CN111199257A - Fault diagnosis method and device for high-speed rail driving equipment - Google Patents
Fault diagnosis method and device for high-speed rail driving equipment Download PDFInfo
- Publication number
- CN111199257A CN111199257A CN202010027587.0A CN202010027587A CN111199257A CN 111199257 A CN111199257 A CN 111199257A CN 202010027587 A CN202010027587 A CN 202010027587A CN 111199257 A CN111199257 A CN 111199257A
- Authority
- CN
- China
- Prior art keywords
- original data
- speed rail
- fault
- data signal
- sound signal
- 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.)
- Pending
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/25—Fusion techniques
- G06F18/254—Fusion techniques of classification results, e.g. of results related to same input data
-
- 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
-
- 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
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/08—Feature extraction
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2218/00—Aspects of pattern recognition specially adapted for signal processing
- G06F2218/12—Classification; Matching
Abstract
The embodiment of the invention provides a method and a device for diagnosing faults of high-speed rail driving equipment, wherein the method comprises the following steps: extracting fault characteristics from one or more collected original data signals of the high-speed rail travelling equipment based on a signal analysis method; performing preliminary fault diagnosis on the high-speed rail driving equipment according to the fault characteristics of each original data signal based on a classifier, and acquiring a preliminary diagnosis result corresponding to each original data signal; and fusing the preliminary diagnosis results corresponding to all the original data signals based on a multi-sensor fusion algorithm to obtain a final diagnosis result of the high-speed rail driving equipment. The embodiment of the invention realizes automatic fault diagnosis of the high-speed rail driving equipment, and the diagnosis result is more accurate.
Description
Technical Field
The invention belongs to the technical field of intelligent transportation, and particularly relates to a fault diagnosis method and device for high-speed rail driving equipment.
Background
The physical system of the high-speed railway traffic safety monitoring information is a multi-factor interactive, dynamic and real-time system, the real-time conversion of time and space of the system enables people, such as locomotive drivers of motor train units, train dispatchers and station watchmen, as well as three basic simple and static components of high-speed trains and lines, to be changed into a complex dynamic composite system, any factor in the system is not an original significant factor, and the risk of the system is enlarged, the operability is weakened and the matching performance is relaxed along with the expansion of time and space and the imbalance of dynamic traffic effect.
The high-speed railway traffic accident can be caused by the imbalance of interaction of any factor in the system, so that the system has extremely important theoretical value and practical significance for systematic research on the problems of the occurrence mechanism of the high-speed railway traffic accident, the safety and reliability analysis of traffic related operators, the safety analysis of traffic equipment, the construction of a safety guarantee system, the safety evaluation of the system and the like.
At present, the high-integration equipment facilities of rail transit bring great convenience to operation enterprises, and simultaneously, provide higher requirements for operation and maintenance personnel, especially an automatic black box system and important electromechanical equipment. The periodic overhaul easily causes over-overhaul or under-overhaul of the equipment, and wastes time and labor; the production data is seriously stocked and difficult to effectively guide and overhaul. Especially, a system which has great influence on driving is used for early finding and preventing accidents, which is a core target of intelligent operation and maintenance development.
Disclosure of Invention
In order to overcome the problems that the existing fault diagnosis method for the high-speed rail driving equipment is time-consuming and labor-consuming and is difficult to diagnose or at least partially solve the problems, the embodiment of the invention provides a fault diagnosis method and a fault diagnosis device for the high-speed rail driving equipment.
According to a first aspect of the embodiments of the present invention, there is provided a fault diagnosis method for a high-speed rail driving device, including:
extracting fault characteristics from one or more collected original data signals of the high-speed rail travelling equipment based on a signal analysis method;
performing preliminary fault diagnosis on the high-speed rail driving equipment according to the fault characteristics of each original data signal based on a classifier, and acquiring a preliminary diagnosis result corresponding to each original data signal;
and fusing the preliminary diagnosis results corresponding to all the original data signals based on a multi-sensor fusion algorithm to obtain a final diagnosis result of the high-speed rail driving equipment.
Specifically, the step of extracting fault features from various collected original data signals of the high-speed rail travelling equipment based on a signal analysis method comprises the following steps:
performing time domain analysis on each original data signal to obtain the time domain characteristics of each original data signal;
performing frequency domain analysis on each original data signal to obtain frequency domain characteristics of each original data signal;
and performing time-frequency analysis on each original data signal to obtain the time-frequency characteristics of each original data signal.
Specifically, the step of performing preliminary fault diagnosis on the high-speed rail driving equipment based on a classifier according to the fault characteristics of each raw data signal and obtaining a preliminary diagnosis result corresponding to each raw data signal includes:
and if the original data signal of the high-speed rail driving equipment is one, respectively inputting the fault characteristics of the original data signal into a learning vectorization neural network and a decision tree model, and performing initial fault diagnosis on the high-speed rail driving equipment to obtain an initial diagnosis result corresponding to the original data signal.
Specifically, the step of performing preliminary fault diagnosis on the high-speed rail driving equipment based on a classifier according to the fault characteristics of each raw data signal and obtaining a preliminary diagnosis result corresponding to each raw data signal includes:
if the original data signal of the high-speed rail driving equipment comprises a sound signal, preprocessing the sound signal into a time-frequency graph, inputting the time-frequency graph into a self-adaptive stacked convolutional neural network model, and obtaining a preliminary diagnosis result corresponding to the sound signal; alternatively, the first and second electrodes may be,
and slicing the sound signal, inputting the sliced sound signal into an end-to-end stacked convolutional neural network model, and obtaining a preliminary diagnosis result corresponding to the sound signal.
Specifically, the step of preprocessing the sound signal into a time-frequency diagram, inputting the time-frequency diagram into a self-adaptive stacked convolutional neural network model, and obtaining a preliminary diagnosis result corresponding to the sound signal includes:
the sound signal is sliced, the sliced sound signal is converted from a time domain to a time-frequency domain based on a wavelet transformation method, a time-frequency diagram of the sound signal is obtained, and the size of the time-frequency diagram is adjusted;
performing feature extraction and dimension reduction on the time-frequency graph after the size is adjusted on the basis of a convolutional layer and a pooling layer in the self-adaptive stacked convolutional neural network model;
inputting the characteristics output by the last convolutional layer in the self-adaptive stacked convolutional neural network model into a full-connection layer;
and acquiring a preliminary diagnosis result corresponding to the sound signal based on a Softmax function according to the output of the full connection layer.
Specifically, the step of slicing the sound signal and inputting the sliced sound signal into an end-to-end stacked convolutional neural network model to obtain a preliminary diagnosis result corresponding to the sound signal includes:
extracting local features of the sound signal based on the first two convolutional layers in the end-to-end stacked convolutional neural network model;
performing non-overlapping maximum pooling operation on local features of the sound signal based on pooling layers after the first two convolutional layers, and outputting frequency-like features;
reshaping the frequency-like characteristics based on the data reshaping layer after the pooling layer;
continuously extracting the characteristics of the reshaped similar frequency characteristics based on the two convolution layers after the data reshaping layer;
inputting the output characteristics of the last convolutional layer in the end-to-end stacked convolutional neural network model into a full connection layer;
and acquiring a preliminary diagnosis result corresponding to the sound signal based on a Softmax function according to the output of the full connection layer.
