CN114684217A - Rail transit health monitoring system and method - Google Patents
Rail transit health monitoring system and method Download PDFInfo
- Publication number
- CN114684217A CN114684217A CN202210259899.3A CN202210259899A CN114684217A CN 114684217 A CN114684217 A CN 114684217A CN 202210259899 A CN202210259899 A CN 202210259899A CN 114684217 A CN114684217 A CN 114684217A
- Authority
- CN
- China
- Prior art keywords
- data
- neural network
- vibration
- layer
- rail transit
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 230000036541 health Effects 0.000 title claims abstract description 95
- 238000012544 monitoring process Methods 0.000 title claims abstract description 54
- 238000000034 method Methods 0.000 title claims abstract description 50
- 238000013528 artificial neural network Methods 0.000 claims abstract description 87
- 230000004044 response Effects 0.000 claims abstract description 30
- 238000007781 pre-processing Methods 0.000 claims abstract description 11
- 239000000835 fiber Substances 0.000 claims description 58
- 239000000523 sample Substances 0.000 claims description 50
- 230000006870 function Effects 0.000 claims description 43
- 230000003287 optical effect Effects 0.000 claims description 36
- 238000012549 training Methods 0.000 claims description 30
- 238000011176 pooling Methods 0.000 claims description 26
- 239000013307 optical fiber Substances 0.000 claims description 23
- 230000008569 process Effects 0.000 claims description 18
- 238000010606 normalization Methods 0.000 claims description 17
- 238000012423 maintenance Methods 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 13
- 238000002310 reflectometry Methods 0.000 claims description 9
- 238000012360 testing method Methods 0.000 claims description 9
- 238000004458 analytical method Methods 0.000 claims description 8
- ORILYTVJVMAKLC-UHFFFAOYSA-N Adamantane Natural products C1C(C2)CC3CC1CC2C3 ORILYTVJVMAKLC-UHFFFAOYSA-N 0.000 claims description 6
- 230000007613 environmental effect Effects 0.000 claims description 6
- 238000004891 communication Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 4
- 230000009467 reduction Effects 0.000 claims description 4
- 239000000853 adhesive Substances 0.000 claims description 3
- 230000001070 adhesive effect Effects 0.000 claims description 3
- 238000003491 array Methods 0.000 claims description 3
- 239000004568 cement Substances 0.000 claims description 3
- 239000003795 chemical substances by application Substances 0.000 claims description 3
- 238000013527 convolutional neural network Methods 0.000 description 18
- 238000001514 detection method Methods 0.000 description 13
- 238000010586 diagram Methods 0.000 description 10
- 230000003862 health status Effects 0.000 description 10
- 238000004590 computer program Methods 0.000 description 7
- 238000011156 evaluation Methods 0.000 description 6
- 230000002159 abnormal effect Effects 0.000 description 5
- 230000004913 activation Effects 0.000 description 5
- 238000012806 monitoring device Methods 0.000 description 5
- 201000010099 disease Diseases 0.000 description 4
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 4
- 210000002569 neuron Anatomy 0.000 description 4
- 230000035945 sensitivity Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000002085 persistent effect Effects 0.000 description 3
- 238000012935 Averaging Methods 0.000 description 2
- 238000013459 approach Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 230000005284 excitation Effects 0.000 description 2
- 230000007774 longterm Effects 0.000 description 2
- 239000000758 substrate Substances 0.000 description 2
- 238000002759 z-score normalization Methods 0.000 description 2
- 230000009471 action Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013523 data management Methods 0.000 description 1
- 230000002950 deficient Effects 0.000 description 1
- 230000006866 deterioration Effects 0.000 description 1
- 230000008376 long-term health Effects 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000005693 optoelectronics Effects 0.000 description 1
- 230000008447 perception Effects 0.000 description 1
- 230000010287 polarization Effects 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 231100000241 scar Toxicity 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000008054 signal transmission Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000010200 validation analysis Methods 0.000 description 1
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61K—AUXILIARY EQUIPMENT SPECIALLY ADAPTED FOR RAILWAYS, NOT OTHERWISE PROVIDED FOR
- B61K9/00—Railway vehicle profile gauges; Detecting or indicating overheating of components; Apparatus on locomotives or cars to indicate bad track sections; General design of track recording vehicles
- B61K9/08—Measuring installations for surveying permanent way
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B61—RAILWAYS
- B61L—GUIDING RAILWAY TRAFFIC; ENSURING THE SAFETY OF RAILWAY TRAFFIC
- B61L15/00—Indicators provided on the vehicle or train for signalling purposes
- B61L15/0054—Train integrity supervision, e.g. end-of-train [EOT] devices
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Mechanical Engineering (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Molecular Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Health & Medical Sciences (AREA)
- Testing Of Devices, Machine Parts, Or Other Structures Thereof (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
Description
技术领域technical field
本发明涉及交通安全监测技术领域,尤其涉及一种轨道交通健康监测系统及方法。The invention relates to the technical field of traffic safety monitoring, in particular to a rail transit health monitoring system and method.
背景技术Background technique
近年来城市轨道运输系统迅猛发展,作为重大结构工程和城市交通命脉重要组成部分的城市轨道交通工程结构,其健康服役对于城市正常运转至关重要。然而,轨道交通工程结构随着运营逐渐伤损,尤其是在温度、列车、基础变形、外部侵扰等复杂载荷条件影响下,随时可能出现病害,影响行车安全。轨道交通工程结构病害表现出频发、突发性,因此,亟待构建面向轨道交通工程结构全时全域安全监测系统,全面及时地掌握车辆和轨道工程结构的健康状况,对潜在的病害和突发的事故隐患及时地预警和报警,为轨道交通安全运营保驾护航。目前我国的城市轨道交通工程结构状态检测、维修大多是定期检修或者事后故障修复的方式,运营护养和维修的成本高、效率低,无法保证监测的实时性。因此,迫切需要利用新技术、新方法,提升轨道交通工程基础设施的安全监测技术和管理水平。传统电类检测系统,在电气化铁路段受到电磁干扰严重,信号无法远距离传输。光纤传感器由于具有长期稳定性佳,信号可远程传输,易于组网,抗电磁干扰等优势,已逐步取代长期稳定性难以得到保证的电类传感器,并在桥梁、隧道、机场、铁路等领域的长期健康监测中被广泛采用。In recent years, the urban rail transit system has developed rapidly. As a major structural project and an important part of the lifeline of urban transportation, the healthy service of the urban rail transit engineering structure is very important for the normal operation of the city. However, the structure of rail transit engineering is gradually damaged with operation, especially under the influence of complex load conditions such as temperature, train, foundation deformation, and external intrusion, diseases may occur at any time, affecting the driving safety. Structural diseases of rail transit engineering show frequent occurrence and suddenness. Therefore, it is urgent to build a full-time and all-domain safety monitoring system for rail transit engineering structures, to comprehensively and timely grasp the health status of vehicles and rail engineering structures, and to prevent potential diseases and emergencies. The hidden dangers of accidents can be warned and alarmed in a timely manner to escort the safe operation of rail transit. At present, most of the structural status detection and maintenance of urban rail transit projects in my country are carried out by means of regular maintenance or post-event repair. The cost of operation, maintenance and maintenance is high and the efficiency is low, and the real-time monitoring cannot be guaranteed. Therefore, it is urgent to use new technologies and new methods to improve the safety monitoring technology and management level of rail transit engineering infrastructure. The traditional electrical detection system is severely affected by electromagnetic interference in the electrified railway section, and the signal cannot be transmitted over a long distance. Optical fiber sensors have gradually replaced electrical sensors whose long-term stability is difficult to guarantee due to their good long-term stability, long-distance signal transmission, easy networking, and anti-electromagnetic interference. Widely used in long-term health monitoring.
现有技术中,发明专利201110009091.1“光纤光栅传感列车车轮踏面状态在线监测系统”公开了一种基于多个光纤光栅组成列车车轮健康状态系统,该专利利用安装在光纤光栅应变传感器检测轮轨耦合作用下钢轨的应力变化,正常车轮通过时会产生平滑的应变响应曲线,当车轮踏面发生异常时,应变响应曲线会发生畸变,通过对畸变信号分析进而判断车轮的故障原因。但该系统检测的可靠性不高,容易造成误报,例如当传感装置发生问题时也会造成应变信号异常,易造成错误判断;发明专利202010666848.3“一种基于光纤分布式振动监测的高铁铁轨安全运行检测方法”采用的是传统的光纤分布式振动检测方法,传统的分布式光纤传感技术由于光纤中散射系数较低,系统的信噪比通常不高,进而影响到分布式传感系统的空间分辨率及检测灵敏度。为解决此问题,研究人员通常采用多次平均的算法或人为增加散射系数的方法。数千次的平均可以改善信噪比,但会造成系统响应时间的增加。而通过紫外线曝光或飞秒激光加工等方法来改善后向散射强度可有效提高信噪比,但其工艺上的均匀性与时效性又存在局限。因此,如何利用光纤光栅传感器进行轨道交通的有效监测是亟待解决的问题。In the prior art, the invention patent 201110009091.1 "Fiber Bragg Grating Sensing Train Wheel Tread Status Online Monitoring System" discloses a train wheel health status system based on multiple optical fiber gratings. Under the action of the stress change of the rail, a smooth strain response curve will be generated when the normal wheel passes, and when the wheel tread is abnormal, the strain response curve will be distorted, and the cause of the wheel failure can be determined by analyzing the distortion signal. However, the detection reliability of the system is not high, and it is easy to cause false alarms. For example, when there is a problem with the sensing device, the strain signal will also be abnormal, which is easy to cause misjudgment; Invention patent 202010666848.3 "A high-speed rail track based on optical fiber distributed vibration monitoring "Safe operation detection method" adopts the traditional optical fiber distributed vibration detection method. Due to the low scattering coefficient in the optical fiber, the traditional distributed optical fiber sensing technology usually has a low signal-to-noise ratio of the system, which in turn affects the distributed sensing system. The spatial resolution and detection sensitivity. To solve this problem, researchers usually use multiple averaging algorithms or artificially increase the scattering coefficient. Averaging over thousands of times can improve the signal-to-noise ratio at the cost of an increase in system response time. However, improving the backscattering intensity by means of ultraviolet exposure or femtosecond laser processing can effectively improve the signal-to-noise ratio, but there are limitations in the uniformity and timeliness of the process. Therefore, how to use fiber grating sensors to effectively monitor rail traffic is an urgent problem to be solved.
