn.0258-2724.20240115

谢明志^{1, 2,},
樊丁萌^{1, 2},
蒋志鹏^{1, 2},
邓飞^{1, 3},
王坤^{1, 2},
韩晨^{1, 2},
杨永清^{1, 2}

1.
西南交通大学土木工程学院，四川成都 610031
2.
西南交通大学桥梁智能与绿色建造全国重点实验室，四川成都 611756
3.
四川金通工程试验检测有限公司，四川成都 610047

基金项目: 国家自然科学基金项目（52322811）；四川省科技计划资助项目（2020YJ0081）

详细信息

作者简介:
谢明志（1985—），男，副教授，博士，硕士研究生导师，研究方向为桥梁智能检测及损伤评估，E-mail：mzxie@home.swjtu.edu.cn

中图分类号: U446；U24
计量
- 文章访问数: 87
- HTML全文浏览量: 40
- PDF下载量: 17
- 被引次数: 0
出版历程
- 收稿日期: 2024-03-08
- 修回日期: 2024-07-11
- 网络出版日期: 2025-03-29

Research Status and Prospects of Computer Vision-Based Crack Detection of Concrete Structure

XIE Mingzhi^{1, 2
,},
FAN Dingmeng^{1, 2},
JIANG Zhipeng^{1, 2},
DENG Fei^{1, 3},
WANG Kun^{1, 2},
HAN Chen^{1, 2},
YANG Yongqing^{1, 2}

1.
School of Civil Engineering, Southwest Jiaotong University, Chengdu 610031, China
2.
State Key Laboratory of Bridge Intelligent and Green Construction, Southwest Jiaotong University, Chengdu 611756, Chengdu 611756, China
3.
Sichuan Jintong Engineering Test and Inspection Co., Ltd., Chengdu 610047, China

摘要

摘要:
裂缝检测作为混凝土结构健康监测的重要内容之一，反映了结构受力及损伤状态，其检测及评估是保障结构安全服役的核心技术. 传统的检测方法时空上覆盖范围有限，受环境、高空等因素影响较大，检测效率及精度相对较低，且较依赖于主观判断，易造成漏检与误检. 基于计算机视觉的检测方法通过搭载数字成像设备进行数据采集、输入、图像处理，对混凝土表面进行自动分析和识别，具有高效、准确、客观等优点，在混凝土结构裂缝智能检测领域应用广泛. 从图像采集、图像处理、识别算法和结构评估4个方面详细阐述基于计算机视觉的混凝土裂缝检测原理、方法和应用；综合评述数字成像技术中裂缝图像采集设备及各种图像预处理方法的适用情况，并分析不同识别算法的优缺点及适用性；与此同时，总结凝练当前研究的不足，分析计算机视觉技术在设备智能化、网络轻量化等方向上的应用及研究中面临的挑战和问题，并提出相应的解决措施，从多源数据融合利用、智能设备轻型化、数字成像与裂缝映射、结构评估高效性及实时性等方面进行展望.
- 计算机视觉 /
- 裂缝识别 /
- 智能检测 /
- 混凝土裂缝 /
- 识别算法 /
- 结构评估
Abstract:
As one of the important contents of health monitoring of concrete structure, crack detection reflects the stress and damage state of the structure, and the detection and evaluation is the core technology to ensure structure safety for service. The traditional detection methods have limited coverage in time and space and are greatly affected by environmental and altitude factors, so the detection efficiency and accuracy are relatively low. Additionally, they are dependent on subjective judgment, which is easy to cause missed detection and false detection. The detection method based on computer vision is equipped with digital imaging equipment for data acquisition, input, and image processing to automatically analyze and identify the concrete surface, which has the advantages of high efficiency, accuracy, and objectivity and is widely used in the field of intelligent crack detection of concrete structures. The principle, method, and application of concrete crack detection based on computer vision were described in detail from four aspects: image acquisition, image processing, recognition algorithm, and structure evaluation. Besides, the application of crack image acquisition equipment and various image preprocessing methods in digital imaging technology was reviewed comprehensively, and the advantages, disadvantages, and applicability of different recognition algorithms were analyzed. At the same time, the shortcomings of current research were summarized, and the challenges and problems faced by the application of computer vision technology for equipment intelligence and lightweight network were analyzed. Then the corresponding solutions were proposed. Prospects are also presented from the aspects of multi-source data fusion and utilization, lightweight intelligent equipment, digital imaging and crack mapping, and high-efficiency and real-time structure evaluation.
- computer vision /
- crack identification /
- intelligent detection /
- concrete crack /
- recognition algorithm /
- structural assessment

HTML全文

现行电气化铁路采用换相（异相）单边供电^[1-2]，以减少负序对三相电网的影响^[3]，进而降低电能质量问题中的三相电压不平衡^[4]. 为此，需要在接触网的各供电分区之间增设电分相以进行隔离. 然而，电分相^[5]的设置导致了无电区，给铁路运行带来诸多问题^[6-10]. 同相供电可以同时解决牵引变电所的负序问题及其出口处的电分相问题. 在文献[11]提出同相供电方案之后，研究者们基于无功型对称补偿技术^[12-17]进一步开发了牵引变电所有功型同相供电方法^[18-19]. 该方法能以最小补偿容量使三相电压不平衡度达标^[20]，同时消除变电所出口处的电分相. 为进一步消除分区所处和全线的电分相和无电区，不仅需要实现两个牵引变电所间的双边供电和更大范围的贯通供电^[18,21-22]，还需应对铁路贯通供电可能对电网带来的新问题.

1. 贯通供电

贯通供电是两个及以上牵引变电所同相供电和分区所双边供电的结合，能够确保现代交直交列车的牵引和（电）制动都需要的牵引网不间断供电. 最简单的贯通供电是采用单相牵引变压器的牵引变电所同相供电和分区所双边供电的结合，如图1所示. 图中，分区所纵向断路器K需要合闸才能实现贯通供电，否则，仅为牵引变电所的同相供电，属于单边供电. 电网中，将分区所断路器K合闸的操作称为合环，合环需满足电网规定的条件，否则无法进行贯通供电.

图 1 同相单边供电和贯通供电空载示意

Figure 1. Co-phase single-end power supply and interconnected power supply under no-load condition

下载: 全尺寸图片幻灯片

虽然俄罗斯、乌克兰等前苏联国家^[2]及韩国的电气化铁路一直采用双边供电，但其牵引变电所仍然采用异相供电，即换相接入电网，并不属于本文定义的贯通供电.

图1（a）所示的同相单边供电牵引网通过单个牵引变电所与电网的单个公共连接点连接后向列车供电. 而图1（b）所示的贯通供电牵引网则通过多个牵引变电所与电网的多个公共连接点连接后向列车供电，此时，牵引网成为电网中的连接支路，牵引供电系统和电网的拓扑结构发生变化. 以牵引网空载情况为例，双边供电的牵引变电所和牵引网串联后在电网不同公共连接点之间又增生了一条新的支路，形成新的电流分量，称为均衡电流或环路电流，或者说增生了一个新的功率分量，称为穿越功率. 与贯通供电相比，由于同相单边供电分区所（SP）的纵向断路器K分闸，牵引网不贯通，空载工况下不会产生穿越功率.

图2为同相单边和贯通供电的列车牵引用电和再生发电工况示意. 从图中可见，同相单边供电时，不同牵引变电所的列车牵引用电功率和列车再生发电功率在贯通供电时得以叠加，虽然列车的用电和发电总量不变，但贯通供电后，从电网用电或向电网发电的最大功率会减小，对应的负序最大值也会减小. 从这个意义上说，贯通供电优于单边供电.

图 2 同相单边和贯通供电运行工况示意

Figure 2. Operation under co-phase single-end power supply and interconnected power supply

下载: 全尺寸图片幻灯片

但与同相单边供电相比，贯通供电给电网带来新的问题，即能否合环以及合环后的穿越功率问题. 本文讨论电网的正常状态，关于故障状态和对应的继电保护技术，可参见文献[18]、[23-24].

2. 贯通供电合环

以图1（b）中选择两个牵引变电所构成的双边供电合环情形加以讨论，如图3所示. 假设实施双边供电的两个变电所SS₁、SS₂的电源分别来自有（等效）电气连接的电网公共连接点PCC₁、PCC₂，当牵引变电所SS₁、SS₂之间分区所的纵向断路器K合闸，就构成双边供电.