Specifically, based on the signal analysis method, the step of extracting fault features from the collected multiple kinds of raw data signals of the high-speed rail travelling equipment further comprises the following steps of:
preprocessing each original data signal according to an interface for collecting each original data signal; wherein the pre-processing comprises signal conversion and cleaning;
and performing data fusion on the processed original data signal based on an improved evidence theory fusion algorithm.
According to a second aspect of the embodiments of the present invention, there is provided a fault diagnosis device for a high-speed rail traveling apparatus, including:
the extraction module is used for extracting fault characteristics from one or more collected original data signals of the high-speed rail travelling equipment based on a signal analysis method;
the diagnosis module is used for carrying out preliminary fault diagnosis on the high-speed rail driving equipment based on a classifier according to the fault characteristics of each original data signal and obtaining a preliminary diagnosis result corresponding to each original data signal;
and the fusion module is used for fusing the preliminary diagnosis results corresponding to all the original data signals based on a multi-sensor fusion algorithm to obtain the final diagnosis result of the high-speed rail driving equipment.
According to a third aspect of the embodiments of the present invention, there is also provided an electronic device, including a memory, a processor, and a computer program stored on the memory and executable on the processor, where the processor calls the program instructions to be able to execute the method for diagnosing the fault of the high-speed rail driving device provided in any one of the various possible implementations of the first aspect.
According to a fourth aspect of the embodiments of the present invention, there is also provided a non-transitory computer-readable storage medium storing computer instructions for causing a computer to execute the method for diagnosing the fault of the high-speed rail traveling equipment provided in any one of the various possible implementation manners of the first aspect.
The embodiment of the invention provides a method and a device for diagnosing faults of high-speed rail driving equipment.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to these drawings without creative efforts.
Fig. 1 is a schematic overall flow chart of a fault diagnosis method for high-speed rail driving equipment according to an embodiment of the present invention;
fig. 2 is a schematic flow chart of state monitoring and fault diagnosis of the high-speed rail driving equipment in the fault diagnosis method for the high-speed rail driving equipment according to the embodiment of the invention;
fig. 3 is a schematic flowchart of multi-signal-source fusion diagnosis in the fault diagnosis method for the high-speed rail driving equipment according to the embodiment of the invention;
fig. 4 is a schematic flow chart of fault information fusion diagnosis in the fault diagnosis method for the high-speed rail driving equipment according to the embodiment of the invention;
fig. 5 is a schematic diagram of a predictive maintenance flow based on equipment health degree evaluation in a fault diagnosis method for a high-speed rail driving equipment according to an embodiment of the present invention;
fig. 6 is a schematic structural diagram of an overall fault diagnosis device for high-speed rail driving equipment according to an embodiment of the present invention;
fig. 7 is a schematic view of an overall structure of an electronic device according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
In an embodiment of the present invention, a method for diagnosing a fault of a high-speed rail driving equipment is provided, and fig. 1 is a schematic overall flow chart of the method for diagnosing a fault of a high-speed rail driving equipment provided in the embodiment of the present invention, where the method includes: s101, extracting fault characteristics from one or more collected original data signals of the high-speed rail travelling equipment based on a signal analysis method;
specifically, a complex sensing technology of a high-speed railway driving safety monitoring information physical system is adopted to acquire real-time operation data of high-speed railway driving equipment. According to the embodiment, a station, line and network three-level operation and maintenance cloud platform is constructed by collecting signals and power supply system data by means of sensing means such as a video voice technology, an intelligent mobile terminal, an online monitoring sensor and an intelligent chip, unified data standard storage, metadata management and data quality monitoring are realized on different platform data such as various service monitoring systems and operation systems, and an intelligent equipment monitoring big data platform is constructed. A unified coding system is constructed by taking intelligent operation and maintenance management as guidance, and a unified data sharing platform for supporting state monitoring and intelligent operation and maintenance management of different professional devices is formed by adopting a mixed coding mode of a bar code and a Radio Frequency Identification (RFID) electronic tag.
The method fully utilizes various means to collect the state information of the high-speed rail driving equipment, and utilizes a signal analysis method to extract fault characteristics, wherein the signal analysis method comprises time domain analysis, frequency domain analysis and time frequency analysis.
S102, performing preliminary fault diagnosis on the high-speed rail driving equipment based on a classifier according to the fault characteristics of each original data signal, and obtaining a preliminary diagnosis result corresponding to each original data signal;
and inputting each extracted fault feature into a classifier, and performing primary fault diagnosis on the high-speed rail driving equipment. Classifiers such as Learning vectorization Neural Networks (LVQ), Convolutional Neural Networks (CNN), and the like. There is a preliminary diagnostic result for each raw data signal.
S103, fusing the preliminary diagnosis results corresponding to all the original data signals based on a multi-sensor fusion algorithm to obtain a final diagnosis result of the high-speed rail driving equipment.
The multi-sensor fusion algorithm may be an IDS (Intrusion Detection System) fusion algorithm, but is not limited to this algorithm. Aiming at the problem that evidence conflict cannot be effectively processed by the most widely applied evidence theory in the multi-sensor fusion algorithm, an improved fusion algorithm is provided. Firstly, the phenomenon of one vote rejection is avoided by utilizing an adjacent borrowing value mode; secondly, jointly correcting the basic probability assignment of the evidence by using a distance function and a Delphi method; then, selecting a corresponding fusion rule according to the relation between the conflict factor and the threshold value to complete data fusion; and finally, applying the improved evidence theory fusion algorithm to the fault diagnosis of the equipment to obtain the fault diagnosis rate higher than that of the original evidence theory.
The multi-sensor fusion algorithm fuses the preliminary diagnosis results corresponding to various original data signals to obtain the final diagnosis result. And displaying the diagnosis result in a visual mode to assist decision-making personnel in making decisions. Fig. 2 is a schematic flow chart of state monitoring and fault diagnosis of the high-speed rail driving equipment.
According to the method, relevant multi-source information is fused, the possible abnormal state of the equipment is predicted, proper predictive maintenance is adopted for the abnormal state in advance, problems are found and processed in advance, and the normal running time of the equipment is prolonged; the self-cognition, learning and reconstruction capabilities are provided for the optimal use and timely maintenance of the equipment, and the equipment health management is realized; the preview speculation is carried out in the virtual equipment health diagnosis model, remote fault diagnosis is realized, faults are solved timely and quickly, maintenance cost is reduced and reduced, and meanwhile, the application of an information physical system technology in high-speed rail driving safety is driven.