发明内容SUMMARY OF THE INVENTION
有鉴于此,有必要提供及一种轨道交通健康监测系统及方法,用以克服现有技术中难以高效利用光纤光栅传感器预测轨道交通情况的问题。In view of this, it is necessary to provide a rail transit health monitoring system and method to overcome the problem that it is difficult to efficiently use fiber grating sensors to predict rail transit conditions in the prior art.
为了解决上述技术问题,本发明提供一种轨道交通健康监测系统,包括:依次通信连接的振动传感网络、数据处理中心以及智能化运维服务平台,其中:In order to solve the above-mentioned technical problems, the present invention provides a rail transit health monitoring system, comprising: a vibration sensor network, a data processing center and an intelligent operation and maintenance service platform connected in sequence in communication, wherein:
所述振动传感网络,包括布设于轨道交通工程结构上的光纤光栅振动传感光缆;The vibration sensing network includes a fiber grating vibration sensing optical cable arranged on a rail transit engineering structure;
所述数据处理中心,用于对所述光纤光栅振动传感光缆、所述光纤光栅应变传感光缆以及所述光纤光栅阵列温度传感光缆的监测信号进行处理,确定响应数据;The data processing center is used to process the monitoring signals of the fiber grating vibration sensing optical cable, the fiber grating strain sensing optical cable, and the fiber grating array temperature sensing optical cable to determine response data;
所述智能化运维服务平台,用于根据所述响应数据进行智能分析,基于智能分析结果,对轨道结构和车辆进行监测。The intelligent operation and maintenance service platform is used to perform intelligent analysis according to the response data, and monitor the track structure and vehicles based on the intelligent analysis results.
进一步地,所述振动传感网络中的光纤光栅阵列都由光纤光栅阵列传感探头制成,所述光纤光栅阵列传感探头包括预设反射率光栅阵列,单根光纤上复用预设规模的光纤光栅传感测点。Further, the fiber grating arrays in the vibration sensing network are all made of fiber grating array sensing probes, and the fiber grating array sensing probes include a preset reflectivity grating array, and a single fiber multiplexes a preset scale. The fiber grating sensing point.
进一步地,每根所述光纤光栅阵列传感探头上的光纤光栅为等间距布设,以等间距的光纤光栅为节点,以相邻的节点形成振动感知单元。Further, the fiber gratings on each of the fiber grating array sensing probes are arranged at equal intervals, and the equally spaced fiber gratings are used as nodes, and adjacent nodes are used to form vibration sensing units.
进一步地,所述振动传感网络中的振动传感光缆与被测结构密贴,其中:Further, the vibration sensing optical cable in the vibration sensing network is closely attached to the measured structure, wherein:
若所述被测结构属于既有线路,则直接用结构胶粘贴于所述被测结构,或采用开槽后放入振动传感光缆后再利用水泥或固化剂将振动传感光缆与所述被测结构固化;If the structure to be tested belongs to an existing line, directly stick it to the structure under test with structural adhesive, or use a slot and then put the vibration sensing optical fiber cable, and then use cement or curing agent to connect the vibration sensing optical fiber cable to the tested structure. The tested structure is cured;
若所述被测结构属于新修线路,直接将振动传感光缆浇筑在混凝土中。If the structure under test belongs to a newly repaired line, directly pour the vibration sensing optical cable into the concrete.
本发明还提供了一种轨道交通健康监测方法,应用于如上所述的轨道交通健康监测系统中的智能化运维服务平台,所述方法包括:The present invention also provides a rail transit health monitoring method, which is applied to the above-mentioned intelligent operation and maintenance service platform in the rail transit health monitoring system. The method includes:
获取响应数据;get response data;
根据所述响应数据进行预处理,确定待测振动数据;Perform preprocessing according to the response data to determine vibration data to be measured;
将所述待测振动数据输入至训练完备的第一神经网络和/或第二神经网络,确定预测轨道结构健康类别和预测车身健康类别。The vibration data to be measured is input into the well-trained first neural network and/or the second neural network, and the predicted track structure health category and the predicted vehicle body health category are determined.
进一步地,所述响应数据包括第i个区间的拾振器在列车接近的第一时刻与列车驶离的第二时刻之间的振动数据,所述根据所述响应数据进行预处理,确定待测数据,包括:Further, the response data includes the vibration data of the vibration pickup in the ith section between the first moment when the train approaches and the second moment when the train leaves, and the preprocessing is performed according to the response data to determine the waiting time. measurement data, including:
将所述振动数据的前后两端添加环境噪声信号,确定填充数据;Adding environmental noise signals to the front and rear ends of the vibration data to determine filling data;
将所述填充数据进行数据归一化处理,确定所述待测数据。Perform data normalization processing on the filling data to determine the data to be measured.
进一步地,所述第一神经网络和/或所述第二神经网络的训练过程包括:Further, the training process of the first neural network and/or the second neural network includes:
当将样本集输入至所述第一神经网络时,输出预测轨道结构健康类别,其中,所述样本集包括样本振动数据与对应的实际轨道结构健康类别;When the sample set is input to the first neural network, the predicted track structure health category is output, wherein the sample set includes sample vibration data and the corresponding actual track structure health category;
当将样本集输入至所述第二神经网络时,输出预测车身健康类别,其中,所述样本集包括所述样本振动数据与对应的实际车身健康类别;When a sample set is input to the second neural network, the predicted body health category is output, wherein the sample set includes the sample vibration data and the corresponding actual body health category;
根据所述预测轨道结构健康类别和所述实际轨道结构健康类别之间的误差,确定第一损失函数,并根据所述第一损失函数训练构建的第一神经网络至收敛;Determine a first loss function according to the error between the predicted track structure health category and the actual track structure health category, and train the constructed first neural network to convergence according to the first loss function;
根据所述预测车身健康类别和所述实际车身健康类别之间的误差,确定第二损失函数,并根据所述第二损失函数训练构建的第二神经网络至收敛。According to the error between the predicted body health category and the actual body health category, a second loss function is determined, and the constructed second neural network is trained to converge according to the second loss function.
进一步地,所述第一神经网络依次包括:输入层、卷积层、最大池化层、Flatten层、全连接层和输出层,其中:Further, the first neural network sequentially includes: an input layer, a convolutional layer, a maximum pooling layer, a Flatten layer, a fully connected layer and an output layer, wherein:
所述输入层,用于输入所述样本集;the input layer, for inputting the sample set;
所述卷积层,用于对所述样本集进行卷积操作,确定第一卷积数据;The convolution layer is used to perform a convolution operation on the sample set to determine the first convolution data;
所述最大池化层,包括池化层和Batch Normalization层,其中,所述池化层用于对所述第一卷积数据进行特征提取和降维处理,所述Batch Normalization层用于对降维后的第一卷积数据进行标准化处理,确定标准化数据;The maximum pooling layer includes a pooling layer and a Batch Normalization layer, wherein the pooling layer is used to perform feature extraction and dimension reduction processing on the first convolutional data, and the Batch Normalization layer is used to reduce the Standardize the first convolution data after dimensioning to determine standardized data;
所述Flatten层,用于将所述标准化数据展开成一维向量;The Flatten layer is used to expand the standardized data into a one-dimensional vector;
所述全连接层,用于将所述一维向量降维,并通过softmax函数进行分类,确定所述预测车身健康类别。The fully connected layer is used for reducing the dimension of the one-dimensional vector, and classifying it through a softmax function to determine the predicted body health category.
进一步地,所述第二神经网络依次包括:BP神经网络和一维神经网络,其中:Further, the second neural network sequentially includes: a BP neural network and a one-dimensional neural network, wherein:
所述BP神经网络,对速度训练集训练,用于输出预测车速,其中,所述速度训练集包括样本振动数据与对应的速度标签;The BP neural network is trained on a speed training set for outputting a predicted vehicle speed, wherein the speed training set includes sample vibration data and corresponding speed labels;
所述一维神经网络,用于依次通过输入层、卷积层、最大池化层、Flatten层、全连接层和输出层对所述样本集进行训练,确定所述预测车身健康类别。The one-dimensional neural network is used to sequentially train the sample set through an input layer, a convolutional layer, a maximum pooling layer, a Flatten layer, a fully connected layer and an output layer to determine the predicted body health category.
进一步地,在所述第一神经网络和/或所述第二神经网络的训练过程中,采用Adam优化器自适应地调整学习率,且所述第一损失函数和所述第二损失函数为多元交叉熵函数。Further, in the training process of the first neural network and/or the second neural network, the Adam optimizer is used to adjust the learning rate adaptively, and the first loss function and the second loss function are Multivariate cross-entropy function.
与现有技术相比,本发明的有益效果包括:首先,对响应数据进行有效的获取;进而,通过对响应数据的预处理,提取出相关有效的数据,形成待测振动数据,保证待测振动数据的准确性;最后,通过第一神经网络和/或第二神经网络,对待测振动数据进行有效的识别和判断,提取其中的特征信息,预测轨道结构健康类别和车身健康类别,保证了快速高效地基于振动信息,实现多方面的交通情况判别。综上,本发明基于振动传感网络,实现高灵敏度地获取响应数据,将待测振动数据输入至相关的神经网络后,仅基于待测振动数据就可得到多方面交通运行信息,保证了算法的高效性和快速性,直接预测出结果后,相关人员可以参考预测结果,及时了解地铁的健康状况和运行车速,以减少意外的发生,保证了反馈的及时性。Compared with the prior art, the beneficial effects of the present invention include: first, effective acquisition of response data; further, by preprocessing the response data, extracting relevant effective data to form vibration data to be measured, ensuring that the vibration data to be measured is formed. The accuracy of the vibration data; finally, through the first neural network and/or the second neural network, the vibration data to be measured can be effectively identified and judged, the characteristic information in it is extracted, and the track structure health category and body health category are predicted, ensuring that Based on vibration information quickly and efficiently, it can realize multi-faceted traffic situation discrimination. To sum up, the present invention is based on the vibration sensing network to obtain response data with high sensitivity. After inputting the vibration data to be measured into the relevant neural network, various aspects of traffic operation information can be obtained only based on the vibration data to be measured, which ensures the algorithm. After directly predicting the results, the relevant personnel can refer to the prediction results to know the health status and running speed of the subway in time, so as to reduce the occurrence of accidents and ensure the timeliness of feedback.