图 3 双边供电等效结构图

Figure 3. Equivalent structure of bilateral power supply

下载: 全尺寸图片幻灯片

实现双边供电必须在同一电网内进行. 电气化铁路的双边供电与电网的电磁合环有所不同：电网电磁合环为三相合环，而双边供电是单相合环；电网电磁合环的开关是常开的，而双边供电的纵向开关是常闭的. 尽管如此，由于双边供电和电网的电磁合环都将改变电网的拓扑结构和运行方式，因此，必须满足电网规定的合环条件.

《青海省电力系统调度规程》^[25]规定了电网的解合环操作，其中合环操作：“必须相位相同，电压差一般允许在20%之内，相角差一般不超过20°，以确保合环时不因环路电流过大引起潮流的变化而超过继电保护、系统稳定和设备容量等方面的限额. 合环后应当及时通知有关调度及厂站. 对于较复杂环网的操作，应当先进行模拟计算. ”

《中国南方电网电力调度管理规程》^[26]解合环操作规定：1）必须确保相序相位正确才能合环；2）解、合环操作必须确保解、合环后潮流不超过稳定极限、设备不过负荷、电压在正常范围内，不引起继电保护和安全自动装置误动；3）具备条件时，合环操作应使用同期装置；4） 500 kV合环时电压差一般不应超过额定电压10%，220 kV不应超过额定电压20%；500 kV系统合环一般应检同期合环，有困难时应启用合环开关的同期装置检查相角差；500 kV合环时相角差一般不应超过20°，220 kV一般不应超过25°.

图3中PCC₁和PCC₂之间的电压差和相角差可以于K合闸之前在分区所测得，也可以在已知PCC₁和PCC₂之间的电气连接关系时通过潮流计算得到^[3].

实际电网中，图3所示PCC₁和PCC₂之间的电气连接方式多种多样，作为举例，这里考虑220 kV输电线路对铁路两个牵引变电所的专用线供电方式，则三相等值电路可由图4描述. 图中，Z_A、Z_B和Z_C为输电线的三相阻抗，L为牵引变电所之间的距离.

图 4 三相等值电路示意

Figure 4. Three-phase equivalent circuit

下载: 全尺寸图片幻灯片

首先，分析分区所纵向断路器K合闸之前PCC₁和PCC₂之间的电压差和相角差是否满足合环规程的规定.

当末端负荷电流 $\dot I$ 通过电阻为R、电抗为X，即阻抗为 $Z = R + {\text{j}}X$ 的线路时，首端电压 ${\dot U_{\text{f}}}$ 与末端电压 ${\dot U_{\text{e}}}$ 的电压降 $\Delta \dot U$ 、电压差 $\Delta U$ 和相位差θ的相量图如所示^[1,15]. $\Delta U$ 和θ分别为^[1,27]

图 5 电压差与相角差示意

Figure 5. Voltage difference and phase angle difference

下载: 全尺寸图片幻灯片

$\begin{split} &\Delta U = \left| {\dot U}_{\mathrm{f}} \right| - \left| {\dot U}_{\mathrm{e}} \right|= \\ & \quad \sqrt {{{\left( {U_{\mathrm{e}} + XI\sin \;\varphi + RI\cos\; \varphi } \right)}^2} + {{\left( {XI\cos \;\varphi - RI\sin \;\varphi } \right)}^2}} - \\& \quad U_{\mathrm{e}} \approx XI\sin \;\varphi + RI\cos\; \varphi, \end{split}$

(1)

$\theta = \arctan \frac{{XI\cos \varphi - RI\sin \varphi }}{{XI\sin \varphi + RI\cos \varphi }},$

(2)

式中： $\varphi$ 为电流 $\dot I$ 的功率因数角，以滞后为正.

220 kV输电线路最大输送功率为500 MV•A^[3]. 计算中，输电线采用2 × LGJQ-500二分裂导线，取单位长阻抗z₀=0.0312 + j0.3052 Ω/km，PCC₂处的输送功率分别取200、300、500 MV•A. 考虑相邻牵引变电所的间距L与输电线匹配，L分别取50、60、70、80 km，输送功率的功率因数分别取0.95和0.90，ΔU和θ的计算结果分别列于表1和表2. 其中，ΔU的百分值以额定电压220 kV为基准.

表 1 功率因数为0.95（滞后）时的电压差ΔU和相角差θ

Table 1. Voltage difference (ΔU) and phase angle difference (θ) at a power factor of 0.95 (lagging)

L/km	传输功率 200 MV•A			传输功率 300 MV•A			传输功率 500 MV•A
L/km	ΔU/kV	ΔU/%	θ/（^o）	ΔU/kV	ΔU/%	θ/（^o）	ΔU/kV	ΔU/%	θ/（^o）
50	3.49	2.74	3.23	5.38	4.23	4.78	9.44	7.43	7.74
60	4.23	3.33	3.85	6.56	5.16	5.69	11.60	9.13	9.16
70	4.99	3.93	4.47	7.77	6.12	6.58	13.85	10.90	10.53
80	5.77	4.54	5.08	9.02	7.10	7.45	16.17	12.73	11.85

下载: 导出CSV

| 显示表格

表 2 功率因数为0.90（滞后）时的电压差ΔU和相角差θ

Table 2. Voltage difference (ΔU) and phase angle difference (θ) at a power factor of 0.90 (lagging)

L/km	传输功率 200 MV•A			传输功率 300 MV•A			传输功率 500 MV•A
L/km	ΔU/kV	ΔU/%	θ/（^o）	ΔU/kV	ΔU/%	θ/（^o）	ΔU/kV	ΔU/%	θ/（^o）
50	4.41	3.47	2.99	6.74	5.30	4.41	11.63	9.16	7.10
60	5.33	4.20	3.56	8.17	6.44	5.23	14.19	11.17	8.37
70	6.26	4.93	4.13	9.64	7.59	6.04	16.81	13.24	9.60
80	7.21	5.68	4.68	11.13	8.76	6.84	19.50	15.35	10.78

下载: 导出CSV

| 显示表格

从图5和表1、2可见，输送（视在）功率的功率因数越高，电压差越小，但相角差越大. 总的看：在牵引变电所间距不超过80 km下（通常如此），双边供电时分区所的合环电压差不大于16.00%，相角差不大于12.00°，低于合环规程的规定值，符合要求.

实践中，双边供电除了仿真计算之外，在合环前应选择牵引网处于空载状态，并检测分区所两侧接触网的电压差和相角差，在允许的电压差和相角差范围内，才能令纵向断路器K合闸.

双边供电一旦合环就可长期运行.

3. 穿越功率分析与计算

牵引供电系统的正常状态可分为牵引、再生和空载3种工况. 其中，空载工况下，双边供电与单边供电才有本质区别，双边供电的牵引供电系统与电网（等效）输电线并联，造成均衡电流或者穿越功率. 从图6可见，空载工况下，双边供电的穿越功率在一个变电所（如SS₁）表现为用电状态，在另一个变电所（如SS₂）则表现为发电状态，更多牵引变电所贯通供电时，穿越功率可能穿越得更远.

图 6 空载工况取流示意

Figure 6. Current distribution under no-load condition

下载: 全尺寸图片幻灯片

3.1 等效电路与均衡电流计算

对应图1和图4，将归算到电网侧的三相等效电路示于图7. 图中，Z_d为三相输电线的相阻抗，即Z_A=Z_B=Z_C=Z_d； ${Z_{{\text{J1}}}}$ 和 ${Z_{{\text{J}}2}}$ 为牵引变电所SS₁和SS₂的进线相阻抗； $Z_{{\text{T1,e}}}$ 和 $Z_{{\text{T}}2,{\mathrm{e}}}$ 为归算到电网侧的单相牵引变压器阻抗； $Z_{\text{q,e}}$ 为归算到电网侧的牵引网阻抗；LC为电力机车.

图 7 双边供电的三相等效电路

Figure 7. Three-phase equivalent circuit for bilateral power supply

下载: 全尺寸图片幻灯片

考虑牵引网空载工况，选择A相或B相可得计算均衡电流的单相等效电路，如所示. 图中，用 ${Z_{\text{d}}} 、{Z_{\text{q}}}$ 分别表示电网输电线与牵引供电系统（包括进线相阻抗）的归算阻抗. 设总电流为 $\dot I$ ，通过电网输电线的电流为 ${\dot I_{\text{d}}}$ ，通过牵引供电系统的电流为 ${\dot I_{\text{q}}}$ ，则有支路电压方程如式（3）所示.