In the embodiment, the holographic sensing system influencing the high-speed rail travelling equipment is established, and the automatic station equipment characteristic identification, the intelligent state sensing and the active data acquisition based on the information physical system are carried out. The method comprises the steps of collecting key technologies and interface standards of equipment state and fault information, analyzing attributes and properties of equipment and key components thereof influencing system faults through a feature recognition technology to form a key equipment operation and maintenance fault point map, carrying out statistical analysis on data according to multiple dimensions such as line type, specialty, equipment type and components, and visually displaying equipment faults, maintenance records and the like.
According to the method, the fault characteristics are extracted from the acquired original data signals of the high-speed rail driving equipment based on a signal analysis method, preliminary fault diagnosis is carried out according to the fault characteristics based on the classifier, the preliminary fault diagnosis results are fused based on a multi-sensor fusion algorithm to obtain the final diagnosis result, so that automatic fault diagnosis of the high-speed rail driving equipment is realized, and the diagnosis result is more accurate.
On the basis of the above embodiment, in this embodiment, based on a signal analysis method, the step of extracting fault features from multiple kinds of raw data signals of the collected high-speed rail driving equipment includes: performing time domain analysis on each original data signal to obtain the time domain characteristics of each original data signal; performing frequency domain analysis on each original data signal to obtain frequency domain characteristics of each original data signal; and performing time-frequency analysis on each original data signal to obtain the time-frequency characteristics of each original data signal.
Specifically, when the time domain features of the original data signals are extracted, the time domain analysis takes a time axis as an abscissa, and dynamic information of various states of mechanical equipment changing along with time in the operation process, such as vibration signals, sound signals, load signals and the like, is collected by using a data acquisition card, so that the time domain signals of the original data signals can be obtained. When extracting the frequency domain characteristics of the original data signal, the frequency domain analysis represents the dynamic change of the signal with the frequency axis as the abscissa. Time domain signals obtained according to time sequence are decomposed into frequency space through Fourier transformation, and therefore amplitude and phase information of frequency components are obtained.
On the basis of the foregoing embodiment, in this embodiment, a classifier is used to perform preliminary fault diagnosis on the high-speed rail driving equipment according to a fault feature of each type of the original data signal, and the step of obtaining a preliminary diagnosis result corresponding to each type of the original data signal includes: and if the original data signal of the high-speed rail driving equipment is one, respectively inputting the fault characteristics of the original data signal into a learning vectorization neural network and a decision tree model, and performing initial fault diagnosis on the high-speed rail driving equipment to obtain an initial diagnosis result corresponding to the original data signal.
Specifically, technologies such as data physical fusion, a neural network and an intelligent algorithm are adopted to diagnose the fault state of the equipment from two angles of single-signal-source multi-sensor integration and multi-signal-source multi-sensor fusion. And obtaining a preliminary diagnosis result of the single signal source by using the classifier, and further performing decision fusion on the preliminary diagnosis result by combining an improved evidence theory to obtain a more reliable device operation state.
Aiming at the problem that a single-signal-source non-integrated model is constrained by a single structure and is difficult to comprehensively reflect the state of equipment, the integrated diagnosis model fusing a learning vectorization neural network and a decision tree based on an improved evidence theory is provided. For example, a bearing is taken as a research object, firstly, the statistical characteristics of multiple information fields are extracted by using the experimental data of the bearing fault; secondly, reducing the dimension of the statistical characteristics by adopting a principal component analysis method, and respectively sending the characteristics of the fan end and the drive end into a learning vectorization neural network and a decision tree model for preliminary bearing fault diagnosis; and finally, transmitting the fault recognition rate of the single model as an evidence into an improved fusion algorithm, and realizing fault diagnosis of the bearing through further decision fusion.
On the basis of the foregoing embodiment, in this embodiment, a classifier is used to perform preliminary fault diagnosis on the high-speed rail driving equipment according to a fault feature of each type of the original data signal, and the step of obtaining a preliminary diagnosis result corresponding to each type of the original data signal includes: if the original data signal of the high-speed rail driving equipment comprises a sound signal, preprocessing the sound signal into a time-frequency graph, inputting the time-frequency graph into a self-adaptive stacked convolutional neural network model, and obtaining a preliminary diagnosis result corresponding to the sound signal; or slicing the sound signal and inputting the sliced sound signal into an end-to-end stacked convolutional neural network model to obtain a preliminary diagnosis result corresponding to the sound signal.
Specifically, a diagnosis method for fusing vibration and sound signals of a multi-source sensor based on a convolutional neural network is provided for solving the problems that a single signal source sensor is limited by single signal source and self performance, and the running state of mechanical equipment is difficult to reflect comprehensively under different working conditions. For example, 200 sound samples are selected by taking a gear as a research object, and a multi-signal-source fusion diagnostic method is provided. Firstly, building a gear box fault diagnosis platform in a semi-anechoic chamber environment and collecting sound signals under different working conditions; secondly, preprocessing the sound signal into a time-frequency diagram, and sending the time-frequency diagram into an Adaptive Stacked Convolutional Neural Network (ASCNN) model to obtain a preliminary diagnosis result of the sound signal. The other method is to slice the sound signal directly into an End-to-End Stacked Convolutional Neural Network (ESCNN) model to obtain a preliminary diagnosis result of the sound signal; and finally, further fusing and deciding the preliminary diagnosis results of other signals except the sound signal and the sound signal in the original data signal by using an improved evidence fusion algorithm to obtain a more accurate and more reliable gear state.
Aiming at the sound signal of the gearbox, an adaptive stacked convolutional neural network model is provided. Converting the signal from a time domain to a time-frequency domain by utilizing wavelet transform according to a conventional feature extraction method, and sending the time-frequency graph with the adjusted size to an ASCNN model for diagnosis; aiming at the sound signals of the gear box, an end-to-end stacked convolutional neural network model is provided, the background dependence of manually extracting features is avoided, the two steps of feature extraction and fault classification are combined into a whole to be finished in a self-adaptive mode in one model, the original sound signals are subjected to slicing processing and then directly sent into an ESCNN model, and the operation of feature extraction and classification recognition is finished by a convolutional layer. And finally, in order to solve the limitation that a single information source cannot sufficiently reflect the comprehensive information of the tested object, a multi-sensor fusion algorithm is utilized to perform fusion decision on the preliminary diagnosis results of other signals and sound signals, so that a more accurate and reliable device operation state is obtained. A flow chart of multi-source sensor fusion diagnosis based on convolutional neural network is shown in fig. 3.
On the basis of the foregoing embodiment, in this embodiment, the step of preprocessing the sound signal into a time-frequency graph and inputting the time-frequency graph into an adaptive stacked convolutional neural network model to obtain a preliminary diagnosis result corresponding to the sound signal includes: the sound signal is sliced, the sliced sound signal is converted from a time domain to a time-frequency domain based on a wavelet transformation method, a time-frequency diagram of the sound signal is obtained, and the size of the time-frequency diagram is adjusted; performing feature extraction and dimension reduction on the time-frequency graph after the size is adjusted on the basis of a convolutional layer and a pooling layer in the self-adaptive stacked convolutional neural network model; inputting the characteristics output by the last convolutional layer in the self-adaptive stacked convolutional neural network model into a full-connection layer; and acquiring a preliminary diagnosis result corresponding to the sound signal based on a Softmax function according to the output of the full connection layer.