附图说明Description of drawings
图1为本发明提供的轨道交通安全监测系统一实施例的流程示意图;1 is a schematic flowchart of an embodiment of a rail transit safety monitoring system provided by the present invention;
图2为本发明提供的轨道交通健康监测方法一实施例的流程示意图;2 is a schematic flowchart of an embodiment of a rail transit health monitoring method provided by the present invention;
图3为本发明提供的图2中步骤S202一实施例的流程示意图;FIG. 3 is a schematic flowchart of an embodiment of step S202 in FIG. 2 provided by the present invention;
图4为本发明提供的第一神经网络和/或第二神经网络的训练过程一实施例的流程示意图;4 is a schematic flowchart of an embodiment of a training process of a first neural network and/or a second neural network provided by the present invention;
图5为本发明提供的第一神经网络一实施例的结构示意图;5 is a schematic structural diagram of an embodiment of a first neural network provided by the present invention;
图6为本发明提供的BP神经网络一实施例的结构示意图;6 is a schematic structural diagram of an embodiment of a BP neural network provided by the present invention;
图7为本发明提供的轨道交通健康监测装置一实施例的结构示意图;7 is a schematic structural diagram of an embodiment of a rail transit health monitoring device provided by the present invention;
图8为本发明提供的电子设备一实施例的结构示意图。FIG. 8 is a schematic structural diagram of an embodiment of an electronic device provided by the present invention.
具体实施方式Detailed ways
下面结合附图来具体描述本发明的优选实施例,其中,附图构成本申请一部分,并与本发明的实施例一起用于阐释本发明的原理,并非用于限定本发明的范围。The preferred embodiments of the present invention are specifically described below with reference to the accompanying drawings, wherein the accompanying drawings constitute a part of the present application, and together with the embodiments of the present invention, are used to explain the principles of the present invention, but are not used to limit the scope of the present invention.
在本发明的描述中,术语“第一”、“第二”仅用于描述目的,而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此,限定有“第一”、“第二”的特征可以明示或者隐含地包括至少一个该特征。此外,“多个”的含义是至少两个,例如两个,三个等,除非另有明确具体的限定。In the description of the present invention, the terms "first" and "second" are only used for the purpose of description, and cannot be understood as indicating or implying relative importance or implying the number of indicated technical features. Thus, a feature delimited with "first", "second" may expressly or implicitly include at least one of that feature. Furthermore, "plurality" means at least two, eg, two, three, etc., unless expressly specifically defined otherwise.
在本发明的描述中,提及“实施例”意味着,结合实施例描述的特定特征、结构或特性可以包含在本发明的至少一个实施例中。在说明书中的各个位置出现该短语并不一定均是指相同的实施例,也不是与其它实施例互斥的独立的或备选的实施例。本领域技术人员显式地和隐式地理解的是,所描述的实施例可以与其它实施例相结合。In the description of the present invention, reference to "an embodiment" means that a particular feature, structure or characteristic described in connection with the embodiment can be included in at least one embodiment of the present invention. The appearances of the phrase in various places in the specification are not necessarily all referring to the same embodiment, nor a separate or alternative embodiment that is mutually exclusive of other embodiments. It is explicitly and implicitly understood by those skilled in the art that the described embodiments may be combined with other embodiments.
本发明提供了一种轨道交通健康监测系统及方法,通过持久化数据样本信息的采集,利用神经网络进行分类,对轨道结构健康和车身健康进行全面地判断,为进一步提高对轨道交通健康状态监测的准确性和高效性提供了新思路。The present invention provides a rail transit health monitoring system and method. Through the collection of persistent data sample information, the neural network is used for classification, and the rail structure health and vehicle body health are comprehensively judged, so as to further improve the monitoring of rail transit health status. The accuracy and efficiency of the system provide new ideas.
在实施例描述之前,对相关词语进行释义:Before the description of the embodiment, the related words are explained:
光纤传感器:是一种将被测对象的状态转变为可测的光信号的传感器。光纤传感器的工作原理是将光源入射的光束经由光纤送入调制器,在调制器内与外界被测参数的相互作用,使光的光学性质如光的强度、波长、频率、相位、偏振态等发生变化,成为被调制的光信号,再经过光纤送入光电器件、经解调器后获得被测参数。整个过程中,光束经由光纤导入,通过调制器后再射出,其中光纤的作用首先是传输光束,其次是起到光调制器的作用;Optical fiber sensor: It is a sensor that converts the state of the measured object into a measurable light signal. The working principle of the optical fiber sensor is to send the light beam incident from the light source into the modulator through the optical fiber, and interact with the external measured parameters in the modulator to make the optical properties of the light such as light intensity, wavelength, frequency, phase, polarization state, etc. After the change, it becomes a modulated optical signal, which is then sent to the optoelectronic device through the optical fiber, and the measured parameter is obtained after the demodulator. In the whole process, the light beam is introduced through the optical fiber, and then exits after passing through the modulator. The function of the optical fiber is to transmit the light beam first, and then to play the role of the optical modulator;
光纤阵列:是利用V形槽(即V槽,V-Groove)基片,把一束光纤或一条光纤带按照规定间隔安装在基片上,所构成的阵列;Optical fiber array: an array formed by using a V-groove (ie, V-groove, V-Groove) substrate to install a bundle of optical fibers or an optical fiber ribbon on the substrate at specified intervals;
光缆:为了满足光学、机械或环境的性能规范而制造的,它是利用置于包覆护套中的一根或多根光纤作为传输媒质并可以单独或成组使用的通信线缆组件。Optical cable: Manufactured to meet optical, mechanical, or environmental performance specifications, it is a communication cable assembly that utilizes one or more optical fibers in a sheathing as the transmission medium and can be used individually or in groups.
基于上述技术名词的描述,现有技术中,光纤传感器对车轮健康状态的监测存在可靠性不高、容易造成误报的问题,且系统信噪比不高,容易造成系统响应时间过长,因而,本发明旨在提出一种快速且高效的轨道交通健康监测方法。Based on the description of the above technical terms, in the prior art, the monitoring of the health status of the wheel by the fiber optic sensor has the problems of low reliability and easy to cause false alarms, and the system signal-to-noise ratio is not high, which is easy to cause the system response time to be too long. , the present invention aims to provide a fast and efficient rail transit health monitoring method.
以下分别对具体实施例进行详细说明:Specific embodiments are described in detail below:
本发明实施例提供了一种轨道交通安全监测系统,结合图1来看,图1为本发明提供的轨道交通安全监测系统一实施例的流程示意图,包括:依次通信连接的振动传感网络1、数据处理中心2以及智能化运维服务平台3,其中:An embodiment of the present invention provides a rail transit safety monitoring system. Referring to FIG. 1 , FIG. 1 is a schematic flowchart of an embodiment of the rail transit safety monitoring system provided by the present invention, including: a
所述振动传感网络1,包括布设于轨道交通工程结构上的光纤光栅振动传感光缆11;The
所述数据处理中心2,用于对所述光纤光栅振动传感光缆、所述光纤光栅应变传感光缆以及所述光纤光栅阵列温度传感光缆的监测信号进行处理,确定传感数据;The
所述智能化运维服务平台3,用于根据所述传感数据进行智能分析,基于智能分析结果,对轨道交通的工程结构和列车状态进行监测。The intelligent operation and
在本发明实施例中,通过设置振动传感网络,利用光纤光栅振动传感光缆、保证监测振动信息的准确性,实现全面的监测;通过设置数据处理中心,对上述多种光缆的监测信息进行数据转化,转化成传感数据;通过设置智能化运维服务平台,对传感数据进行相应的智能分析,从智能分析的结果,有效反馈轨道交通的工程结构和列车状态,以此保证对列车实时状态和工程结构的运行情况的快速监测,并及时发出相应的报警处理。In the embodiment of the present invention, by setting up a vibration sensing network and using fiber grating vibration sensing optical cables to ensure the accuracy of monitoring vibration information, comprehensive monitoring is achieved; by setting up a data processing center, the monitoring information of the above-mentioned various optical cables is monitored. Data is transformed into sensor data; by setting up an intelligent operation and maintenance service platform, corresponding intelligent analysis is carried out on the sensor data. From the results of intelligent analysis, the engineering structure and train status of rail transit can be effectively fed back, so as to ensure the safety of trains. Real-time status and rapid monitoring of the operation of the engineering structure, and timely issue of corresponding alarm processing.
作为优选的实施例,所述振动传感网络中的光纤光栅阵列都由光纤光栅阵列传感探头制成,所述光纤光栅阵列传感探头包括预设反射率光栅阵列,单根光纤上复用预设规模的光纤光栅传感测点。As a preferred embodiment, the fiber grating arrays in the vibration sensing network are all made of fiber grating array sensing probes, and the fiber grating array sensing probes include a preset reflectivity grating array, which is multiplexed on a single fiber. Fiber Bragg grating sensing points of preset scale.
在本发明实施例中,利用光纤光栅阵列传感探头保证光纤光栅的复用容量,仅需敷设数根传感光缆组成的传感网络即可覆盖数十公里的轨道交通线路,实现全域无盲区的智能感知,便于监测信号的全面采集。In the embodiment of the present invention, the fiber grating array sensing probe is used to ensure the multiplexing capacity of the fiber grating, and it is only necessary to lay a sensing network composed of several sensing optical cables to cover tens of kilometers of rail transit lines, so as to achieve no blind spots in the whole area. The intelligent perception facilitates the comprehensive collection of monitoring signals.
其中,预设反射率光栅阵列优选为极弱反射率光栅阵列,每根光纤探头上的光纤光栅为等间距布设,反射带宽为2~3nm,反射率一致性好,依据单根光纤上光纤光栅的复用容量,反射率范围为-30~-50dB,反射率越低单根光纤上复用容量越大。Among them, the preset reflectivity grating array is preferably a very weak reflectivity grating array, the fiber gratings on each fiber probe are arranged at equal intervals, the reflection bandwidth is 2-3 nm, and the reflectivity consistency is good. According to the fiber grating on a single fiber The reflectivity ranges from -30 to -50dB. The lower the reflectivity, the greater the multiplexing capacity on a single fiber.