图 8 均衡电流示意

Figure 8. Equalizing current

下载: 全尺寸图片幻灯片

${\dot I_{\text{q}}}{{\textit{Z}}_{\text{q}}} = {{\textit{Z}}_{\text{d}}}{\dot I_{\text{d}}}.$

(3)

设Z_J1=Z_J2=Z_J，则求得均衡电流I_q为

${\dot I_{\text{q}}} = \frac{{{{\textit{Z}}_{\text{d}}}}}{{{{\textit{Z}}_{\text{q}}}}}{\dot I_{\text{d}}} = \frac{{{{\textit{Z}}_{\text{d}}}}}{{2{{\textit{Z}}_{\text{J}}} + ({{\textit{Z}}_{\text{T}}} + \dfrac{1}{2}{{Z}_{\text{q,e}}})k_{\text{T}}^2}}{\dot I_{\text{d}}},$

(4)

式中：k_T为牵引变压器变比，即电网进线的线电压与牵引母线额定电压的比值.

定义分流比为η，即均衡电流与电网输电线中的电流之比，则由式（4）得

$\eta = \left| {\frac{{{{\dot I}_{\text{q}}}}}{{{{\dot I}_{\text{d}}}}}} \right| = \left| {\frac{{{{\textit{Z}}_{\text{d}}}}}{{{{\textit{Z}}_{\text{q}}}}}} \right| = \left| {\frac{{{{\textit{Z}}_{\text{d}}}}}{{2{{\textit{Z}}_{\text{J}}} + ({{\textit{Z}}_{\text{T}}} + \dfrac{1}{2}{{Z}_{\text{q,e}}})k_{\text{T}}^2}}} \right|.$

(5)

精确计算需要掌握具体数据. 假设电网PCC₁和PCC₂之间220 kV输电线的长度为50 km，PCC₁（PCC₂）到牵引变电所SS₁（SS₂）的进线长度为10 km，输电线单位长阻抗同前，牵引变压器均为单相变压器，额定容量均为31.5 MV•A，短路电压百分比为10.5%，则求得归算到牵引侧的漏抗Z_T= 0.2134 + j2.52 Ω，相邻牵引变电所间距50 km，直供方式牵引网空载电压为27.5 kV，变比k_T=8，单线铁路、单链型悬挂牵引网单位长阻抗z=0.232 + j0.515 Ω/km，代入式（5）可得η=1.44%. 工程上，可以用

$\eta \approx \frac{1}{{k_{\text{T}}^2}},$

(6)

进行估算，计算得 $\eta$ =1/64≈1.56%，接近且略大于精确计算值.

3.2 穿越功率计算与测量

设电网输电线的额定电压为U_N，由式（3）得到穿越功率S_q为

${S _{\text{q}}} = {U_{\text{N}}}{I_{\text{q}}} = \frac{{{{\textit{Z}}_{\text{d}}}}}{{{{\textit{Z}}_{\text{q}}}}}{U_{\text{N}}}{I_{\text{d}}},$

(7)

或者已知电网输电线三相功率S时，由式（8）估算得牵引网（单相）穿越功率.

${S _{\text{q}}} \approx \frac{S}{{\sqrt 3 k_{\text{T}}^{\text{2}}}}.$

(8)

一般110 kV输电线最大输送功率为50 MV•A，则直供方式时，变比k_T=4，由式（8）估算得双边供电最大穿越功率为1.8 MV•A，占比3.6%；220 kV输电线最大输送功率为500 MV•A，则直供方式时，k_T=8，由式（8）估算得双边供电穿最大越功率为4.6 MV•A，占比1.8%.

上述定义和讨论的均衡电流和穿越功率均沿线路传输，故称为纵向分量. 穿越功率可进一步分为有功分量和无功分量（一般为感性），都属于纵向分量. 因此，如果穿越功率存在，则在牵引变电所进线、牵引馈线以及牵引网的任何方便的部位都能测量得到. 还有一个像负荷一样的横向分量，这就是线路的容性无功功率，也叫充电功率. 穿越功率取决于电网对牵引供电系统的供电方式和电网潮流，充电功率则沿输电线和牵引网分布，取决于其长度和电压等级. 线路充电功率在分区所纵向断路器合闸前后的总量不变，但合闸后将在系统中重新分配，此时也存在一个分流点. 当分流点与分区所不重合时，分区所会测到这个分量不是0，而当分流点与分区所重合时，分区所的这个分量就是0.

要判断分区所合环后是否存在穿越功率，则应在牵引网空载时测量牵引网的有功分量，才能准确反映穿越功率情况：若检测到的有功分量为0，则穿越功率中的（纵向）无功分量亦为0，穿越功率即为0；而此时若检测到的无功分量不为0，只能说明存在充电功率（横向分量），不是穿越功率. 下面会看到，当电网对铁路牵引变电所进行树形供电时，理论上穿越功率为0.

4. 穿越功率解决方案

正常运行状态下，双边供电与单边供电在潮流分布上的唯一区别就是穿越功率问题. 由上分析可知，穿越功率占比很小，不会对电网潮流造成大的影响. 但是，为不产生穿越功率或不向电网发电（分量），仅达到与单边供电相同或更好的效果，必须关注并解决穿越功率的问题.

1）树形供电

电网的变电站用同一（分段）母线公共连接点（PCC）为多个铁路变电所供电，这种供电方式称为树形供电^[19,22,28]，如图9所示. 此时，对应图3，可认为PCC₁与PCC₂合并，电网输电线长度为0，即令式（6）中Z_d=0，得η=0. 因此，铁路实施树形双边供电不会产生均衡电流和穿越功率，而负序等电能质量影响与铁路单边供电的处理方式相同.

图 9 树形供电方式示意

Figure 9. Tree-structured power supply

下载: 全尺寸图片幻灯片

我国西北地区地广人稀，电网对铁路牵引变电所多进行树形供电，为开展和实施树形双边供电提供了得天独厚的条件. 国铁集团于2020年在科技研究开发计划中立项开展“电气化铁路牵引变电所双边供电试验”研究，通过综合比选，考虑电网电源条件，试验段选取在海拔约3000 m、线路坡度16‰的格库铁路（青海段）东柴山至花土沟车站区段，长度约75 km. 格库铁路为国铁Ⅰ级客货共线的单线双向电气化铁路，设计速度120 km/h，使用HXD1C系列电力机车. 2021年7月试验取得圆满成功，仿真和试验证明在树形供电方式下，双边供电不产生穿越功率，具备投运条件.

交通运输部国家“交通强国”试点中的“巴准铁路贯通式同相供电工程化”项目也是在电网树形供电下的铁路贯通供电应用实例^[26]. 该铁路位于内蒙古鄂尔多斯市，为国铁Ⅰ级复线电气化重载运煤铁路，东起点岱沟站，向西经海勒斯壕南站与包神铁路在巴图塔站接轨，全长128 km. 改造前设有4座牵引变电所，贯通供电改造后由电网的川掌变电站树形供电至保留的四道柳和纳林川两个牵引变电所，牵引变电所均采用组合式同相供电方案^[11]：主变更换为不等边Scott接线，其M座额定容量为40 MV•A，T座同相供电装置采用5 MV•A运行 + 5 MV•A备用方式，2014年6月25日8时58分四道柳和纳林川变电所之间的分区所合环成功，2024年9月实现全线贯通，这是世界首条取消全部电分相和无电区的贯通供电线路.

2）合建所

将牵引变电所与电力（动照）变电所合建在一起，牵引变压器和电力（动照）变压器接入同一进线母线，如图10的SS₂所示. 假设穿越功率由变电所SS₁流向SS₂，只要电力变压器的电力负荷大于穿越功率，则穿越功率就被“淹没”，合建所总体对电网表现为用电状态，而不会向电网发电. 由于穿越功率通常较小，只需电力负荷大于该值，即可解决穿越功率向电网发电的问题.

图 10 牵引电力合建所示意

Figure 10. Co-built power and traction substation

下载: 全尺寸图片幻灯片

3）穿越功率利用

首先识别穿越功率，例如通过光纤实时传递双边供电两端牵引馈线的有功功率，设流向牵引网为正，当该功率非0且两者之和为0时，则判定该功率为穿越功率，再在穿越功率末端的牵引变电所牵引母线上将穿越功率通过交/直/交（AC/DC/AC）装置将穿越功率转移到三相电力用电母线上，供电力负荷利用，或者通过交/直（AC/DC）装置与储能装置（ES）连接进行储能利用，为了更好利用和控制储能装置容量，一般串接直/直（DC/DC）变换装置，如图11所示. 图中，TT为牵引变压器，PTa为电压互感器，CTa为电流互感器.