Specifically, aiming at the fault problem of the gearbox under the variable rotating speed working condition, the time-frequency graph of the sound signal is acted by convolution operation, the required features are extracted in a self-adaptive mode, similar semantics are combined by pooling operation, the calculation scale is reduced, and an ASCNN gearbox fault diagnosis model is built. The model comprises an input layer, three convolution layers, two sub-sampling layers, namely a pooling layer, a full-link layer and an output layer. And realizing the self-adaptive feature extraction and dimension reduction of the sound signal time-frequency diagram through the stacking operation of a plurality of convolution layers and pooling layers.
And for each sound signal acquired by the acceleration sensor, segmenting the sound signal and carrying out time-frequency transformation to obtain a time-frequency diagram, and further adjusting the size of the time-frequency diagram to be 32 x 32 so as to meet the input requirement of ASCNN. Then, a 'convolution-pooling' stacking operation is carried out, 5 convolution kernels with the size of 28 × 28 are set to act on the convolution layer 1, the step size is moved to 1, and 2 × 2 maximal pooling operation is adopted, and the step size is moved to 2; the second convolution layer has 10 convolution kernels with the size of 10 x 10, the moving step length is 1, and the parameters of the sub-sampling layer 2 are equal to the previous pooling layer; the third convolutional layer is provided with 10 4 x 4 convolutional kernels with step size 1.
In order to extract local features as much as possible, a small-size convolution kernel filtering time-frequency graph is adopted in a convolution part. Then a full connection layer containing 200 hidden units and a Logistic-regression layer for fault classification by using a Softmax function, and the final output layer outputs the identification precision of 10 fault types of the gearbox.
And (3) randomly initializing the weight when the network training is started, and reversely propagating and correcting the network weight by calculating errors of the predicted value and the true value in the training process until a termination condition is met. From the time-frequency diagram samples corresponding to each gearbox fault, 75% of the samples are randomly selected as a training set, and the remaining 25% are selected as a testing set. The ASCNN obtains a trained prediction model by using 75% of training samples to learn and memorize fault characteristics in a self-adaptive manner, then sends 25% of test sets into a trained ASCNN network, and obtains a prediction result by means of a Softmax function.
On the basis of the above embodiment, in this embodiment, the step of slicing the sound signal and inputting the sliced sound signal into an end-to-end stacked convolutional neural network model to obtain a preliminary diagnosis result corresponding to the sound signal includes: extracting local features of the sound signal based on the first two convolutional layers in the end-to-end stacked convolutional neural network model; performing non-overlapping maximum pooling operation on local features of the sound signal based on pooling layers after the first two convolutional layers, and outputting frequency-like features; reshaping the frequency-like characteristics based on the data reshaping layer after the pooling layer; continuously extracting the characteristics of the reshaped similar frequency characteristics based on the two convolution layers after the data reshaping layer; inputting the output characteristics of the last convolutional layer in the end-to-end stacked convolutional neural network model into a full connection layer; and acquiring a preliminary diagnosis result corresponding to the sound signal based on a Softmax function according to the output of the full connection layer.
Specifically, the sound signal is directly input into an end-to-end stacked convolutional neural network, the convolutional layer is utilized to automatically learn and extract the characteristics required by the model, and then the fault identification is carried out on the running state of the gear. The ESCNN model includes an input layer, four convolutional layers, two sub-sampling layers, a full link layer, and an output layer. The advantage of end-to-end is that the input layer directly inputs the sound signal, the step of manually extracting the characteristics is omitted, and the characteristic extraction is finished by the first two convolutional layers. The local features of the sound signal are extracted using small-sized convolution kernel sliding on convolutional layers 1 and 2. In extracting features from the audio signal, a first convolutional layer is provided with 40 1 × 8 convolutional kernels with step size 1, and a second convolutional layer is provided with 40 × 8 convolutional kernels with step size 1 to extract different audio features.
And sending the sound segment with the time length of 1s after segmentation into an input layer, and obtaining the output characteristics of the time sequence by adopting the maximal pooling operation with the size of 160 x 1 and the moving step length of 160 without overlapping after two layers of convolution. Since convolution and pooling act on the time sequence, each 40-dimensional vector represents a frequency-like feature of the corresponding 10ms time region. The next convolution target is fault identification, and the class frequency characteristics are used as images and sent to an identification stage. And reshaping the data through a Reshape function to obtain a two-dimensional matrix of 40 × 160, which is similar to the convolution operation of sending the time-frequency diagram to the next layer. The two convolutional layers then continue to extract and learn the features of the different sound signals. Then a fully connected layer containing 200 hidden units is prepared for fault identification. And the next logic-regression layer for fault classification by using a Softmax function, wherein the next layer comprises a class label corresponding to the sample. The final output layer then outputs 10 failure accuracies of the gear.
The weight of each layer is initialized randomly at the beginning of the model, and is continuously corrected and optimized by using error back propagation in the training process. After training is finished, the test samples are sent to an ESCNN model, and the identification precision of each type of fault is obtained by comparing the real label and the predicted label of each sample.
On the basis of the foregoing embodiments, before the step of extracting the fault feature from the collected multiple kinds of raw data signals of the high-speed rail driving equipment in this embodiment, the method further includes: preprocessing each original data signal according to an interface for collecting each original data signal; wherein the pre-processing comprises signal conversion and cleaning; and performing data fusion on the processed original data signal based on an improved evidence theory fusion algorithm.
Specifically, the characteristics representing the nature of the fault are extracted from the original data signals and sent to a classifier, and the input characteristics are classified and identified autonomously through the learning of a large number of samples, so that the fault diagnosis process is completed. Firstly, preprocessing an original data signal, performing signal conversion and cleaning to remove redundancy according to different acquisition interface requirements, and then performing comprehensive analysis processing by data fusion and feature fusion step by step. In the data layer fusion, the original data signals are subjected to fast Fourier transform, wavelet analysis and the like to extract characteristic data, so that the characteristics of the original data signals are obtained, and the characteristic layers are fused.
In the embodiment, artificial intelligence methods such as a neural network, a support vector machine, a decision tree, a fault tree and the like are used for diagnosis, so that a more accurate diagnosis result is obtained. Firstly, data fusion is carried out, feature extraction is carried out according to the fused data, then feature fusion is carried out, and the result of the feature fusion is sent to a decision layer. And (3) carrying out final decision making by using intelligent fusion algorithms such as Bayes theory, maximum likelihood estimation, expert system, clustering, Kalman filtering, weighted least square method, evidence theory and the like, and assisting managers in maintaining equipment, as shown in figure 4.
When the collected original data signals are preprocessed, relevant databases are established by extracting pre-maintenance, alarming, remote measuring, remote signaling, operation, recorded information and the like. And (3) establishing a logic fault standardization rule among equipment, forming a fault diagnosis expert knowledge base, and finishing data labeling preprocessing by using an isolated forest algorithm. By intelligent information integration, the intelligent state monitoring, data query, report function, state evaluation and other functions of the high-speed rail driving equipment are realized.