其中,预设规模的光纤光栅传感测点优选为超大规模的光纤光栅传感测点,便于采集多种监测信号。Among them, the fiber grating sensing measuring point of the preset scale is preferably an ultra-large-scale fiber grating sensing measuring point, which is convenient for collecting various monitoring signals.
作为优选的实施例,每根所述光纤光栅阵列传感探头上的光纤光栅为等间距布设,以等间距的光纤光栅为节点,以相邻的节点形成振动感知单元。As a preferred embodiment, the fiber gratings on each of the fiber grating array sensing probes are arranged at equal intervals, and the equally spaced fiber gratings are used as nodes, and adjacent nodes are used to form vibration sensing units.
在本发明实施例中,利用相邻的节点形成高灵敏度干涉型振动感知单元,使得拾振器可感知微小振动异常信息,保证了系统探测的灵敏性。In the embodiment of the present invention, a high-sensitivity interference-type vibration sensing unit is formed by using adjacent nodes, so that the vibration pickup can sense the abnormal information of small vibration, and the sensitivity of the system detection is ensured.
作为更具体的实施例,以等间距的光纤光栅为节点,以相邻的节点形成高灵敏度干涉型振动感知单元。其中,所述光纤光栅阵列节点间距可调。As a more specific embodiment, fiber gratings with equal spacing are used as nodes, and adjacent nodes are used to form a high-sensitivity interference-type vibration sensing unit. Wherein, the node spacing of the fiber grating array is adjustable.
特殊地,本发明节点间距l设定大于等于3米。特别地以节点间距5米为例,每个拾振器能感知的最小相位差Δφ<10mrad,依据公式:In particular, the node spacing l of the present invention is set to be greater than or equal to 3 meters. In particular, taking the node spacing of 5 meters as an example, the minimum phase difference Δφ<10mrad that each vibration pickup can perceive, according to the formula:
式中:Δφ-相邻光纤光栅节点间光信号相位差;λ-光信号波长;Δl-相邻光栅节点间光栅长度的变化量;In the formula: Δφ - optical signal phase difference between adjacent fiber grating nodes; λ - optical signal wavelength; Δl - grating length variation between adjacent grating nodes;
其中,每个拾振单元可感知的最小应变量<50nε,从而可感知轮轨耦合激励下由于车辆和轨道的劣化或损伤所造成微小振动异常信息,显著提升系统探测能力。Among them, the minimum perceptible strain of each vibration pickup unit is less than 50nε, which can sense the abnormal information of micro vibration caused by the deterioration or damage of the vehicle and the track under the excitation of the wheel-rail coupling, which significantly improves the detection capability of the system.
在本发明实施例中,利用相邻的节点形成高灵敏度干涉型振动感知单元,使得拾振器可感知微小振动异常信息,保证了系统探测的灵敏性。In the embodiment of the present invention, a high-sensitivity interference-type vibration sensing unit is formed by using adjacent nodes, so that the vibration pickup can sense the abnormal information of small vibration, and the sensitivity of the system detection is ensured.
作为优选的实施例,所述振动传感网络中的振动传感光缆与被测结构密贴,其中:As a preferred embodiment, the vibration sensing optical cable in the vibration sensing network is closely attached to the measured structure, wherein:
若所述被测结构属于既有线路,则直接用结构胶粘贴于所述被测结构,或采用开槽后放入振动传感光缆后再利用水泥或固化剂将振动传感光缆与所述被测结构固化;If the structure to be tested belongs to an existing line, directly stick it to the structure under test with structural adhesive, or use a slot and then put the vibration sensing optical fiber cable, and then use cement or curing agent to connect the vibration sensing optical fiber cable to the tested structure. The tested structure is cured;
若所述被测结构属于新修线路,直接将振动传感光缆浇筑在混凝土中。If the structure under test belongs to a newly repaired line, directly pour the vibration sensing optical cable into the concrete.
在本发明实施例中,通过不同情况的布设方法,将被测结构与振动传感光缆进行高效的固化。In the embodiment of the present invention, the structure under test and the vibration sensing optical cable are efficiently cured through different layout methods.
作为优选的实施例,所述数据处理中心包括解调仪表、采集数据网络平台,其中:As a preferred embodiment, the data processing center includes a demodulation instrument and a network platform for collecting data, wherein:
所述解调仪表,用于将所述光纤光栅振动传感光缆、所述光纤光栅应变传感光缆以及所述光纤光栅阵列温度传感光缆的所述监测信号转化为数字信号;The demodulation instrument is used to convert the monitoring signals of the fiber grating vibration sensing optical cable, the fiber grating strain sensing optical cable, and the fiber grating array temperature sensing optical cable into digital signals;
所述采集数据网络平台,用于实时采集所述数字信号,转化为所述传感数据上传至所述智能化运维服务平台。The data collection network platform is used to collect the digital signal in real time, convert it into the sensor data and upload it to the intelligent operation and maintenance service platform.
在本发明实施例中,采用计算机软硬件及网络平台具备信息采集功能,将光纤光栅阵列传感光缆接入光纤光栅阵列传感解调仪表,传感光缆感知的光信号经解调仪表转化为数字信号,利用计算机软硬件及网络平台实时采集传感网络中每个感知单元的监测数据,实现轨道交通工程结构全时全域的检测。In the embodiment of the present invention, the computer software and hardware and the network platform are used to have the function of information collection, the fiber grating array sensing optical cable is connected to the fiber grating array sensing demodulation instrument, and the optical signal sensed by the sensing optical cable is converted into the demodulation instrument through the demodulation instrument. Digital signal, using computer software, hardware and network platform to collect real-time monitoring data of each sensing unit in the sensor network, to realize the full-time and all-domain detection of rail transit engineering structures.
作为优选的实施例,结合图2来看,图2为本发明提供的轨道交通健康监测方法一实施例的流程示意图,包括步骤S201至步骤S203,其中:As a preferred embodiment, with reference to FIG. 2, FIG. 2 is a schematic flowchart of an embodiment of a rail transit health monitoring method provided by the present invention, including steps S201 to S203, wherein:
在步骤S201中,获取响应数据;In step S201, obtain response data;
在步骤S202中,根据所述响应数据进行预处理,确定待测振动数据;In step S202, preprocessing is performed according to the response data to determine vibration data to be measured;
在步骤S203中,将所述待测振动数据输入至训练完备的第一神经网络和/或第二神经网络,确定预测轨道结构健康类别和预测车身健康类别。In step S203, the vibration data to be measured is input into the well-trained first neural network and/or the second neural network, and the predicted track structure health category and the predicted vehicle body health category are determined.
在本发明实施例中,首先,对响应数据进行有效的获取;进而,通过对响应数据的预处理,提取出相关有效的数据,形成待测振动数据,保证待测振动数据的准确性;最后,通过第一神经网络和/或第二神经网络,对待测振动数据进行有效的识别和判断,提取其中的特征信息,预测轨道结构健康类别和车身健康类别,保证了快速高效地基于振动信息,实现多方面的交通情况判别。In the embodiment of the present invention, first, the response data is effectively obtained; then, through the preprocessing of the response data, relevant valid data is extracted to form the vibration data to be measured, so as to ensure the accuracy of the vibration data to be measured; finally , through the first neural network and/or the second neural network, to effectively identify and judge the vibration data to be measured, extract the feature information, and predict the track structure health category and body health category, which ensures fast and efficient vibration information based on Realize multi-faceted traffic situation discrimination.
作为更具体的实施例,超弱反射率光纤光栅振动传感光缆沿轨道结构通长敷设,轨道线路被光纤光栅节点划分为N个区间:As a more specific embodiment, the ultra-weak reflectivity fiber grating vibration sensing optical cable is laid along the track structure, and the track line is divided into N intervals by the fiber grating node:
上式中,L为轨道线路长度,l为光纤光栅间距。In the above formula, L is the track line length, and l is the fiber grating spacing.
其中,本发明能实时采集所有区间上N个拾振器的信号,采样频率≥1000Hz,但是本发明中利用的是轮轨耦合激励的数据,实时采集的绝大部分是无用数据,徒增存储容量,浪费系统资源。为了便于数据分析和数据管理,减少无效数据十分有必要。一般地,一个持久化的数据样本包括如下信息:区间标记i(1≤i≤N);列车接近区间i的时刻t0;列车驶离区间i的时刻t1;t0和t1时间段内区间i拾振器连续采集的振动数据。Among them, the present invention can collect the signals of N vibration pickups in all intervals in real time, and the sampling frequency is ≥ 1000 Hz, but the data of the wheel-rail coupling excitation is used in the present invention, and most of the real-time collected data are useless data, and the storage is increased. capacity, wasting system resources. In order to facilitate data analysis and data management, it is necessary to reduce invalid data. Generally, a persistent data sample includes the following information: section mark i (1≤i≤N); time t0 when the train approaches section i; time t1 when the train leaves section i; Vibration data collected continuously by the vibrator.
一般地,将可选取区间i-k的拾振器检测到有车驶过时记为t0,区间i+k的拾振器检测到有车驶过时记为t1,其中:Generally, the time when the vibrator in the range i-k can be selected to detect a car passing by is recorded as t0, and the time when the vibrator in the interval i+k detects a car passing by is recorded as t1, where:
上式中,Lc为列车长度。In the above formula, L c is the train length.
作为优选的实施例,结合图3来看,图3为本发明提供的图2中步骤S202一实施例的流程示意图,步骤S202包括步骤S301至步骤S302,其中:As a preferred embodiment, referring to FIG. 3 , FIG. 3 is a schematic flowchart of an embodiment of step S202 in FIG. 2 provided by the present invention. Step S202 includes steps S301 to S302, wherein:
在步骤S301中,将所述振动数据的前后两端添加环境噪声信号,确定填充数据;In step S301, an environmental noise signal is added to the front and rear ends of the vibration data to determine filling data;
在步骤S302中,将所述填充数据进行数据归一化处理,确定所述待测数据。In step S302, data normalization is performed on the filling data to determine the data to be measured.
在本发明实施例中,利用数据填充和数据归一化处理,保证待测振动数据的准确性,以进一步保证识别结果的准确性。In the embodiment of the present invention, data filling and data normalization are used to ensure the accuracy of the vibration data to be measured, so as to further ensure the accuracy of the identification result.