图 11 穿越功率利用示意

Figure 11. Through power utilization

下载: 全尺寸图片幻灯片

多数情况下，电网对铁路的供电方式是非树形的，这时就有穿越功率产生^[12-14]，例如，广州地铁18号和22号线采用同相供电，在陇枕变电所和陈头岗变电所进行了贯通供电试验，试验测试显示，穿越功率最大为2.2 MV•A.

5. 多功能的智能牵引变电所

与单边供电相比，贯通供电仍会产生负序功率和再生发电功率，但其最大值通常较小. 此外，贯通供电可能还会产生穿越功率. 由于再生发电功率和穿越功率的发电分量可以在牵引变电所进行统一处理，因此应提出一种多功能的牵引变电所供电方案，如图12所示. 图中：TSB为牵引母线；部分功率直流变换器（TPC）和三端口变换器（TPC）均为单相变流器，两者串联构成同相供电装置；DPC为三相交直变流器构成的能量转换装置，与TPC配合向铁路三相用电系统送电；ES为储能装置；RES为新能源发电装置；HPF为高通滤波器；PT为电压互感器；CT1和CT2为电流互感器；A、B、C为牵引变电所进行三相母线；a、b、c则为铁路三相用电系统的母线.

图 12 智能牵引变电所方案

Figure 12. Scheme of intelligent traction substation

下载: 全尺寸图片幻灯片

该方案借助单相组合式同相供电方案，仍然以Scott牵引变压器的M座牵引绕组为供电主通道，通过共用直流母线，增加新的潮流调节通道，最大程度复用变流器，减低变流器容量，增加性价比，使全寿命周期成本降至最低. 由于这种多功能供电方案可以根据供电需求进行灵活配置，同时实现牵引供电系统状态数据离线和在线智慧预测与监测，故亦可称为智能牵引变电所.

智能牵引变电所按照需要可以配置以下几个部分：1） M座牵引绕组构成单相牵引变压器，为供电主通道，简单、经济、可靠；2）同相供电装置，由TPC、PPC串联构成，按需配置，使负序达标；3）发电功率及穿越功率发电分量利用通道由TPC和DPC或ES构成，联通铁路三相用电或融入储能系统，能源高效融合利用，按需配置；4）新能源发电通道，按需配置，接入地点可在变电所或车站，实现铁路沿线新能源消纳；5）高通滤波通道，按需配置，消除高次谐波和谐振.

一般同相供电装置的容量仅占M座牵引绕组（主变）容量1/4甚至更小，穿越和再生功率利用通道容量也在5 MV•A及以下，可有效减少设备重复与投资.

智能牵引变电所具有故障导向安全特性，即使图12中的②～⑤通道退出运行，也能够保持牵引供电.

此外，借助单相组合式同相供电方案，还可以构造共用交流牵引母线的智能牵引变电所方案.

智能牵引变电所具有以下多种功能：1）可以补偿负序，使三相电压不平衡达标，不对电网造成影响；2）可以补偿无功功率，提高功率因数，提高和稳定网压；3）可以消除高次谐波和谐振；4）可以调控系统潮流，将剩余再生发电功率和穿越功率发电分量转变为用电功率，不向电网发电，不影响电网发电潮流调度；5）可以促进铁路沿线新能源消纳等.

以上功能可以根据工程实际需要进行灵活组合.

6. 结　论

1）在一般牵引变电所间距（不超过80 km）下，双边供电时分区所的合环电压差不大于16.00%，相角差不大于12.00°，符合电网合环规定.

2）与单边供电相比，双边供电和贯通供电可能会产生穿越功率. 该穿越功率可以通过计算和测量获得. 以110 kV输电线为例，其最大输送功率为50 MV•A，采用直供方式时，牵引网的最大穿越功率为1.8 MV•A，占电网输送功率的3.6%；而220 kV输电线的最大输送功率为500 MV•A，直供方式下牵引网的最大穿越功率为4.6 MV•A，占比为1.8%.

3）消除穿越功率向电网发电的影响，可以采取电网树形供电、合建变电所和穿越功率利用等方案加以解决.

4）可以进一步发展多功能的智能牵引变电所，根据实际需求进行灵活配置，综合解决负序、再生发电和穿越功率等对电网的影响问题，从而实现电气化铁路的更大范围贯通供电，推动电网和铁路的双赢发展.

致谢：黄小红、王辉、张戬、王帅博士参加了相关研究工作，在此表示感谢.

图 1 桥上PRSS无人机示意^[23]

Figure 1. Schematic diagrams of PRSS unmanned aerial vehicle (UAV) on the bridge ^[23]

下载: 全尺寸图片幻灯片

图 2 智能检测机器人

Figure 2. Intelligent detection robot

下载: 全尺寸图片幻灯片

图 3 水下机器人

Figure 3. Underwater robot

下载: 全尺寸图片幻灯片

图 4 环形爬壁机器人系统原理图^[35]

Figure 4. Schematic diagram of circular wall-climbing robot system ^[35]

下载: 全尺寸图片幻灯片

图 5 不同类型结构表面的裂缝图像采集^[36]

Figure 5. Crack image acquisition of different types of structural surfaces^[36]

下载: 全尺寸图片幻灯片

图 6 基于云点的语义分割^[43]

Figure 6. Semantic segmentation based on point cloud ^[43]

下载: 全尺寸图片幻灯片

图 7 图像裁剪与拼接^[59]

Figure 7. Image cropping and stitching ^[59]

下载: 全尺寸图片幻灯片

图 8 几种基于深度学习的算法

Figure 8. Several algorithms based on deep learning

下载: 全尺寸图片幻灯片

图 9 Skele-Marker方法的图像处理步骤^[76]

Figure 9. Image processing steps of Skele-Marker method ^[76]

下载: 全尺寸图片幻灯片

图 10 骨架提取产生的毛刺^[92]

Figure 10. Burrs from skeleton extraction ^[92]

下载: 全尺寸图片幻灯片

图 11 4种裂缝宽度测量方法

Figure 11. Four measurement methods for crack width

下载: 全尺寸图片幻灯片

图 12 2种裂缝宽度测量方法

Figure 12. Two measurement methods for crack width

下载: 全尺寸图片幻灯片

图 13 Benchmarks前10名中基于Transformer的数量

Figure 13. Number of Transformer-based algorithms in the top 10 benchmarks

下载: 全尺寸图片幻灯片

图 14 SA-1B与常见的语义分割数据集对比^[98]

Figure 14. Comparison of SA-1B with common semantic segmentation datasets ^[98]

下载: 全尺寸图片幻灯片

图 15 基于应力-位移曲线定义的损伤指标^[109]

Figure 15. Damage index based on stress-displacement curve ^[109]

下载: 全尺寸图片幻灯片

表 1 图像降噪算法分类

Table 1. Classification of image noise reduction algorithms

算法名称	算法描述	算法优缺点	适用情况
均值滤波^[53]	线性滤波器，利用相邻模板像素的均值代替中心像素值	均值滤波简单、计算方便，但缺点是不能很好地去除噪声点，同时容易破坏图像细节导致图像模糊	适用于轻度噪声的平滑处理，能较好抑制高斯噪声
中值滤波^[54-55]	非线性滤波器，利用相邻模板像素的中值代替中心像素值，减少图像模糊程度；在此基础上提出了自适应中值滤波器等	中值滤波实现简单，计算效率高，但对于较小的结构或细节，可能会造成模糊效果	适用于处理椒盐噪声等离散噪声，但不适合处理大尺寸裂缝图像
高斯滤波^[51]	线性滤波器，对邻域像素进行加权平均，在平滑噪声的同时保持图像原本的灰度分布结构特征. 在此基础上进一步发展出双边滤波器等	高斯滤波能较好地消除高斯噪声，但可能会导致图像边缘模糊，同时对大尺寸图像来说计算成本过大	适用于平滑图像并保持图像边缘的情况，能很好抑制高斯噪声
形态学滤波^[52]	从传统的形态学开闭运算的角度出发，设计了一种多角度多结构元素加权组合的基于形态学原理的滤波方法，如开闭运算、顶帽变换等	形态学滤波能够有效地去除图像中的小尺度细节，同时保持图像的主要边缘和结构特征不变，但对较大的噪声去除效果不明显.	适用于去除图像中的斑点噪声或孤立噪声
双边滤波^[56]	通过考虑距离因素和像素值差异的影响，使其在去噪的同时，能够很好地保留图像的特征信息	双边滤波器能够在平滑图像的同时有效地保留图像的边缘信息. 但对于一些细小的纹理和细节，可能会导致失真	适用于平滑图像时保持图像的边缘信息和细节的情况，能很好抑制高斯噪声及随机噪声