When data level fusion is carried out, data acquired by the original sensors are directly analyzed and processed for data sources of the same type. The feature level fusion obtains corresponding feature vectors by extracting the features of the measured values of the sensors, and then comprehensively analyzes and processes the feature vectors, so that the main features of information are reserved, a certain degree of information compression is realized, the real-time performance is ensured, and the method belongs to the fusion of intermediate levels. And the decision-level fusion carries out preliminary decision on the characteristic vectors of all the sensors from a specific decision problem, and then recombines and evaluates the preliminary diagnosis results according to a certain rule and reliability to obtain an optimal decision for a specific decision target.
In addition, by using big data analysis methods such as high-frequency fault clustering and the like and an artificial intelligence machine learning alternation technology, high-frequency faults are diagnosed and judged and the positioning of key parts is prevented and corrected based on the relation diagnosis of state data, and faults are predicted, as shown in fig. 5. And (3) in combination with maintenance management, carrying out intelligent inspection on machine room equipment based on an IoT (Internet of Things) perception technology, constructing an accurate three-dimensional model in the tunnel by using a laser radar, and mining the hidden fault danger characteristics based on big data analysis.
The equipment life prediction in predictive maintenance based on equipment health assessment is based on a fault curve. Firstly, obtaining the service life of equipment based on a high-acceleration test method; accumulating the service time of the equipment in the working process, and carrying out normalization calculation on the loads under different working conditions based on an artificial intelligence method to form weighted working time; and obtaining the service life prediction fault and service life of the equipment based on the weighted working time and the high-acceleration test.
By utilizing a novel information interaction platform with data interaction, communication and video and voice visualization, system fault study and judgment, self-starting and self-defining of a maintenance service work order of a system-specific key device are adopted, and the priority and operation and maintenance optimization scheme of main equipment facilities for preventive maintenance and predictive maintenance is obtained. Meanwhile, an evaluation model of preventive maintenance is established according to the failure mode and the influence analysis result, and an intelligent maintenance plan and maintenance resource scheduling are automatically generated. The robot-based robot transition is completed through personnel track monitoring and operation process guidance, and repeated and unstable manual operation is avoided. The intelligent monitoring and operation decision of the whole life cycle management and control maintenance of the key equipment system is achieved.
And designing a trend prediction module on the basis of an LSTM (Long Short-Term Memory) algorithm, and predicting the future trend of different data. An abnormity detection model is built through the XGboost model, a plurality of different abnormity detection models are trained according to a data set of station equipment, an abnormity detection module diagnoses abnormity according to a real-time data source loaded corresponding abnormity detection model, fault alarming and management analysis of a minimum maintainable unit are realized through a visual intelligent monitoring module, a predefined processing flow can be called, and automatic fault non-intervention preprocessing is realized.
And establishing an intelligent operation and maintenance service application and operation and maintenance management module around the maintenance service and the fault association constraint condition of the signal and power supply key equipment. The operation and maintenance service application provides the priority and operation and maintenance optimization scheme of main equipment facilities for preventive maintenance and predictive maintenance from the view of operation and maintenance service for equipment state monitoring, abnormal alarm, trend prediction, reliability evaluation and fault statistics. The operation and maintenance management carries out data classification on key equipment data according to different levels of maintenance work areas and operation and maintenance personnel, establishes records of a full life cycle and a maintenance ledger, and makes contents such as maintenance flow, maintenance regulation, intelligent maintenance guidance, maintenance plan management and the like.
In another embodiment of the invention, a fault diagnosis device for high-speed rail traveling equipment is provided, and the device is used for realizing the method in the foregoing embodiments. Therefore, the description and definition in the foregoing embodiments of the fault diagnosis method for the high-speed rail traveling equipment may be used for understanding the respective execution modules in the embodiments of the present invention. Fig. 6 is a schematic diagram of an overall structure of a fault diagnosis device for high-speed rail traveling equipment according to an embodiment of the present invention, where the device includes an extraction module 601, a diagnosis module 602, and a fusion module 603, where:
the extraction module 601 is used for extracting fault features from one or more collected original data signals of the high-speed rail travelling equipment based on a signal analysis method;
the extraction module 601 collects state information of the high-speed rail driving equipment by various means, and extracts fault features by using a signal analysis method, wherein the signal analysis method comprises time domain analysis, frequency domain analysis and time frequency analysis.
The diagnosis module 602 is configured to perform preliminary fault diagnosis on the high-speed rail driving equipment according to the fault feature of each type of the original data signal based on a classifier, and obtain a preliminary diagnosis result corresponding to each type of the original data signal;
the diagnosis module 602 inputs each extracted fault feature into the classifier to perform preliminary fault diagnosis on the high-speed rail driving equipment. Classifiers such as learning vectorized neural networks and convolutional neural networks. There is a preliminary diagnostic result for each raw data signal.
The fusion module 603 is configured to fuse the preliminary diagnosis results corresponding to all the raw data signals based on a multi-sensor fusion algorithm, and obtain a final diagnosis result of the high-speed rail driving equipment.
The fusion module 603 fuses the preliminary diagnosis results corresponding to the plurality of raw data signals by using a multi-sensor fusion algorithm to obtain a final diagnosis result. And displaying the diagnosis result in a visual mode to assist decision-making personnel in making decisions.
According to the method, the fault characteristics are extracted from the acquired original data signals of the high-speed rail driving equipment based on a signal analysis method, preliminary fault diagnosis is carried out according to the fault characteristics based on the classifier, the preliminary fault diagnosis results are fused based on a multi-sensor fusion algorithm to obtain the final diagnosis result, so that automatic fault diagnosis of the high-speed rail driving equipment is realized, and the diagnosis result is more accurate.
On the basis of the foregoing embodiment, the extraction module in this embodiment is specifically configured to: performing time domain analysis on each original data signal to obtain the time domain characteristics of each original data signal; performing frequency domain analysis on each original data signal to obtain frequency domain characteristics of each original data signal; and performing time-frequency analysis on each original data signal to obtain the time-frequency characteristics of each original data signal.
On the basis of the foregoing embodiment, the diagnosis module in this embodiment is specifically configured to: and if the original data signal of the high-speed rail driving equipment is one, respectively inputting the fault characteristics of the original data signal into a learning vectorization neural network and a decision tree model, and performing initial fault diagnosis on the high-speed rail driving equipment to obtain an initial diagnosis result corresponding to the original data signal.
On the basis of the foregoing embodiment, the diagnosis module in this embodiment is specifically configured to: if the original data signal of the high-speed rail driving equipment is a sound signal, preprocessing the sound signal into a time-frequency graph, inputting the time-frequency graph into a self-adaptive stacked convolutional neural network model, and obtaining a preliminary diagnosis result corresponding to the sound signal; and slicing the sound signal, inputting the sliced sound signal into an end-to-end stacked convolutional neural network model, and obtaining a preliminary diagnosis result corresponding to the sound signal.