作为优选的实施例,结合图4来看,图4为本发明提供的第一神经网络和/或第二神经网络的训练过程一实施例的流程示意图,包括步骤S401至步骤S404,其中:As a preferred embodiment, referring to FIG. 4, FIG. 4 is a schematic flowchart of an embodiment of the training process of the first neural network and/or the second neural network provided by the present invention, including steps S401 to S404, wherein:
在步骤S401中,当将样本集输入至所述第一神经网络时,输出预测轨道结构健康类别,其中,所述样本集包括样本振动数据与对应的实际轨道结构健康类别;In step S401, when a sample set is input to the first neural network, a predicted track structure health category is output, wherein the sample set includes sample vibration data and a corresponding actual track structure health category;
在步骤S402中,当将样本集输入至所述第二神经网络时,输出预测车身健康类别,其中,所述样本集包括所述样本振动数据与对应的实际车身健康类别;In step S402, when a sample set is input to the second neural network, a predicted body health category is output, wherein the sample set includes the sample vibration data and the corresponding actual body health category;
在步骤S403中,根据所述预测轨道结构健康类别和所述实际轨道结构健康类别之间的误差,确定第一损失函数,并根据所述第一损失函数训练构建的第一神经网络至收敛;In step S403, a first loss function is determined according to the error between the predicted track structure health category and the actual track structure health category, and the constructed first neural network is trained to converge according to the first loss function;
在步骤S404中,根据所述预测车身健康类别和所述实际车身健康类别之间的误差,确定第二损失函数,并根据所述第二损失函数训练构建的第二神经网络至收敛。In step S404, a second loss function is determined according to the error between the predicted body health category and the actual body health category, and the constructed second neural network is trained to converge according to the second loss function.
在本发明实施例中,利用样本集对第一神经网络和第二神经网络进行有效地训练,完成对轨道结构健康类别和车身健康类别的有效识别。In the embodiment of the present invention, the first neural network and the second neural network are effectively trained by using the sample set, and the effective identification of the track structure health category and the vehicle body health category is completed.
作为优选的实施例,结合图5来看,图5为本发明提供的第一神经网络一实施例的结构示意图,所述第一神经网络依次包括:输入层、卷积层、最大池化层、Flatten层、全连接层和输出层,其中:As a preferred embodiment, referring to FIG. 5 , FIG. 5 is a schematic structural diagram of an embodiment of the first neural network provided by the present invention. The first neural network sequentially includes: an input layer, a convolution layer, and a maximum pooling layer , Flatten layer, fully connected layer and output layer, where:
所述输入层,用于输入所述样本集;the input layer, for inputting the sample set;
所述卷积层,用于对所述样本集进行卷积操作,确定第一卷积数据;The convolution layer is used to perform a convolution operation on the sample set to determine the first convolution data;
所述最大池化层,包括池化层和Batch Normalization层,其中,所述池化层用于对所述第一卷积数据进行特征提取和降维处理,所述Batch Normalization层用于对降维后的第一卷积数据进行标准化处理,确定标准化数据;The maximum pooling layer includes a pooling layer and a Batch Normalization layer, wherein the pooling layer is used to perform feature extraction and dimension reduction processing on the first convolutional data, and the Batch Normalization layer is used to reduce the Standardize the first convolution data after dimensioning to determine standardized data;
所述Flatten层,用于将所述标准化数据展开成一维向量;The Flatten layer is used to expand the standardized data into a one-dimensional vector;
所述全连接层,用于将所述一维向量降维,并通过softmax函数进行分类,确定所述预测轨道结构健康类别。The fully connected layer is used for reducing the dimension of the one-dimensional vector, and classifying it through a softmax function to determine the predicted track structure health category.
在本发明实施例中,构建第一神经网络,以实现对样本集的特征提取,并确定与实际轨道结构健康类别的对应关系,完成网络收敛,以进行对车身健康类别的有效预测。In the embodiment of the present invention, a first neural network is constructed to extract the features of the sample set, determine the corresponding relationship with the actual track structure health category, complete the network convergence, and effectively predict the vehicle body health category.
下面以一个具体的实施例说明轨道结构健康评估算法流程如下:The following describes the algorithm flow of the track structure health assessment with a specific embodiment as follows:
第一步,地铁信号采集:将采集到的数据存入数据库;The first step, subway signal collection: save the collected data into the database;
第二步,由于地铁的速度不一致,在进站和出站的时候,地铁速度较慢,在中途运行过程中运行速度较快,因此得到的信号长短会不一致,因此需要对持续时间较短的信号进行数据填充,在信号前后两端添加环境噪声,使其长度一致,均为729个数据点;In the second step, due to the inconsistent speed of the subway, the subway speed is slower when entering and exiting the station, and the speed is faster during the middle operation, so the length of the obtained signal will be inconsistent, so it is necessary to adjust the duration of the shorter time. The signal is filled with data, and environmental noise is added at the front and rear ends of the signal to make the length the same, with 729 data points;
第三步,数据归一化:由于得到的信号幅值不一致,参差不齐,不利于神经网络的训练和收敛,因此有必要对数据进行归一化预处理。目前常用的归一化方式有min-max标准化和Z-score标准化两种方法。本发明实施例采用Z-score标准化,其具体公式为:The third step, data normalization: Since the obtained signal amplitudes are inconsistent and uneven, which is not conducive to the training and convergence of the neural network, it is necessary to normalize the data preprocessing. At present, the commonly used normalization methods include min-max normalization and Z-score normalization. The embodiment of the present invention adopts Z-score standardization, and its specific formula is:
其中,μ为所有样本数据的均值,σ为所有样本数据的标准差。经过处理的数据符合标准正态分布,即均值为0,标准差为1。Among them, μ is the mean of all sample data, and σ is the standard deviation of all sample data. The processed data conformed to a standard normal distribution, that is, a mean of 0 and a standard deviation of 1.
第四步,神经网络训练并预测:得到的数据按照9:1的比例分为训练集和测试集。卷积神经网络主要用于轨道健康状况的检测。主要分类为以下六类:(1)轨道健康;(2)轨道断轨;(3)轨道波磨;(4)扣件松动;(5)弹条断裂;(6)线下基础病害。对于根据大数据来预测结果的问题,卷积神经网络往往表现出较好的结果。卷积神经网络根据使用对象,又可以分为一维卷积神经网络和二维卷积神经网络等。二维卷积神经网络主要用于图像的识别和分类。而针对文本类数据,一维卷积神经网络,往往更加适合。因此本发明实施例采用一维卷积神经网络,训练集用于对神经网络的训练,通过损失函数的调节和反向传播原理,不断对模型进行优化,直至达到良好的性能;测试集用于评估模型的泛化能力和轨道预测评估能力;The fourth step, neural network training and prediction: the obtained data is divided into training set and test set according to the ratio of 9:1. Convolutional Neural Networks are mainly used for the detection of track health. Mainly classified into the following six categories: (1) track health; (2) track broken; (3) track corrugation; (4) loose fasteners; (5) spring clip fracture; (6) offline basic diseases. For the problem of predicting outcomes from big data, convolutional neural networks tend to show better results. Convolutional neural network can be divided into one-dimensional convolutional neural network and two-dimensional convolutional neural network according to the object of use. Two-dimensional convolutional neural networks are mainly used for image recognition and classification. For textual data, one-dimensional convolutional neural networks are often more suitable. Therefore, the embodiment of the present invention adopts a one-dimensional convolutional neural network, and the training set is used for training the neural network. Through the adjustment of the loss function and the principle of back propagation, the model is continuously optimized until good performance is achieved; the test set is used for the training of the neural network. Evaluate the generalization ability and orbit prediction evaluation ability of the model;
其中,结合图5来看,主要由输入层,卷积层(Conv1D),池化层(Pooling),全连接层(FC),输出层组成。其中地铁振动信号作为输入张量,送入神经网络进行训练。首先经过一维卷积层,卷积核尺寸设置为9,卷积核数量设置为32,数据张量经过卷积层之后,进入池化层。池化层,是对原数据进行特征提取和降维处理,能够大大减少数据的维度,加快模型的训练和收敛。池化层有平均池化和最大池化两种,在实践中,最大池化层往往能够取得较好的效果,因此本发明实施例拟采用最大池化层。为了防止模型训练的过拟合,在每个池化层之后,都跟着一个Batch Normalization层,它能够再一次对数据进行标准化,加快模型收敛和防止过拟合。经过三轮卷积和池化操作后,通过Flatten操作,将二维张量展开成一维向量,通过全连接层降维最后通过softmax函数分成六类,输出分类结果,用于对轨道健康状况的评估,其具体参数设置见下表1:Among them, combined with Figure 5, it is mainly composed of an input layer, a convolution layer (Conv1D), a pooling layer (Pooling), a fully connected layer (FC), and an output layer. The subway vibration signal is used as the input tensor and sent to the neural network for training. First go through the one-dimensional convolution layer, the size of the convolution kernel is set to 9, the number of convolution kernels is set to 32, and the data tensor enters the pooling layer after passing through the convolution layer. The pooling layer is to perform feature extraction and dimensionality reduction processing on the original data, which can greatly reduce the dimension of the data and speed up the training and convergence of the model. There are two types of pooling layers: average pooling and maximum pooling. In practice, the maximum pooling layer can often achieve better results. Therefore, the embodiment of the present invention intends to use the maximum pooling layer. In order to prevent overfitting of model training, after each pooling layer, there is a Batch Normalization layer, which can standardize the data again, speed up model convergence and prevent overfitting. After three rounds of convolution and pooling operations, the two-dimensional tensor is expanded into a one-dimensional vector through the Flatten operation, and the dimension is reduced by the fully connected layer and finally divided into six categories by the softmax function. Evaluation, its specific parameter settings are shown in Table 1 below:
表1Table 1
其中,要想对神经网络进行训练,首先得对它的训练过程进行编译,其中有三个重要的参数,即优化器,损失函数,和评价指标。本发明拟采用的优化器是Adam优化器,之所以采用它,是因为能够根据不同的参数自适应地调整学习率,更加适用于大规模的数据及参数的场景。损失函数采用多元交叉熵函数,是因为本问题是多分类问题,更加适合多元交叉熵函数。其具体公式如下所示:Among them, in order to train a neural network, it is necessary to compile its training process first. There are three important parameters, namely optimizer, loss function, and evaluation index. The optimizer to be used in the present invention is the Adam optimizer, which is used because the learning rate can be adjusted adaptively according to different parameters, and is more suitable for large-scale data and parameter scenarios. The loss function adopts the multivariate cross entropy function, because this problem is a multi-classification problem, which is more suitable for the multivariate cross entropy function. Its specific formula is as follows:
上式中,y是真实标签,是预测标签;In the above formula, y is the true label, is the predicted label;
至于评价指标,选择准确率作为对模型分类性能的评价指标。准确率是指分类正确的样本数量占总数量之比;As for the evaluation index, the accuracy rate is selected as the evaluation index for the classification performance of the model. Accuracy refers to the ratio of the number of correctly classified samples to the total number;
第五步,输出结果,上位机显示:神经网络预测出结果后,将结果显示在上位机界面上,相关人员可以参考预测结果,及时了解轨道的健康状况,及时加以处理,以减少意外的发生。The fifth step is to output the result and display it on the host computer: after the neural network predicts the result, the result is displayed on the interface of the host computer. The relevant personnel can refer to the predicted result, understand the health status of the track in time, and deal with it in time to reduce the occurrence of accidents .