下载: 导出CSV

表 2 部分开源的混凝土裂缝数据库

Table 2. Part of open-source concrete crack databases

数据集名称	图像大小/像素	图像数量/张	图像来源	标注类型	下载地址
크랙 Dataset^[64]	1280 × 720	713	混凝土建筑	目标检测	https://universe.roboflow.com/hyegeun/-rxh4q
Concrete crack images for classification^[65]	227 × 227	40000	混凝土建筑	图像分类	https://data.mendeley.com/datasets/5y9wdsg2zt/2
Dataset^[66]	256 × 256	919	混凝土隧道	语义分割	https://pan.baidu.com/s/10W01KQqMS8FFAoRpKl4-Qw密码：e9d2
Bridge crack library 2.0^[45]	256 × 256	23400	混凝土桥梁（GAN 生成）	语义分割	https://doi.org/10.7910/DVN/TUFAJT
Crack dataset^[67]	224 × 224	800	混凝土建筑	语义分割	https://drive.google.com/open?id=1cplcUBmgHfD82YQTWnn1dssK2Z_xRpjx
Crack forest-dataset^[68]	480 × 320	118	混凝土道路	语义分割	https://github.com/cuilimeng/CrackForest-dataset
SDNET2018^[69]	256 × 256	56000	混凝土桥梁、建筑、路面	图像分类	https://digitalcommons.usu.edu/all_datasets/48/
Crack-detection^[70]	224 × 224	6069	混凝土桥梁	图像分类	https://github.com/tjdxxhy/crack-detection

下载: 导出CSV

表 3 不同采集设备对检测精度的影响

Table 3. Influence of different equipment on detection accuracy

图像采集方式	识别精度/mm	应用场景
民用无人机	0.50	价格便宜，小巧灵活，可以高效地完成大面积的裂缝检测，适用各种场景
行业无人机	0.10	价格昂贵，体型较大，适合桥梁、大坝等空旷场景
水下机器人	1.00	适用于水库大坝、海底隧道、桥墩水下部分等水下混凝土结构
智能检测车	0.10	载荷较大，可搭载多个摄像头，主要用于桥梁、隧道等场景
微距摄像头	0.01	测量精度高，但要求近距离拍摄，且检测范围小，适合实验室使用

下载: 导出CSV

参考文献(112)