On the basis of the foregoing embodiment, the diagnosis module in this embodiment is further configured to: the sound signal is sliced, the sliced sound signal is converted from a time domain to a time-frequency domain based on a wavelet transformation method, a time-frequency diagram of the sound signal is obtained, and the size of the time-frequency diagram is adjusted; performing feature extraction and dimension reduction on the time-frequency graph after the size is adjusted on the basis of a convolutional layer and a pooling layer in the self-adaptive stacked convolutional neural network model; inputting the characteristics output by the last convolutional layer in the self-adaptive stacked convolutional neural network model into a full-connection layer; and acquiring a preliminary diagnosis result corresponding to the sound signal based on a Softmax function according to the output of the full connection layer.
On the basis of the foregoing embodiment, the diagnosis module in this embodiment is further configured to: extracting local features of the sound signal based on the first two convolutional layers in the end-to-end stacked convolutional neural network model; performing non-overlapping maximum pooling operation on local features of the sound signal based on pooling layers after the first two convolutional layers, and outputting frequency-like features; reshaping the frequency-like characteristics based on the data reshaping layer after the pooling layer; continuously extracting the characteristics of the reshaped similar frequency characteristics based on the two convolution layers after the data reshaping layer; inputting the output characteristics of the last convolutional layer in the end-to-end stacked convolutional neural network model into a full connection layer; and acquiring a preliminary diagnosis result corresponding to the sound signal based on a Softmax function according to the output of the full connection layer.
On the basis of the above embodiment, the present embodiment further includes a preprocessing module, configured to preprocess each of the raw data signals according to an interface for acquiring each of the raw data signals; wherein the pre-processing comprises signal conversion and cleaning; and performing data fusion on the processed original data signal based on an improved evidence theory fusion algorithm.
Fig. 7 illustrates a physical structure diagram of an electronic device, and as shown in fig. 7, the electronic device may include: a processor (processor)701, a communication Interface (Communications Interface)702, a memory (memory)703 and a communication bus 704, wherein the processor 701, the communication Interface 702 and the memory 703 complete communication with each other through the communication bus 704. The processor 701 may call logic instructions in the memory 703 to perform the following method: extracting fault characteristics from one or more collected original data signals of the high-speed rail travelling equipment based on a signal analysis method; performing preliminary fault diagnosis on the high-speed rail driving equipment according to the fault characteristics of each original data signal based on a classifier, and acquiring a preliminary diagnosis result corresponding to each original data signal; and fusing the preliminary diagnosis results corresponding to all the original data signals based on a multi-sensor fusion algorithm to obtain a final diagnosis result of the high-speed rail driving equipment.
In addition, the logic instructions in the memory 703 can be implemented in the form of software functional units and stored in a computer readable storage medium when the software functional units are sold or used as independent products. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), a magnetic disk or an optical disk, and other various media capable of storing program codes.
The present embodiments provide a non-transitory computer-readable storage medium storing computer instructions that cause a computer to perform the methods provided by the above method embodiments, for example, including: extracting fault characteristics from one or more collected original data signals of the high-speed rail travelling equipment based on a signal analysis method; performing preliminary fault diagnosis on the high-speed rail driving equipment according to the fault characteristics of each original data signal based on a classifier, and acquiring a preliminary diagnosis result corresponding to each original data signal; and fusing the preliminary diagnosis results corresponding to all the original data signals based on a multi-sensor fusion algorithm to obtain a final diagnosis result of the high-speed rail driving equipment.
Those of ordinary skill in the art will understand that: all or part of the steps for implementing the method embodiments may be implemented by hardware related to program instructions, and the program may be stored in a computer readable storage medium, and when executed, the program performs the steps including the method embodiments; and the aforementioned storage medium includes: various media that can store program codes, such as ROM, RAM, magnetic or optical disks.
The above-described embodiments of the apparatus are merely illustrative, and the units described as separate parts may or may not be physically separate, and parts displayed as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the solution of the present embodiment. One of ordinary skill in the art can understand and implement it without inventive effort.
Through the above description of the embodiments, those skilled in the art will clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and certainly can also be implemented by hardware. With this understanding in mind, the above-described technical solutions may be embodied in the form of a software product, which can be stored in a computer-readable storage medium such as ROM/RAM, magnetic disk, optical disk, etc., and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in the embodiments or some parts of the embodiments.
Finally, it should be noted that: the above examples are only intended to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.
Claims (10)
1. A fault diagnosis method for high-speed rail traveling equipment is characterized by comprising the following steps:
extracting fault characteristics from one or more collected original data signals of the high-speed rail travelling equipment based on a signal analysis method;
performing preliminary fault diagnosis on the high-speed rail driving equipment according to the fault characteristics of each original data signal based on a classifier, and acquiring a preliminary diagnosis result corresponding to each original data signal;
and fusing the preliminary diagnosis results corresponding to all the original data signals based on a multi-sensor fusion algorithm to obtain a final diagnosis result of the high-speed rail driving equipment.
2. The method for diagnosing the fault of the high-speed rail driving equipment according to claim 1, wherein the step of extracting the fault feature from the collected multiple kinds of raw data signals of the high-speed rail driving equipment based on a signal analysis method comprises the following steps:
performing time domain analysis on each original data signal to obtain the time domain characteristics of each original data signal;
performing frequency domain analysis on each original data signal to obtain frequency domain characteristics of each original data signal;
and performing time-frequency analysis on each original data signal to obtain the time-frequency characteristics of each original data signal.
3. The method for diagnosing the fault of the high-speed rail driving equipment according to claim 1, wherein the step of performing preliminary fault diagnosis on the high-speed rail driving equipment according to the fault characteristics of each original data signal based on a classifier, and acquiring the preliminary diagnosis result corresponding to each original data signal comprises the steps of:
and if the original data signal of the high-speed rail driving equipment is one, respectively inputting the fault characteristics of the original data signal into a learning vectorization neural network and a decision tree model, and performing initial fault diagnosis on the high-speed rail driving equipment to obtain an initial diagnosis result corresponding to the original data signal.
4. The method for diagnosing the fault of the high-speed rail driving equipment according to claim 1, wherein the step of performing preliminary fault diagnosis on the high-speed rail driving equipment according to the fault characteristics of each original data signal based on a classifier, and acquiring the preliminary diagnosis result corresponding to each original data signal comprises the steps of:
if the original data signal of the high-speed rail driving equipment comprises a sound signal, preprocessing the sound signal into a time-frequency graph, inputting the time-frequency graph into a self-adaptive stacked convolutional neural network model, and obtaining a preliminary diagnosis result corresponding to the sound signal; alternatively, the first and second electrodes may be,
and slicing the sound signal, inputting the sliced sound signal into an end-to-end stacked convolutional neural network model, and obtaining a preliminary diagnosis result corresponding to the sound signal.