作为优选的实施例,结合图6来看,图6为本发明提供的BP神经网络一实施例的结构示意图,所述第二神经网络依次包括:BP神经网络和一维神经网络,其中:As a preferred embodiment, referring to FIG. 6 , FIG. 6 is a schematic structural diagram of an embodiment of a BP neural network provided by the present invention. The second neural network sequentially includes: a BP neural network and a one-dimensional neural network, wherein:
所述BP神经网络,对速度训练集训练,用于输出预测车速,其中,所述速度训练集包括样本振动数据与对应的速度标签;The BP neural network is trained on a speed training set for outputting a predicted vehicle speed, wherein the speed training set includes sample vibration data and corresponding speed labels;
所述一维神经网络,用于依次通过输入层、卷积层、最大池化层、Flatten层、全连接层和输出层对所述样本集进行训练,确定所述预测车身健康类别。The one-dimensional neural network is used to sequentially train the sample set through an input layer, a convolutional layer, a maximum pooling layer, a Flatten layer, a fully connected layer and an output layer to determine the predicted body health category.
在本发明实施例中,构建第二神经网络,以实现对样本集的特征提取,并确定与实际车速、实际车身健康类别的对应关系,完成网络收敛,以进行对车身健康类别的有效预测。In the embodiment of the present invention, a second neural network is constructed to extract the features of the sample set, determine the corresponding relationship with the actual vehicle speed and the actual vehicle body health category, and complete the network convergence to effectively predict the vehicle body health category.
作为优选的实施例,在所述第一神经网络和/或所述第二神经网络的训练过程中,采用Adam优化器自适应地调整学习率,且所述第一损失函数和所述第二损失函数为多元交叉熵函数。As a preferred embodiment, in the training process of the first neural network and/or the second neural network, the Adam optimizer is used to adaptively adjust the learning rate, and the first loss function and the second The loss function is a multivariate cross-entropy function.
在本发明实施例中,利用Adam优化器和多元交叉熵函数,保证训练过程的快速性和有效性。In the embodiment of the present invention, the Adam optimizer and the multivariate cross-entropy function are used to ensure the rapidity and effectiveness of the training process.
下面以一个具体的实施例说明车辆健康评估算法流程如下:The following describes the process of the vehicle health assessment algorithm with a specific embodiment as follows:
第1步,地铁信号采集:将采集到的数据存入数据库,使其持久化;
第2步,数据填充:由于地铁的速度不一致,在进站和出站的时候,地铁速度较慢,在中途运行过程中运行速度较快,得到的信号长短会不一致,因此需要对持续时间较短的信号进行数据填充,在信号前后两端添加环境噪声,使其长度一致,均为729个数据点;
第3步,数据归一化:由于得到的信号幅值不一致,参差不齐,不利于神经网络的训练和收敛,因此有必要对数据进行归一化预处理。目前常用的归一化方式有min-max标准化和Z-score标准化两种方法。本发明实施例采用Z-score标准化,其具体公式为:The third step, data normalization: Since the obtained signal amplitudes are inconsistent and uneven, which is not conducive to the training and convergence of the neural network, it is necessary to normalize the data preprocessing. At present, the commonly used normalization methods include min-max normalization and Z-score normalization. The embodiment of the present invention adopts Z-score standardization, and its specific formula is:
其中,μ为所有样本数据的均值,σ为所有样本数据的标准差。经过处理的数据符合标准正态分布,即均值为0,标准差为1。Among them, μ is the mean of all sample data, and σ is the standard deviation of all sample data. The processed data conformed to a standard normal distribution, that is, a mean of 0 and a standard deviation of 1.
第4步,神经网络预测:本模块分为两部分,一部分是BP神经网络,另一部分是一维卷积神经网络。BP神经网络结构如图6所示,用于对地铁车速的检测。一维卷积神经网络如图5所示,用于车辆状态的分类。对于地铁车速检测,是一个典型的标量回归问题,利用BP神经网络的非线性学习能力来搭建模型,使用事先人工标记好车速标签的训练集样本进行训练,然后用于车速预测。本模型由输入层,隐藏层,输出层组成。输入层为之前提取的地铁振动信号,隐藏层为32个神经元,每个神经元之后都跟着一个Relu激活函数,最后为一个神经元的输出层(无激活函数),用于输出预测的地铁车速:Step 4, neural network prediction: This module is divided into two parts, one part is BP neural network, and the other part is one-dimensional convolutional neural network. The structure of BP neural network is shown in Figure 6, which is used to detect the subway speed. A one-dimensional convolutional neural network is shown in Figure 5 for the classification of vehicle states. For subway speed detection, it is a typical scalar regression problem. The nonlinear learning ability of the BP neural network is used to build a model, and the training set samples that have been manually marked with speed labels are used for training, and then used for speed prediction. This model consists of an input layer, a hidden layer, and an output layer. The input layer is the previously extracted subway vibration signal, the hidden layer is 32 neurons, each neuron is followed by a Relu activation function, and the last is the output layer of a neuron (without activation function), which is used to output the predicted subway Speed:
其中,每个隐藏层的神经元的输出为:Among them, the output of each hidden layer neuron is:
f(x)=σ(∑wx+b)f(x)=σ(∑wx+b)
其中,w为权重,b为偏差,σ为激活函数,此处为Relu激活函数,其具体公式为:Among them, w is the weight, b is the deviation, σ is the activation function, here is the Relu activation function, and its specific formula is:
f(x)=max(0,x)f(x)=max(0,x)
其中,之所以采用激活函数,是为了增强数据的非线性,更加有利于模型的训练;Among them, the reason why the activation function is used is to enhance the nonlinearity of the data, which is more conducive to the training of the model;
其中,由于为回归问题,因此损失函数选择为均方差函数(MSE,mean squarederror),评价指标选择平均绝对误差(MAE,mean absolute error),其公式分别为:Among them, because it is a regression problem, the loss function is selected as the mean squared error (MSE, mean squarederror), and the evaluation index is selected as the mean absolute error (MAE, mean absolute error). The formulas are:
其中,优化器采用Adam优化器,经过150次训练之后,模型验证集收敛到0.7的平均绝对误差,可以较准确地用于地铁车速的预测;Among them, the optimizer adopts the Adam optimizer. After 150 times of training, the model validation set converges to an average absolute error of 0.7, which can be more accurately used for the prediction of subway speed;
其中,一维卷积神经网络主要用于地铁车身健康状况的检测。主要分类为以下四类:Among them, the one-dimensional convolutional neural network is mainly used for the detection of the health status of the subway body. Mainly classified into the following four categories:
(1)车辆健康;(2)车轮扁疤缺陷;(3)车轮偏心;(4)车轮松动。对于根据大数据来预测结果的问题,卷积神经网络往往表现出较好的结果。卷积神经网络根据使用对象,又可以分为一维卷积神经网络和二维卷积神经网络等。二维卷积神经网络主要用于图像的识别和分类。而针对文本类数据,一维卷积神经网络,往往更加适合;(1) The vehicle is healthy; (2) The wheel flat scar is defective; (3) The wheel is eccentric; (4) The wheel is loose. For the problem of predicting outcomes from big data, convolutional neural networks tend to show better results. Convolutional neural network can be divided into one-dimensional convolutional neural network and two-dimensional convolutional neural network according to the object of use. Two-dimensional convolutional neural networks are mainly used for image recognition and classification. For textual data, one-dimensional convolutional neural networks are often more suitable;
其中,一维卷积神经网络结构如图5所示,其结构、训练过程的编译、损失函数的选择与上述第一神经网络一致,但分类标签有所不同,通过全连接层降维最后通过softmax函数分成四类,输出分类结果,用于对地铁健康状况的评估,在此不再赘述;Among them, the structure of the one-dimensional convolutional neural network is shown in Figure 5. Its structure, the compilation of the training process, and the selection of the loss function are the same as those of the first neural network above, but the classification labels are different. The softmax function is divided into four categories, and the output classification results are used to evaluate the health status of the subway, which will not be repeated here;
第5步,输出结果,上位机显示:神经网络预测出结果后,将结果显示在上位机界面上,相关人员可以参考预测结果,及时了解地铁的健康状况和运行车速,以减少意外的发生。Step 5: Output the result and display it on the host computer: After the neural network predicts the result, the result is displayed on the host computer interface, and the relevant personnel can refer to the forecast result to know the health status and running speed of the subway in time to reduce the occurrence of accidents.
本发明实施例还提供了一种轨道交通健康监测装置,结合图7来看,图7为本发明提供的轨道交通健康监测装置一实施例的结构示意图,轨道交通健康监测装置700包括:An embodiment of the present invention also provides a rail transit health monitoring device. Referring to FIG. 7 , FIG. 7 is a schematic structural diagram of an embodiment of the rail transit health monitoring device provided by the present invention. The rail transit
获取单元701,用于获取响应数据;an obtaining
处理单元702,用于根据所述响应数据进行预处理,确定待测振动数据;a
预测单元703,用于将所述待测振动数据输入至训练完备的第一神经网络和/或第二神经网络,确定预测轨道结构健康类别和预测车身健康类别。The
轨道交通健康监测装置的各个单元的更具体实现方式可以参见对于上述轨道交通健康监测方法的描述,且具有与之相似的有益效果,在此不再赘述。For a more specific implementation manner of each unit of the rail transit health monitoring device, reference may be made to the description of the above-mentioned rail transit health monitoring method, which has similar beneficial effects, and will not be repeated here.