[1]	ZHAO S Z, KANG F, LI J J. Intelligent segmentation method for blurred cracks and 3D mapping of width nephograms in concrete dams using UAV photogrammetry[J]. Automation in Construction, 2024, 157: 105145.1-105145.19.
[2]	ZHOU Z, ZHANG J J, GONG C J. Hybrid semantic segmentation for tunnel lining cracks based on Swin Transformer and convolutional neural network[J]. Computer-Aided Civil and Infrastructure Engineering, 2023, 38(17): 2491-2510. doi: 10.1111/mice.13003
[3]	LIANG K, SU R K L. Seismic behavior of non-seismically designed shear walls with moderate shear span under high axial loads[J]. Engineering Structures, 2024, 300: 117223.1-117223.16.
[4]	李晓华. 制造业的数实融合:表现、机制与对策[J]. 改革与战略,2022,38(5): 42-54.
[5]	SINGH M P, SEN S, PATHAK H, et al. Early age cracking relevant to mass concrete dam structures during the construction schedule[J]. Construction and Building Materials, 2024, 411: 134739.1-134739.15.
[6]	LIU Y T, ZHENG D J, CAO E H, et al. Cracking risk analysis of face slabs in concrete face rockfill dams during the operation period[J]. Structures, 2022, 40: 621-632. doi: 10.1016/j.istruc.2022.04.054
[7]	WANG Y J, WU Z M, QU F M, et al. Numerical investigation on crack propagation process of concrete gravity dams under static and dynamic loads with in-crack reservoir pressure[J]. Theoretical and Applied Fracture Mechanics, 2022, 117: 103221.1-103221.17.
[8]	SHENG X W, XIAO S M, ZHENG W Q, et al. Experimental and finite element investigations on hydration heat and early cracks in massive concrete piers[J]. Case Studies in Construction Materials, 2023, 18: e01926.1-e01926.14.
[9]	WANG H M, RANADE R, OKUMUS P. Seismic fragility of reinforced concrete bridge columns utilizing ductile fiber-reinforced concrete covers[J]. Structure and Infrastructure Engineering, 2023, 19(5): 708-730. doi: 10.1080/15732479.2021.1973040
[10]	VISHWANATH B S, BANERJEE S. Considering uncertainty in corrosion process to estimate life-cycle seismic vulnerability and risk of aging bridge piers[J]. Reliability Engineering & System Safety, 2023, 232: 109050.1-109050.15.
[11]	DING Y, ZHOU Z, WEI Y, et al. Experimental and numerical study on the seismic performance of barbell-shaped shear walls with partial application of UHPC[J]. Soil Dynamics and Earthquake Engineering, 2024, 179: 108525.1-108525.13.
[12]	HU H, LAN Y D, LIU Y, et al. Experimental study on the damage performance and forced response of concrete lining in fault-crossing tunnel[J]. Scientific Reports, 2024, 14: 2120.1-2120.23. doi: 10.1038/s41598-024-72324-z
[13]	DING Z, HUANG X, SUN M M, et al. Experimental study on the influence of cracks on tunnel vibration under subway train load[J]. Tunnelling and Underground Space Technology, 2023, 142: 105444.1-105444.18.
[14]	VY V, LEE Y, BAK J, et al. Damage localization using acoustic emission sensors via convolutional neural network and continuous wavelet transform[J]. Mechanical Systems and Signal Processing, 2023, 204: 110831.1-110831.19.
[15]	CHEN L, LU X R, DENG D P, et al. Optimized wavelet and wavelet packet transform techniques for assessing crack behavior in curved segments of arched beam bridges spanning rivers[J]. Water, 2023, 15(22): 3977.1-3977.23.
[16]	蒋宇静,吴法博,张学朋,等. 基于激光多普勒测振的混凝土损伤量化评价研究[J]. 中国矿业大学学报,2023,52(5): 889-903. JIANG Yujing, WU Fabo, ZHANG Xuepeng, et al. Non-destructive quantitative assessment on the damage of concrete structure based on laser Doppler vibrometer technology[J]. Journal of China University of Mining & Technology, 2023, 52(5): 889-903.
[17]	陈江,熊杰,李宇驰,等. 基于热效应的涉水混凝土工程裂缝监测方法[J]. 工程科学与技术,2023,55(1): 296-303. CHEN Jiang, XIONG Jie, LI Yuchi, et al. Crack monitoring method for wading concrete engineering based on thermal effects[J]. Advanced Engineering Sciences, 2023, 55(1): 296-303.
[18]	SUN W, DU J, QIAO P Z, et al. Smart piezoelectric module-based wave energy measurement for crack damage evaluation of concrete beams[J]. Measurement, 2023, 221: 113463.1-113463.16.
[19]	钟新谷,彭雄,沈明燕. 基于无人飞机成像的桥梁裂缝宽度识别可行性研究[J]. 土木工程学报,2019,52(4): 52-61. ZHONG Xingu, PENG Xiong, SHEN Mingyan. Study on the feasibility of identifying concrete crack width with images acquired by unmanned aerial vehicles[J]. China Civil Engineering Journal, 2019, 52(4): 52-61.
[20]	CHOU K W, HE L G, HSU K, et al. Application of an unmanned aerial vehicle for crack measurement using image calibration supported by laser projectors[J]. Multiscale Science and Engineering, 2021, 3(3): 225-235.
[21]	SANCHEZ-CUEVAS P J, RAMON-SORIA P, ARRUE B, et al. Robotic system for inspection by contact of bridge beams using UAVs[J]. Sensors, 2019, 19(2): E305.1-E305.17.
[22]	JIN T, ZHANG W, CHEN C L, et al. Deep-learning- and unmanned aerial vehicle-based structural crack detection in concrete[J]. Buildings, 2023, 13(12): 3114.1-3114.15.
[23]	SALAAN C J O, OKADA Y, MIZUTANI S, et al. Close visual bridge inspection using a UAV with a passive rotating spherical shell[J]. Journal of Field Robotics, 2018, 35(6): 850-867. doi: 10.1002/rob.21781
[24]	WANG H F, ZHAI L, HUANG H, et al. Measurement for cracks at the bottom of bridges based on tethered creeping unmanned aerial vehicle[J]. Automation in Construction, 2020, 119: 103330.1-103330.13.
[25]	XU Z, WANG Y W, HAO X T, et al. Crack detection of bridge concrete components based on large-scene images using an unmanned aerial vehicle[J]. Sensors, 2023, 23(14): 6271.1-6271.14.
[26]	KIM B, NATARAJAN Y, SRI PREETHAA K R, et al. Real-time assessment of surface cracks in concrete structures using integrated deep neural networks with autonomous unmanned aerial vehicle[J]. Engineering Applications of Artificial Intelligence, 2024, 129: 107537.1-107537.12.
[27]	CHEN Y, MEI T, WANG X J, et al. A bridge crack image detection and classification method based on climbing robot[C]//2016 35th Chinese Control Conference (CCC). Chengdu: IEEE, 2016: 4037-4042.
[28]	MENG Q L, YANG J B, ZHANG Y, et al. A robot system for rapid and intelligent bridge damage inspection based on deep-learning algorithms[J]. Journal of Performance of Constructed Facilities, 2023, 37(6): 04023052.1-04023052.14.
[29]	GIBB S, LA H M, LE T, et al. Nondestructive evaluation sensor fusion with autonomous robotic system for civil infrastructure inspection[J]. Journal of Field Robotics, 2018, 35(6): 988-1004. doi: 10.1002/rob.21791
[30]	LI X K, GAO C, GUO Y C, et al. Cable surface damage detection in cable-stayed bridges using optical techniques and image mosaicking[J]. Optics & Laser Technology, 2019, 110: 36-43.
[31]	YANG L, LI B, FENG J L, et al. Automated wall-climbing robot for concrete construction inspection[J]. Journal of Field Robotics, 2023, 40(1): 110-129. doi: 10.1002/rob.22119
[32]	谢文高,张怡孝,刘爱荣,等. 基于水下机器人与数字图像技术的混凝土结构表面裂缝检测方法[J]. 工程力学,2022,39(增1): 64-70. XIE Wengao, ZHANG Yixiao, LIU Airong, et al. Detection method of concrete structure surface cracks based on underwater robot and digital image technology[J]. Engineering Mechanics, 2022, 39(S1): 64-70.
[33]	YAMAN O, TUNCER T, ERTAM F. An automated crack detection method for underwater structures based on multilevel DWT and LPQ feature generation[J]. Multimedia Tools and Applications, 2023, 82(27): 42331-42352. doi: 10.1007/s11042-023-15229-9
[34]	邱昕捷,韩凤磊,赵望源. 水下机器人实时智能裂缝检测算法[J]. 哈尔滨工程大学学报,2023,44(5): 774-782. doi: 10.11990/jheu.202108022 QIU Xinjie, HAN Fenglei, ZHAO Wangyuan. Real-time intelligent crack detection algorithm for underwater vehicles[J]. Journal of Harbin Engineering University, 2023, 44(5): 774-782. doi: 10.11990/jheu.202108022
[35]	JANG K, AN Y K, KIM B, et al. Automated crack evaluation of a high-rise bridge pier using a ring-type climbing robot[J]. Computer-Aided Civil and Infrastructure Engineering, 2021, 36(1): 14-29. doi: 10.1111/mice.12550
[36]	LIU Y F, NIE X, FAN J S, et al. Image-based crack assessment of bridge piers using unmanned aerial vehicles and three-dimensional scene reconstruction[J]. Computer-Aided Civil and Infrastructure Engineering, 2020, 35(5): 511-529. doi: 10.1111/mice.12501
[37]	刘宇飞,樊健生,孔思宇,等. 多视角几何三维重建法识别工程结构缺损与变形[J]. 工程力学,2020,37(9): 103-111. doi: 10.6052/j.issn.1000-4750.2019.10.0598 LIU Yufei, FAN Jiansheng, KONG Siyu, et al. Detection of structural defect and deformation based on multi-view geometric three-dimensional reconstruction method[J]. Engineering Mechanics, 2020, 37(9): 103-111. doi: 10.6052/j.issn.1000-4750.2019.10.0598
[38]	LIU Y F, CHO S, SPENCER B F J, et al. Concrete crack assessment using digital image processing and 3D scene reconstruction[J]. Journal of Computing in Civil Engineering, 2016, 30(1): 04014124.1-04014124.19.
[39]	WOO H J, SEO D M, KIM M S, et al. Localization of cracks in concrete structures using an unmanned aerial vehicle[J]. Sensors, 2022, 22(17): 6711.1-6711.13.
[40]	ZENG Q Q, FAN G, WANG D Y, et al. A systematic approach to pixel-level crack detection and localization with a feature fusion attention network and 3D reconstruction[J]. Engineering Structures, 2024, 300: 117219.1-117219.17.
[41]	HUA L X, DENG J H, SHI Z F, et al. Single-stripe-enhanced spacetime stereo reconstruction for concrete defect identification[J]. Automation in Construction, 2023, 156: 105136.1-105136.13.
[42]	DENG L, SUN T, YANG L, et al. Binocular video-based 3D reconstruction and length quantification of cracks in concrete structures[J]. Automation in Construction, 2023, 148: 104743.1-104743.15.
[43]	XIAO J L, FAN J S, LIU Y F, et al. Region of interest (ROI) extraction and crack detection for UAV-based bridge inspection using point cloud segmentation and 3D-to-2D projection[J]. Automation in Construction, 2024, 158: 105226.1-105226.18.
[44]	LIU Y F, GAO W L, ZHAO T T, et al. A rapid bridge crack detection method based on deep learning[J]. Applied Sciences, 2023, 13(17): 9878.1-9878.17.
[45]	JIN T, YE X W, LI Z X. Establishment and evaluation of conditional GAN-based image dataset for semantic segmentation of structural cracks[J]. Engineering Structures, 2023, 285: 116058.1-116058.15.
[46]	ZHONG J T, JU H Y, ZHANG W G, et al. A deeper generative adversarial network for grooved cement concrete pavement crack detection[J]. Engineering Applications of Artificial Intelligence, 2023, 119: 105808.1-105808.13.
[47]	KHODAMBASHI S, MOGHADDAM M E. An impulse noise fading technique based on local histogram processing[C]//2009 IEEE International Symposium on Signal Processing and Information Technology (ISSPIT). Ajman: IEEE, 2009: 95-100.
[48]	ZHANG L N, ZHANG L J, ZHANG L D. Application research of digital media image processing technology based on wavelet transform[J]. EURASIP Journal on Image and Video Processing, 2018, 2018(1): 138.1-138.10.
[49]	GAO H, WEI P, KE J. Color enhancement and image defogging in HSI based on Retinex model[C]//2015 International Conference on Optical Instruments and Technology: Optoelectronic Imaging and Processing Technology. Beijing: SPIE, 2015: 9-16.
[50]	OH J K, JANG G, OH S, et al. Bridge inspection robot system with machine vision[J]. Automation in Construction, 2009, 18(7): 929-941. doi: 10.1016/j.autcon.2009.04.003
[51]	孙亮,邢建春,谢立强,等. 基于自适应阈值Canny算法的裂缝检测方法研究[J]. 微型机与应用,2017,36(5): 35-37,41. SUN Liang, XING Jianchun, XIE Liqiang, et al. An adaptive threshold-based Canny algorithm for crack detection[J]. Microcomputer & Its Applications, 2017, 36(5): 35-37,41.
[52]	王楹. 一种改进的桥梁裂缝图像滤波算法[J]. 重庆交通大学学报(自然科学版),2017,36(12): 13-17. doi: 10.3969/j.issn.1674-0696.2017.12.03 WANG Ying. An improved filtering algorithm for bridge crack image[J]. Journal of Chongqing Jiaotong University (Natural Science), 2017, 36(12): 13-17. doi: 10.3969/j.issn.1674-0696.2017.12.03
[53]	李晋惠. 公路路面裂缝类病害图像处理算法研究[J]. 计算机工程与应用,2003,39(35): 212-213,232. doi: 10.3321/j.issn:1002-8331.2003.35.069 LI Jinhui. Image processing algorithm for detecting the pavement crack diseases[J]. Computer Engineering and Applications, 2003, 39(35): 212-213,232. doi: 10.3321/j.issn:1002-8331.2003.35.069
[54]	张雪峰,闫慧. 基于中值滤波和分数阶滤波的图像去噪与增强算法[J]. 东北大学学报(自然科学版),2020,41(4): 482-487. ZHANG Xuefeng, YAN Hui. Image denoising and enhancement algorithm based on Median filtering and fractional order filtering[J]. Journal of Northeastern University (Natural Science), 2020, 41(4): 482-487.
[55]	王松林. 一种改进的自适应加权中值去噪算法的研究[D]. 武汉:武汉科技大学,2017.
[56]	余博,郭雷,钱晓亮,等. 一种新的自适应双边滤波算法[J]. 应用科学学报,2012,30(5): 517-523. doi: 10.3969/j.issn.0255-8297.2012.05.013 YU Bo, GUO Lei, QIAN Xiaoliang, et al. A new adaptive bilateral filtering[J]. Journal of Applied Sciences, 2012, 30(5): 517-523. doi: 10.3969/j.issn.0255-8297.2012.05.013