5. The method according to claim 4, wherein the step of preprocessing the sound signal into a time-frequency diagram and inputting the time-frequency diagram into an adaptive stacked convolutional neural network model to obtain a preliminary diagnosis result corresponding to the sound signal comprises:
the sound signal is sliced, the sliced sound signal is converted from a time domain to a time-frequency domain based on a wavelet transformation method, a time-frequency diagram of the sound signal is obtained, and the size of the time-frequency diagram is adjusted;
performing feature extraction and dimension reduction on the time-frequency graph after the size is adjusted on the basis of a convolutional layer and a pooling layer in the self-adaptive stacked convolutional neural network model;
inputting the characteristics output by the last convolutional layer in the self-adaptive stacked convolutional neural network model into a full-connection layer;
and acquiring a preliminary diagnosis result corresponding to the sound signal based on a Softmax function according to the output of the full connection layer.
6. The method for diagnosing the fault of the high-speed rail traveling equipment according to claim 4, wherein the step of slicing the sound signal and inputting the sliced sound signal into an end-to-end stacked convolutional neural network model to obtain a preliminary diagnosis result corresponding to the sound signal comprises the following steps:
extracting local features of the sound signal based on the first two convolutional layers in the end-to-end stacked convolutional neural network model;
performing non-overlapping maximum pooling operation on local features of the sound signal based on pooling layers after the first two convolutional layers, and outputting frequency-like features;
reshaping the frequency-like characteristics based on the data reshaping layer after the pooling layer;
continuously extracting the characteristics of the reshaped similar frequency characteristics based on the two convolution layers after the data reshaping layer;
inputting the output characteristics of the last convolutional layer in the end-to-end stacked convolutional neural network model into a full connection layer;
and acquiring a preliminary diagnosis result corresponding to the sound signal based on a Softmax function according to the output of the full connection layer.
7. The method for diagnosing the fault of the high-speed rail traveling equipment according to any one of claims 1 to 6, wherein the step of extracting the fault feature from the collected multiple kinds of raw data signals of the high-speed rail traveling equipment based on a signal analysis method further comprises the following steps of:
preprocessing each original data signal according to an interface for collecting each original data signal; wherein the pre-processing comprises signal conversion and cleaning;
and performing data fusion on the processed original data signal based on an improved evidence theory fusion algorithm.
8. The utility model provides a high-speed railway driving equipment failure diagnosis device which characterized in that includes:
the extraction module is used for extracting fault characteristics from one or more collected original data signals of the high-speed rail travelling equipment based on a signal analysis method;
the diagnosis module is used for carrying out preliminary fault diagnosis on the high-speed rail driving equipment based on a classifier according to the fault characteristics of each original data signal and obtaining a preliminary diagnosis result corresponding to each original data signal;
and the fusion module is used for fusing the preliminary diagnosis results corresponding to all the original data signals based on a multi-sensor fusion algorithm to obtain the final diagnosis result of the high-speed rail driving equipment.
9. An electronic device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the program implements the steps of the method for diagnosing a malfunction of a driving device for a high-speed rail according to any one of claims 1 to 7.
10. A non-transitory computer readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the method for diagnosing a malfunction of a high-speed rail traveling apparatus according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010027587.0A CN111199257A (en) | 2020-01-10 | 2020-01-10 | Fault diagnosis method and device for high-speed rail driving equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010027587.0A CN111199257A (en) | 2020-01-10 | 2020-01-10 | Fault diagnosis method and device for high-speed rail driving equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111199257A true CN111199257A (en) | 2020-05-26 |
Family
ID=70747144
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010027587.0A Pending CN111199257A (en) | 2020-01-10 | 2020-01-10 | Fault diagnosis method and device for high-speed rail driving equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111199257A (en) |
Cited By (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111931806A (en) * | 2020-06-23 | 2020-11-13 | 广州杰赛科技股份有限公司 | Equipment fault diagnosis method and device for multi-sensor data fusion |
| CN112660211A (en) * | 2021-01-16 | 2021-04-16 | 湖南科技大学 | Intelligent operation and maintenance management system for railway locomotive |
| CN112733588A (en) * | 2020-08-13 | 2021-04-30 | 精英数智科技股份有限公司 | Machine running state detection method and device and electronic equipment |
| CN113326896A (en) * | 2021-06-25 | 2021-08-31 | 国网上海市电力公司 | Fusion sensing method based on multiple types of sensors |
| CN113343855A (en) * | 2021-06-09 | 2021-09-03 | 西南交通大学 | Rolling bearing fault diagnosis system and method based on guide type sub-field self-adaption |
| CN113468210A (en) * | 2021-06-08 | 2021-10-01 | 上海交通大学 | Robot fault diagnosis method and system based on characteristic engineering |
| CN113532138A (en) * | 2021-07-06 | 2021-10-22 | 广东工业大学 | Roller kiln sintering zone difference detection algorithm based on decision fusion framework |
| CN113655341A (en) * | 2021-09-10 | 2021-11-16 | 国网山东省电力公司鱼台县供电公司 | Power distribution network fault positioning method and system |
| CN113962261A (en) * | 2021-10-21 | 2022-01-21 | 中国人民解放军空军航空大学 | Depth network model for radar signal sorting |
| CN114167837A (en) * | 2021-12-02 | 2022-03-11 | 中国路桥工程有限责任公司 | Intelligent fault diagnosis method and system for railway signal system |
| CN114625110A (en) * | 2022-03-25 | 2022-06-14 | 上海富欣智能交通控制有限公司 | Fault diagnosis method, device and system and intelligent rail transit system |
| CN117176507A (en) * | 2023-11-02 | 2023-12-05 | 上海鉴智其迹科技有限公司 | Data analysis method, device, electronic equipment and storage medium |
| CN113962261B (en) * | 2021-10-21 | 2024-05-14 | 中国人民解放军空军航空大学 | Deep network model construction method for radar signal sorting |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101110155A (en) * | 2007-08-27 | 2008-01-23 | 北京交通大学 | Built-in intelligent fault diagnosing device based on data inosculating pattern recognition and method thereof |