本发明实施例还提供了一种计算机可读存储介质,其上存储有计算机程序,该程序被处理器执行时,实现如上所述的轨道交通健康监测方法。Embodiments of the present invention further provide a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, implements the above-mentioned rail transit health monitoring method.
一般来说,用于实现本发明方法的计算机指令的可以采用一个或多个计算机可读的存储介质的任意组合来承载。非临时性计算机可读存储介质可以包括任何计算机可读介质,除了临时性地传播中的信号本身。In general, computer instructions for implementing the methods of the present invention may be carried in any combination of one or more computer-readable storage media. A non-transitory computer-readable storage medium may include any computer-readable medium except for the temporarily propagated signal itself.
计算机可读存储介质例如可以是——但不限于——电、磁、光、电磁、红外线、或半导体的系统、装置或器件,或者任意以上的组合。计算机可读存储介质的更具体的例子(非穷举的列表)包括:具有一个或多个导线的电连接、便携式计算机磁盘、硬盘、随机存取存储器(RAM)、只读存储器(ROM)、可擦式可编程只读存储器(EPROM或闪存)、光纤、便携式紧凑磁盘只读存储器(CD-ROM)、光存储器件、磁存储器件、或者上述的任意合适的组合。在本发明实施例件中,计算机可读存储介质可以是任何包含或存储程序的有形介质,该程序可以被指令执行系统、装置或者器件使用或者与其结合使用。The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus or device, or a combination of any of the above. More specific examples (a non-exhaustive list) of computer readable storage media include: electrical connections having one or more wires, portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), Erasable programmable read only memory (EPROM or flash memory), optical fiber, portable compact disk read only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the above. In the embodiments of the present invention, the computer-readable storage medium may be any tangible medium that contains or stores a program, and the program may be used by or in combination with an instruction execution system, apparatus, or device.
可以以一种或多种程序设计语言或其组合来编写用于执行本发明操作的计算机程序代码,程序设计语言包括面向对象的程序设计语言—诸如Java、Smalltalk、C++,还包括常规的过程式程序设计语言—诸如“C”语言或类似的程序设计语言,特别是可以使用适于神经网络计算的Python语言和基于TensorFlow、PyTorch等平台框架。程序代码可以完全地在用户计算机上执行、部分地在用户计算机上执行、作为一个独立的软件包执行、部分在用户计算机上部分在远程计算机上执行、或者完全在远程计算机或服务器上执行。在涉及远程计算机的情形中,远程计算机可以通过任意种类的网络——包括局域网(LAN)或广域网(WAN)—连接到用户计算机,或者,可以连接到外部计算机(例如利用因特网服务提供商来通过因特网连接)。Computer program code for carrying out operations of the present invention may be written in one or more programming languages, including object-oriented programming languages—such as Java, Smalltalk, C++, but also conventional procedural languages, or a combination thereof. Programming language - such as "C" language or similar programming language, especially Python language suitable for neural network computing and platform frameworks based on TensorFlow, PyTorch, etc. can be used. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (eg, using an Internet service provider through Internet connection).
本发明实施例还提供了一种电子设备,结合图8来看,图8为本发明提供的电子设备一实施例的结构示意图,电子设备800包括处理器801、存储器802及存储在存储器802上并可在处理器801上运行的计算机程序,处理器801执行程序时,实现如上所述的轨道交通健康监测方法。An embodiment of the present invention also provides an electronic device. Referring to FIG. 8 , FIG. 8 is a schematic structural diagram of an embodiment of the electronic device provided by the present invention. The
作为优选的实施例,上述电子设备800还包括显示器803,用于显示处理器801执行如上所述的轨道交通健康监测方法。As a preferred embodiment, the above-mentioned
示例性的,计算机程序可以被分割成一个或多个模块/单元,一个或者多个模块/单元被存储在存储器802中,并由处理器801执行,以完成本发明。一个或多个模块/单元可以是能够完成特定功能的一系列计算机程序指令段,该指令段用于描述计算机程序在电子设备800中的执行过程。例如,计算机程序可以被分割成上述实施例中的获取单元701、处理单元702及预测单元703,各单元的具体功能如上所述,在此不一一赘述。Exemplarily, the computer program may be divided into one or more modules/units, and the one or more modules/units are stored in the
电子设备800可以是带可调摄像头模组的桌上型计算机、笔记本、掌上电脑或智能手机等设备。The
其中,处理器801可能是一种集成电路芯片,具有信号的处理能力。上述的处理器801可以是通用处理器,包括中央处理器(Central Processing Unit,CPU)、网络处理器(Network Processor,NP)等;还可以是数字信号处理器(Digital Signal Processor,DSP)、专用集成电路(Application Specific Integrated Circuit,ASIC)、现成可编程门阵列(Field-Programmable Gate Array,FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。可以实现或者执行本发明实施例中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。Wherein, the
其中,存储器802可以是,但不限于,随机存取存储器(Random Access Memory,RAM),只读存储器(Read Only Memory,ROM),可编程只读存储器(Programmable Read-OnlyMemory,PROM),可擦除只读存储器(Erasable Programmable Read-Only Memory,EPROM),电可擦除只读存储器(Electric Erasable Programmable Read-Only Memory,EEPROM)等。其中,存储器802用于存储程序,所述处理器801在接收到执行指令后,执行所述程序,前述本发明实施例任一实施例揭示的流程定义的方法可以应用于处理器801中,或者由处理器801实现。Wherein, the
其中,显示器803可以是LCD显示屏,也可以是LED显示屏。例如,手机上的显示屏。The
可以理解的是,图8所示的结构仅为电子设备800的一种结构示意图,电子设备800还可以包括比图8所示更多或更少的组件。图8中所示的各组件可以采用硬件、软件或其组合实现。It can be understood that the structure shown in FIG. 8 is only a schematic structural diagram of the
根据本发明上述实施例提供的计算机可读存储介质和电子设备,可以参照根据本发明实现如上所述的轨道交通健康监测方法具体描述的内容实现,并具有与如上所述的轨道交通健康监测方法类似的有益效果,在此不再赘述。The computer-readable storage medium and electronic device provided according to the above-mentioned embodiments of the present invention can be implemented with reference to the content specifically described in the implementation of the above-mentioned rail transit health monitoring method according to the present invention, and have the same characteristics as the above-mentioned rail transit health monitoring method. Similar beneficial effects are not repeated here.
本发明公开了一种轨道交通健康监测系统及方法,首先,对响应数据进行有效的获取;进而,通过对响应数据的预处理,提取出相关有效的数据,形成待测振动数据,保证待测振动数据的准确性;最后,通过第一神经网络和/或第二神经网络,对待测振动数据进行有效的识别和判断,提取其中的特征信息,预测轨道结构健康类别和车身健康类别,保证了快速高效地基于振动信息,实现多方面的交通情况判别。The invention discloses a rail transit health monitoring system and method. First, the response data is effectively acquired; further, through the preprocessing of the response data, relevant and effective data are extracted to form vibration data to be measured, so as to ensure the vibration data to be measured. The accuracy of the vibration data; finally, through the first neural network and/or the second neural network, the vibration data to be measured can be effectively identified and judged, the feature information in it is extracted, and the track structure health category and body health category are predicted, ensuring that Based on vibration information quickly and efficiently, it can realize multi-faceted traffic situation discrimination.
本发明技术方案,基于振动传感网络,实现高灵敏度地获取响应数据,将待测振动数据输入至相关的神经网络后,仅基于待测振动数据就可得到多方面交通运行信息,保证了算法的高效性和快速性,直接预测出结果后,相关人员可以参考预测结果,及时了解地铁的健康状况和运行车速,以减少意外的发生,保证了反馈的及时性。The technical scheme of the present invention is based on the vibration sensing network to achieve high-sensitivity acquisition of response data. After inputting the vibration data to be measured into the relevant neural network, various traffic operation information can be obtained only based on the vibration data to be measured, which ensures the algorithm After directly predicting the results, the relevant personnel can refer to the prediction results to know the health status and running speed of the subway in time, so as to reduce the occurrence of accidents and ensure the timeliness of feedback.
以上所述,仅为本发明较佳的具体实施方式,但本发明的保护范围并不局限于此,任何熟悉本技术领域的技术人员在本发明揭露的技术范围内,可轻易想到的变化或替换,都应涵盖在本发明的保护范围之内。The above description is only a preferred embodiment of the present invention, but the protection scope of the present invention is not limited to this. Substitutions should be covered within the protection scope of the present invention.