[57]	LEE J H, GWON G H, KIM I H, et al. A motion deblurring network for enhancing UAV image quality in bridge inspection[J]. Drones, 2023, 7(11): 657.1-657.20.
[58]	XIE R P, YAO J, LIU K, et al. Automatic multi-image stitching for concrete bridge inspection by combining point and line features[J]. Automation in Construction, 2018, 90: 265-280. doi: 10.1016/j.autcon.2018.02.021
[59]	秦海伟,陆从飞,蒋永生,等. 轨道式攀爬机器人系统设计与图像处理[J]. 中国安全科学学报,2019,29(7): 117-122. QIN Haiwei, LU Congfei, JIANG Yongsheng, et al. Design of orbital climbing robot system and image processing[J]. China Safety Science Journal, 2019, 29(7): 117-122.
[60]	张昊宇,丁勇,李登华. 基于三维重建的结构表面裂缝检测方法[J]. 工业建筑,2023,54(5): 60-67. ZHANG Haoyu, DING Yong, LI Denghua. Astructural surface crack detection method based on 3d reconstruction[J]. Industrial Construction, 2023, 54(5): 60-67.
[61]	丁威,俞珂,舒江鹏. 基于深度学习和无人机的混凝土结构裂缝检测方法[J]. 土木工程学报,2021,54(增1): 1-12. DING Wei, YU Ke, SHU Jiangpeng. Method for detecting cracks in concrete structures based on deep learning and UAV[J]. China Civil Engineering Journal, 2021, 54(S1): 1-12.
[62]	AL-HUDA Z, PENG B, ALI ALGBURI R N, et al. Weakly supervised pavement crack semantic segmentation based on multi-scale object localization and incremental annotation refinement[J]. Applied Intelligence, 2023, 53(11): 14527-14546. doi: 10.1007/s10489-022-04212-w
[63]	PENNADA S, PERRY M, MCALORUM J, et al. Threshold-based BRISQUE-assisted deep learning for enhancing crack detection in concrete structures[J]. Journal of Imaging, 2023, 9(10): 218.1-218.19.
[64]	HYEGEUN. 크랙 object detection dataset[DB/OL]. Roboflow: HYEGEUN, 2022[2024-07-02]. https://universe.roboflow.com/hyegeun/-rxh4q.
[65]	ZHANG L, YANG F, DANIEL ZHANG Y, et al. Road crack detection using deep convolutional neural network[C]//2016 IEEE International Conference on Image Processing (ICIP). Phoenix: IEEE, 2016: 3708-3712.
[66]	REN Y P, HUANG J S, HONG Z Y, et al. Image-based concrete crack detection in tunnels using deep fully convolutional networks[J]. Construction and Building Materials, 2020, 234: 117367.1-117367.12.
[67]	YANG X C, LI H, YU Y T, et al. Automatic pixel‐level crack detection and measurement using fully convolutional network[J]. Computer-Aided Civil and Infrastructure Engineering, 2018, 33(12): 1090-1109. doi: 10.1111/mice.12412
[68]	SHI Y, CUI L M, QI Z Q, et al. Automatic road crack detection using random structured forests[J]. IEEE Transactions on Intelligent Transportation Systems, 17(12): 3434-3445.
[69]	DORAFSHAN S, THOMAS R J, MAGUIRE M. SDNET2018: an annotated image dataset for non-contact concrete crack detection using deep convolutional neural networks[J]. Data in Brief, 2018, 21: 1664-1668. doi: 10.1016/j.dib.2018.11.015
[70]	XU H Y, SU X, WANG Y, et al. Automatic bridge crack detection using a convolutional neural network[J]. Applied Sciences, 2019, 9(14): 2867.1-2867.14.
[71]	LAXMAN K C, TABASSUM N, AI L, et al. Automated crack detection and crack depth prediction for reinforced concrete structures using deep learning[J]. Construction and Building Materials, 2023, 370: 130709.1-130709.11.
[72]	SUN Z, CAETANO E, PEREIRA S, et al. Employing histogram of oriented gradient to enhance concrete crack detection performance with classification algorithm and Bayesian optimization[J]. Engineering Failure Analysis, 2023, 150: 107351.1-107351.15.
[73]	CROGNALE M, DE IULIIS M, RINALDI C, et al. Damage detection with image processing: a comparative study[J]. Earthquake Engineering and Engineering Vibration, 2023, 22(2): 333-345. doi: 10.1007/s11803-023-2172-1
[74]	周颖,刘彤. 基于计算机视觉的混凝土裂缝识别[J]. 同济大学学报(自然科学版),2019,47(9): 1277-1285. ZHOU Ying, LIU Tong. Computer vision-based crack detection and measurement on concrete structure[J]. Journal of Tongji University (Natural Science), 2019, 47(9): 1277-1285.
[75]	杨才千,李帅,王博昆,等. 基于动态阈值的混凝土裂缝高抗噪提取及识别方法[J]. 东南大学学报(自然科学版),2021,51(6): 967-972. YANG Caiqian, LI Shuai, WANG Bokun, et al. High anti-noise extraction and identification method for concrete cracks based on dynamic threshold[J]. Journal of Southeast University (Natural Science Edition), 2021, 51(6): 967-972.
[76]	DOW H, PERRY M, MCALORUM J, et al. Skeleton-based noise removal algorithm for binary concrete crack image segmentation[J]. Automation in Construction, 2023, 151: 104867.1-104867.10.
[77]	张振海,季坤,党建武. 基于桥梁裂缝识别模型的桥梁裂缝病害识别方法[J]. 吉林大学学报(工学版),2023,53(5): 1418-1426. ZHANG Zhenhai, JI Kun, DANG Jianwu. Crack identification method for bridge based on BCEM model[J]. Journal of Jilin University (Engineering and Technology Edition), 2023, 53(5): 1418-1426.
[78]	刘春,艾克然木•艾克拜尔,蔡天池. 面向建筑健康监测的无人机自主巡检与裂缝识别[J]. 同济大学学报(自然科学版),2022,50(7): 921-932,918. LIU Chun, AKBAR Akram, CAI Tianchi. UAV autonomous inspection and crack detection towards building health monitoring[J]. Journal of Tongji University (Natural Science), 2022, 50(7): 921-932,918.
[79]	REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas: IEEE, 2016: 779-788.
[80]	NABIZADEH E, PARGHI A. Vision-based concrete crack detection using deep learning-based models[J]. Asian Journal of Civil Engineering, 2023, 24(7): 2389-2403. doi: 10.1007/s42107-023-00648-8
[81]	黄贲,康飞,唐玉. 基于目标检测的混凝土坝裂缝实时检测方法[J]. 清华大学学报(自然科学版),2023,63(7): 1078-1086. HUANG Ben, KANG Fei, TANG Yu. A real-time detection method for concrete dam cracks based on an object detection algorithm[J]. Journal of Tsinghua University (Science and Technology), 2023, 63(7): 1078-1086.
[82]	YE G T, QU J S, TAO J T, et al. Autonomous surface crack identification of concrete structures based on the YOLOv7 algorithm[J]. Journal of Building Engineering, 2023, 73: 106688.1-106688.15.
[83]	WU Y, HAN Q B, JIN Q L, et al. LCA-YOLOv8-seg: an improved lightweight YOLOv8-seg for real-time pixel-level crack detection of dams and bridges[J]. Applied Sciences, 2023, 13(19): 10583.1-10583.17.
[84]	XU G, YUE Q R, LIU X G. Deep learning algorithm for real-time automatic crack detection, segmentation, qualification[J]. Engineering Applications of Artificial Intelligence, 2023, 126: 107085.1-107085.22.
[85]	LI G, FANG Z Y, MOHAMMED A M, et al. Automated bridge crack detection based on improving encoder–decoder network and strip pooling[J]. Journal of Infrastructure Systems, 2023, 29(2): 04023004.1-04023004.13.
[86]	CHEN B, ZHANG H, WANG G J, et al. Automatic concrete infrastructure crack semantic segmentation using deep learning[J]. Automation in Construction, 2023, 152: 104950.1-104950.13.
[87]	CHEN Y L, DONG S J, HU B, et al. A dynamic semantic segmentation algorithm with encoder-crossor-decoder structure for pixel-level building cracks[J]. Measurement Science and Technology, 2024, 35(2): 025139.1-025139.17.
[88]	WENG X X, HUANG Y C, LI Y N, et al. Unsupervised domain adaptation for crack detection[J]. Automation in Construction, 2023, 153: 104939.1-104939.11.
[89]	余加勇,李锋,薛现凯,等. 基于无人机及Mask R-CNN的桥梁结构裂缝智能识别[J]. 中国公路学报,2021,34(12): 80-90. doi: 10.3969/j.issn.1001-7372.2021.12.007 YU Jiayong, LI Feng, XUE Xiankai, et al. Intelligent identification of bridge structural cracks based on unmanned aerial vehicle and mask R-CNN[J]. China Journal of Highway and Transport, 2021, 34(12): 80-90. doi: 10.3969/j.issn.1001-7372.2021.12.007
[90]	BAE H, AN Y K. Computer vision-based statistical crack quantification for concrete structures[J]. Measurement, 2023, 211: 112632.1-112632.13.
[91]	孙晓贺,施成华,刘凌晖,等. 基于改进的种子填充算法的混凝土裂缝图像识别系统[J]. 华南理工大学学报(自然科学版),2022,50(5): 127-136,146. doi: 10.12141/j.issn.1000-565X.210602 SUN Xiaohe, SHI Chenghua, LIU Linghui, et al. Concrete crack image recognition system based on improved seed filling algorithm[J]. Journal of South China University of Technology (Natural Science Edition), 2022, 50(5): 127-136,146. doi: 10.12141/j.issn.1000-565X.210602
[92]	MENG Q C, HU L, WAN D, et al. Image-based concrete cracks identification under complex background with lightweight convolutional neural network[J]. KSCE Journal of Civil Engineering, 2023, 27(12): 5231-5242. doi: 10.1007/s12205-023-0923-1
[93]	余加勇,刘宝麟,尹东,等. 基于YOLOv5和U-Net3+ 的桥梁裂缝智能识别与测量[J]. 湖南大学学报(自然科学版),2023,50(5): 65-73. YU Jiayong, LIU Baolin, YIN Dong, et al. Intelligent identification and measurement of bridge cracks based on YOLOv5 and U-Net3+ [J]. Journal of Hunan University (Natural Sciences), 2023, 50(5): 65-73.
[94]	VASWANI A, SHAZEER N, PARMAR N, et al. Attention is all you need[C]//2017 Advances in Neural Information Processing Systems. Long Beach: NIPS, 2017: 5999-6009.
[95]	DOSOVITSKIY A, BEYER L, KOLESNIKOV A, et al. AN image is worth 16x16 words: transformers for image recognition at scale[C]//ICLR 2021-9th International Conference on Learning Representations. Virtual: ICLR, 2021: 1-21.
[96]	WU Y J, LI S Q, ZHANG J G, et al. Dual attention transformer network for pixel-level concrete crack segmentation considering camera placement[J]. Automation in Construction, 2024, 157: 105166.1-105166.15.
[97]	CHU H H, DENG L, YUAN H Q, et al. A transformer and self-cascade operation-based architecture for segmenting high-resolution bridge cracks[J]. Automation in Construction, 2024, 158: 105194.1-105194.13.
[98]	KIRILLOV A, MINTUN E, RAVI N, et al. Segment Anything[C/OL]//2023 IEEE/CVF International Conference on Computer Vision (ICCV). Paris: IEEE, 2023: 3992-4003.
[99]	MORTON COLBERT Z, ARRINGTON D, FOOTE M, et al. Repurposing traditional U-net predictions for sparse SAM prompting in medical image segmentation[J]. Biomedical Physics & Engineering Express, 2024, 10(2): 025004.1-025004.10.
[100]	LI Y, WANG D, YUAN C, et al. Enhancing agricultural image segmentation with an agricultural segment anything model adapter[J]. Sensors, 2023, 23(18): 7884.1-7884.16.
[101]	OSCO L P, WU Q S, DE LEMOS E L, et al. The Segment anything model (SAM) for remote sensing applications: from zero to one shot[J]. International Journal of Applied Earth Observation and Geoinformation, 2023, 124: 103540.1-103540.18.
[102]	ZHENG K Q, NI S W, ZHANG Y H, et al. Post-cracking shear stiffness model of reinforced concrete beams[J]. Buildings, 2023, 13(11): 2814.1-2814.13.
[103]	SAITO S, MAKI T, TSUCHIYA S, et al. Strain based damage indicators for finite element analysis of reinforced concrete structures[J]. Structures, 2023, 58: 105473.1-105473.11.
[104]	KABOODKHANI M, BAYESTEH H, HAMIDIA M. Energy-based damage assessment of RC frames with non-seismic beam-column joint detailing using crack image processing techniques[J]. Engineering Failure Analysis, 2024, 155: 107723.1-107723.20.
[105]	AFZALI M, HAMIDIA M, SAFI M. Data-driven strength-based seismic damage index measurement for RC columns using crack image-derived parameters[J]. Measurement, 2023, 218: 113155.1-113155.16.
[106]	JAMSHIDIAN S, HAMIDIA M. Post-earthquake damage assessment for RC columns using crack image complexity measures[J]. Bulletin of Earthquake Engineering, 2023, 21(13): 6029-6063. doi: 10.1007/s10518-023-01745-4
[107]	XU S, TANG H D, WANG X R, et al. Assessment of geometric parameters of segmented crack on concrete building facade using deep learning[J]. Structures, 2023, 57: 105188.1-105188.14.
[108]	ASJODI A H, DOLATSHAHI K M, BURTON H V. Three-dimensional fragility surface for reinforced concrete shear walls using image-based damage features[J]. Earthquake Engineering & Structural Dynamics, 2023, 52(8): 2533-2553.
[109]	ZHOU K Y, LEI D, CHUN P J, et al. Evaluation of BFRP strengthening and repairing effects on concrete beams using DIC and YOLO-v5 object detection algorithm[J]. Construction and Building Materials, 2024, 411: 134594.1-134594.16.
[110]	WAY A C, VILJOEN C. Reliability performance in tension governed reinforced concrete reservoirs as a function of target crack width[J]. Structural Concrete, 2023, 24(6): 7726-7741. doi: 10.1002/suco.202201159
[111]	LIU P F, CHEN J Y, XU Q, et al. Seismic stability analysis of CSG dams considering the effect of tension crack based on genetic algorithm with an improved initial population strategy[J]. Soil Dynamics and Earthquake Engineering, 2023, 175: 108210.1-108210.10.
[112]	BORAH M M, SIL A. Service-life estimation of a reinforced concrete bridge structure exposed to chloride-contaminated environments and variable traffic loads[J]. Journal of Risk and Uncertainty in Engineering Systems, Part A: Civil Engineering, 2023, 9(4): 04023028.1-04023028.18.

施引文献

附加材料(0)

访问统计

点击查看大图

图(15) / 表(3)

计量

文章访问数: 87
HTML全文浏览量: 40
PDF下载量: 17
被引次数: 0

1. 贯通供电
2. 贯通供电合环
3. 穿越功率分析与计算
3.1 等效电路与均衡电流计算
3.2 穿越功率计算与测量
4. 穿越功率解决方案
5. 多功能的智能牵引变电所
6. 结　论

基于计算机视觉的混凝土结构裂缝检测研究现状与展望

doi: 10.3969/j.issn.0258-2724.20240115

作者简介: 谢明志（1985—），男，副教授，博士，硕士研究生导师，研究方向为桥梁智能检测及损伤评估，E-mail：mzxie@home.swjtu.edu.cn

计量

出版历程

Research Status and Prospects of Computer Vision-Based Crack Detection of Concrete Structure

1. 贯通供电

2. 贯通供电合环

3. 穿越功率分析与计算

3.1 等效电路与均衡电流计算

3.2 穿越功率计算与测量

4. 穿越功率解决方案

5. 多功能的智能牵引变电所

6. 结 论

计量

出版历程

目录

1. 贯通供电

2. 贯通供电合环

3. 穿越功率分析与计算

3.1 等效电路与均衡电流计算

3.2 穿越功率计算与测量

4. 穿越功率解决方案

5. 多功能的智能牵引变电所

6. 结 论

作者简介:
谢明志（1985—），男，副教授，博士，硕士研究生导师，研究方向为桥梁智能检测及损伤评估，E-mail：mzxie@home.swjtu.edu.cn

6. 结　论

6. 结　论