| CN103901882A (en) * | 2014-04-15 | 2014-07-02 | 北京交通大学 | Online monitoring fault diagnosis system and method of train power system |
| CN104833534A (en) * | 2015-04-21 | 2015-08-12 | 广州市地下铁道总公司 | Train running fault diagnosis device based on multi-source information fusion, and method |
| CN106772080A (en) * | 2016-12-21 | 2017-05-31 | 哈尔滨工业大学 | Space lithium ion battery accelerated degradation test time equivalence modeling method |
| CN107832913A (en) * | 2017-10-11 | 2018-03-23 | 微梦创科网络科技（中国）有限公司 | The Forecasting Methodology and system to monitoring data trend based on deep learning |
| JP2019129593A (en) * | 2018-01-24 | 2019-08-01 | 東芝三菱電機産業システム株式会社 | Preventive maintenance device |
-
2020
- 2020-01-10 CN CN202010027587.0A patent/CN111199257A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101110155A (en) * | 2007-08-27 | 2008-01-23 | 北京交通大学 | Built-in intelligent fault diagnosing device based on data inosculating pattern recognition and method thereof |
| CN103901882A (en) * | 2014-04-15 | 2014-07-02 | 北京交通大学 | Online monitoring fault diagnosis system and method of train power system |
| CN104833534A (en) * | 2015-04-21 | 2015-08-12 | 广州市地下铁道总公司 | Train running fault diagnosis device based on multi-source information fusion, and method |
| CN106772080A (en) * | 2016-12-21 | 2017-05-31 | 哈尔滨工业大学 | Space lithium ion battery accelerated degradation test time equivalence modeling method |
| CN107832913A (en) * | 2017-10-11 | 2018-03-23 | 微梦创科网络科技（中国）有限公司 | The Forecasting Methodology and system to monitoring data trend based on deep learning |
| JP2019129593A (en) * | 2018-01-24 | 2019-08-01 | 東芝三菱電機産業システム株式会社 | Preventive maintenance device |
Non-Patent Citations (4)
| Title |
|---|
| 姚雪梅: ""多源数据融合的设备状态监测与智能诊断研究"" * |
| 姚雪梅: ""多源数据融合的设备状态监测与智能诊断研究"", 《中国博士学位论文全文数据库 工程科技Ⅱ辑》, no. 05, pages 2 * |
| 张双江: ""基于XGBoost和LSTM的智能监控系统的设计与实现"", 《中国优秀硕士学位论文全文数据库 信息科技辑》, vol. 1, no. 07, pages 20 * |
| 甘祖旺;: "加速寿命试验在产品变工况下退化数据归一化中的应用研究", 自动化与信息工程, no. 06, pages 23 - 28 * |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111931806A (en) * | 2020-06-23 | 2020-11-13 | 广州杰赛科技股份有限公司 | Equipment fault diagnosis method and device for multi-sensor data fusion |
| CN112733588A (en) * | 2020-08-13 | 2021-04-30 | 精英数智科技股份有限公司 | Machine running state detection method and device and electronic equipment |
| CN112660211A (en) * | 2021-01-16 | 2021-04-16 | 湖南科技大学 | Intelligent operation and maintenance management system for railway locomotive |
| CN113468210A (en) * | 2021-06-08 | 2021-10-01 | 上海交通大学 | Robot fault diagnosis method and system based on characteristic engineering |
| CN113343855B (en) * | 2021-06-09 | 2022-09-16 | 西南交通大学 | Rolling bearing fault diagnosis system and method based on guide type sub-field self-adaption |
| CN113343855A (en) * | 2021-06-09 | 2021-09-03 | 西南交通大学 | Rolling bearing fault diagnosis system and method based on guide type sub-field self-adaption |
| CN113326896A (en) * | 2021-06-25 | 2021-08-31 | 国网上海市电力公司 | Fusion sensing method based on multiple types of sensors |
| CN113532138A (en) * | 2021-07-06 | 2021-10-22 | 广东工业大学 | Roller kiln sintering zone difference detection algorithm based on decision fusion framework |
| CN113655341A (en) * | 2021-09-10 | 2021-11-16 | 国网山东省电力公司鱼台县供电公司 | Power distribution network fault positioning method and system |
| CN113655341B (en) * | 2021-09-10 | 2024-01-23 | 国网山东省电力公司鱼台县供电公司 | Fault positioning method and system for power distribution network |
| CN113962261A (en) * | 2021-10-21 | 2022-01-21 | 中国人民解放军空军航空大学 | Depth network model for radar signal sorting |
| CN113962261B (en) * | 2021-10-21 | 2024-05-14 | 中国人民解放军空军航空大学 | Deep network model construction method for radar signal sorting |
| CN114167837B (en) * | 2021-12-02 | 2023-09-15 | 中国路桥工程有限责任公司 | Intelligent fault diagnosis method and system for railway signal system |
| CN114167837A (en) * | 2021-12-02 | 2022-03-11 | 中国路桥工程有限责任公司 | Intelligent fault diagnosis method and system for railway signal system |
| CN114625110A (en) * | 2022-03-25 | 2022-06-14 | 上海富欣智能交通控制有限公司 | Fault diagnosis method, device and system and intelligent rail transit system |
| CN114625110B (en) * | 2022-03-25 | 2023-08-29 | 上海富欣智能交通控制有限公司 | Fault diagnosis method, device and system and intelligent rail transit system |
| CN117176507A (en) * | 2023-11-02 | 2023-12-05 | 上海鉴智其迹科技有限公司 | Data analysis method, device, electronic equipment and storage medium |
| CN117176507B (en) * | 2023-11-02 | 2024-02-23 | 上海鉴智其迹科技有限公司 | Data analysis method, device, electronic equipment and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111199257A (en) | Fault diagnosis method and device for high-speed rail driving equipment | |
| RU2540830C2 (en) | Adaptive remote maintenance of rolling stocks | |
| CN110390262A (en) | Video analysis method, apparatus, server and storage medium | |
| CN114585983B (en) | Method, device and system for detecting abnormal operation state of equipment | |
| CN110264440B (en) | Large-scale train displacement fault detection method and system based on deep learning | |
| Liu et al. | Industrial AI enabled prognostics for high-speed railway systems | |
| CN103699698A (en) | Method and system for track traffic failure recognition based on improved Bayesian algorithm | |
| Zang et al. | Methods for fault diagnosis of high-speed railways: A review | |
| CN114267178B (en) | Intelligent operation maintenance method and device for station | |
| CN117172414A (en) | Building curtain construction management system based on BIM technology | |
| CN114297935A (en) | Airport terminal building departure optimization operation simulation system and method based on digital twin | |
| KR20200142993A (en) | Diagnosing and modeling method of an engine condition | |
| CN114707401A (en) | Fault early warning method and device for signal system equipment | |
| Brahimi et al. | Development of a prognostics and health management system for the railway infrastructure—Review and methodology | |
| CN106682835A (en) | Data-driven complex electromechanical system service quality state evaluation method | |
| CN110737976A (en) | mechanical equipment health assessment method based on multi-dimensional information fusion | |
| CN112862233A (en) | Fault relevance analysis system and method based on Internet of vehicles data | |
| Shubinsky et al. | Application of machine learning methods for predicting hazardous failures of railway track assets | |
| Ji et al. | Rail track condition monitoring: A review on deep learning approaches | |
| Ochkasov et al. | Usage of intelligent technologies in choosing the strategy of technical maintenance of locomotives | |
| CN116842379A (en) | Mechanical bearing residual service life prediction method based on DRSN-CS and BiGRU+MLP models | |
| Zeng et al. | Rail break prediction and cause analysis using imbalanced in-service train data | |
| CN115623004A (en) | Block chain-based rail transit equipment management system and method | |
| CN113807704A (en) | Intelligent algorithm platform construction method for urban rail transit data | |
| Sharma et al. | The fundamentals and strategies of maintenance, repair, and overhaul (MRO) in Industry 4.0 |
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 |