Claims (10)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210259899.3A CN114684217B (en) | 2022-03-16 | 2022-03-16 | Rail transit health monitoring system and method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210259899.3A CN114684217B (en) | 2022-03-16 | 2022-03-16 | Rail transit health monitoring system and method |
Publications (2)
Publication Number | Publication Date |
---|---|
CN114684217A true CN114684217A (en) | 2022-07-01 |
CN114684217B CN114684217B (en) | 2024-03-01 |
Family
ID=82138179
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210259899.3A Active CN114684217B (en) | 2022-03-16 | 2022-03-16 | Rail transit health monitoring system and method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN114684217B (en) |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115420476A (en) * | 2022-11-03 | 2022-12-02 | 高勘(广州)技术有限公司 | Optical cable-based railway health monitoring method, device, equipment and storage medium |
CN115808468A (en) * | 2022-12-08 | 2023-03-17 | 中铁第四勘察设计院集团有限公司 | Ballastless track disease rapid identification method based on vibration waveform |
CN116223763A (en) * | 2022-12-05 | 2023-06-06 | 湖北国际物流机场有限公司 | Method for monitoring health state of runway surface of airport runway landing zone |
CN118082914A (en) * | 2024-01-12 | 2024-05-28 | 通号(西安)轨道交通工业集团有限公司北京分公司 | Railway signal axle counting system |
CN119091641A (en) * | 2024-11-07 | 2024-12-06 | 武汉理工大学 | A vehicle continuous tracking method, device and electronic device based on grating array |
Citations (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106503642A (en) * | 2016-10-18 | 2017-03-15 | 长园长通新材料股份有限公司 | A kind of model of vibration method for building up for being applied to optical fiber sensing system |
CN108647786A (en) * | 2018-07-10 | 2018-10-12 | 电子科技大学 | The rotating machinery on-line fault monitoring method of neural network is fought based on depth convolution |
US20180357542A1 (en) * | 2018-06-08 | 2018-12-13 | University Of Electronic Science And Technology Of China | 1D-CNN-Based Distributed Optical Fiber Sensing Signal Feature Learning and Classification Method |
CN109446187A (en) * | 2018-10-16 | 2019-03-08 | 浙江大学 | Complex equipment health status monitoring method based on attention mechanism and neural network |
CN109815892A (en) * | 2019-01-22 | 2019-05-28 | 武汉理工大学 | Signal recognition method of distributed fiber grating sensor network based on CNN |
CN109829538A (en) * | 2019-02-28 | 2019-05-31 | 苏州热工研究院有限公司 | A kind of equipment health Evaluation method and apparatus based on deep neural network |
CN109829577A (en) * | 2019-01-17 | 2019-05-31 | 北京交通大学 | Track train operating status prediction technique based on deep neural network structural model |
CN110175434A (en) * | 2019-06-19 | 2019-08-27 | 西南交通大学 | A kind of rail fastener system injury detection method based on convolutional neural networks |
CN110502991A (en) * | 2019-07-18 | 2019-11-26 | 武汉理工大学 | Internal combustion engine health monitoring method and system based on random convolutional neural network structure |
CN110751073A (en) * | 2019-10-12 | 2020-02-04 | 武汉理工大学 | Pipeline early damage mode identification method based on distributed optical fiber sensing and deep learning |
CN111071300A (en) * | 2020-02-12 | 2020-04-28 | 太原理工大学 | High-speed train rail transit fault safety monitoring and early warning system and signal processing method |
CN111242021A (en) * | 2020-01-10 | 2020-06-05 | 电子科技大学 | A Distributed Optical Fiber Vibration Signal Feature Extraction and Recognition Method |
CN112580153A (en) * | 2020-12-29 | 2021-03-30 | 成都运达科技股份有限公司 | Health state management system and method for vehicle running gear monitoring component |
CN113469060A (en) * | 2021-07-02 | 2021-10-01 | 浙大城市学院 | Multi-sensor fusion convolution neural network aeroengine bearing fault diagnosis method |
CN113465924A (en) * | 2021-06-21 | 2021-10-01 | 武汉理工大学 | Bearing fault diagnosis method and system based on improved BP neural network |
US20210312801A1 (en) * | 2020-04-07 | 2021-10-07 | Nec Laboratories America, Inc | Traffic monitoring using distributed fiber optic sensing |
CN113639646A (en) * | 2021-05-24 | 2021-11-12 | 武汉理工大学 | Full-time global online monitoring device and method for deformation of large engineering structure |
CN113761649A (en) * | 2021-08-05 | 2021-12-07 | 江苏大学 | Intelligent automobile tire eccentric wear prediction method based on one-dimensional convolutional neural network |
-
2022
- 2022-03-16 CN CN202210259899.3A patent/CN114684217B/en active Active
Patent Citations (18)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106503642A (en) * | 2016-10-18 | 2017-03-15 | 长园长通新材料股份有限公司 | A kind of model of vibration method for building up for being applied to optical fiber sensing system |
US20180357542A1 (en) * | 2018-06-08 | 2018-12-13 | University Of Electronic Science And Technology Of China | 1D-CNN-Based Distributed Optical Fiber Sensing Signal Feature Learning and Classification Method |
CN108647786A (en) * | 2018-07-10 | 2018-10-12 | 电子科技大学 | The rotating machinery on-line fault monitoring method of neural network is fought based on depth convolution |
CN109446187A (en) * | 2018-10-16 | 2019-03-08 | 浙江大学 | Complex equipment health status monitoring method based on attention mechanism and neural network |
CN109829577A (en) * | 2019-01-17 | 2019-05-31 | 北京交通大学 | Track train operating status prediction technique based on deep neural network structural model |
CN109815892A (en) * | 2019-01-22 | 2019-05-28 | 武汉理工大学 | Signal recognition method of distributed fiber grating sensor network based on CNN |
CN109829538A (en) * | 2019-02-28 | 2019-05-31 | 苏州热工研究院有限公司 | A kind of equipment health Evaluation method and apparatus based on deep neural network |
CN110175434A (en) * | 2019-06-19 | 2019-08-27 | 西南交通大学 | A kind of rail fastener system injury detection method based on convolutional neural networks |
CN110502991A (en) * | 2019-07-18 | 2019-11-26 | 武汉理工大学 | Internal combustion engine health monitoring method and system based on random convolutional neural network structure |
CN110751073A (en) * | 2019-10-12 | 2020-02-04 | 武汉理工大学 | Pipeline early damage mode identification method based on distributed optical fiber sensing and deep learning |
CN111242021A (en) * | 2020-01-10 | 2020-06-05 | 电子科技大学 | A Distributed Optical Fiber Vibration Signal Feature Extraction and Recognition Method |
CN111071300A (en) * | 2020-02-12 | 2020-04-28 | 太原理工大学 | High-speed train rail transit fault safety monitoring and early warning system and signal processing method |
US20210312801A1 (en) * | 2020-04-07 | 2021-10-07 | Nec Laboratories America, Inc | Traffic monitoring using distributed fiber optic sensing |
CN112580153A (en) * | 2020-12-29 | 2021-03-30 | 成都运达科技股份有限公司 | Health state management system and method for vehicle running gear monitoring component |
CN113639646A (en) * | 2021-05-24 | 2021-11-12 | 武汉理工大学 | Full-time global online monitoring device and method for deformation of large engineering structure |
CN113465924A (en) * | 2021-06-21 | 2021-10-01 | 武汉理工大学 | Bearing fault diagnosis method and system based on improved BP neural network |
CN113469060A (en) * | 2021-07-02 | 2021-10-01 | 浙大城市学院 | Multi-sensor fusion convolution neural network aeroengine bearing fault diagnosis method |
CN113761649A (en) * | 2021-08-05 | 2021-12-07 | 江苏大学 | Intelligent automobile tire eccentric wear prediction method based on one-dimensional convolutional neural network |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115420476A (en) * | 2022-11-03 | 2022-12-02 | 高勘(广州)技术有限公司 | Optical cable-based railway health monitoring method, device, equipment and storage medium |
CN116223763A (en) * | 2022-12-05 | 2023-06-06 | 湖北国际物流机场有限公司 | Method for monitoring health state of runway surface of airport runway landing zone |
CN116223763B (en) * | 2022-12-05 | 2024-03-26 | 湖北国际物流机场有限公司 | Method for monitoring health state of runway surface of airport runway landing zone |
CN115808468A (en) * | 2022-12-08 | 2023-03-17 | 中铁第四勘察设计院集团有限公司 | Ballastless track disease rapid identification method based on vibration waveform |
CN115808468B (en) * | 2022-12-08 | 2024-08-23 | 中铁第四勘察设计院集团有限公司 | Quick identification method for ballastless track diseases based on vibration waveforms |
CN118082914A (en) * | 2024-01-12 | 2024-05-28 | 通号(西安)轨道交通工业集团有限公司北京分公司 | Railway signal axle counting system |
CN119091641A (en) * | 2024-11-07 | 2024-12-06 | 武汉理工大学 | A vehicle continuous tracking method, device and electronic device based on grating array |
Also Published As
Publication number | Publication date |
---|---|
CN114684217B (en) | 2024-03-01 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN114684217B (en) | Rail transit health monitoring system and method | |
Yang et al. | Pipeline safety early warning by multifeature-fusion CNN and LightGBM analysis of signals from distributed optical fiber sensors | |
Liu et al. | Real-time remote measurement of distance using ultra-wideband (UWB) sensors | |
CN114575927A (en) | Rail transit safety monitoring system and method | |
Wang et al. | Research on application of deep convolutional network in high-speed railway track inspection based on distributed fiber acoustic sensing | |
CN101303802A (en) | On-line automatic pre-judgment method and system for truck overload | |
WO2022056677A1 (en) | Monitoring, collection and analysis systems and methods thereof, and device, storage medium, program and program product | |
CN117824756B (en) | Building structure monitoring method, equipment and storage medium | |
US20230024104A1 (en) | Identification of false transformer humming using machine learning | |
CN108444662A (en) | A kind of bridge damnification on-line monitoring method based on degree/day effect | |
CN105755950A (en) | Intelligent optical-fiber inhaul-cable damping support system | |
CN118890090B (en) | Optical cable intelligent monitoring method and system based on digital twin | |
CN119252040B (en) | Road and bridge pavement environment monitoring and early warning method and system | |
CN118246134B (en) | Double-tower cable-stayed bridge life cycle safety control system based on machine learning | |
Ju et al. | Deflection prediction and early warning for long-span bridges using a fusion ATT-BiGRU model and real-time updated thresholds | |
CN118182574B (en) | Train positioning and tunnel security real-time monitoring method and system and storage medium | |
CN106646670A (en) | Distributed micrometeorological monitoring method for transmission lines | |
CN114676540B (en) | Overhead transmission line icing galloping prediction method based on multi-information fusion | |
CN114417698B (en) | Risk monitoring system and evaluation method for external environment along track traffic line | |
CN116952146A (en) | Surrounding rock displacement real-time observation device for urban underground space engineering | |
CN116320824A (en) | Thing allies oneself with electric energy meter of electrified energy quality module | |
Xu et al. | A Cross‐Diagnosis Method for Determining the Structural Condition of a Long‐Span Suspension Bridge Based on the Spatial Windows of Distributed Strain Data | |
CN118129081B (en) | Active safety early warning system based on neural network | |
Jia et al. | Online sensing method for transmission line conductor ice cover based on fiber optic sensing information fusion and continuous wavelet decomposition | |
Zhang et al. | Intelligent perception for multiple intrusion events of urban long-line infrastructures with distributed optical fiber sensing |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |