Railway vehicle traction filter screen detection method, device and equipment and storage medium

Abstract

The invention discloses a railway vehicle traction filter screen detection method, device and equipment and a storage medium, and relates to the technical field of railway vehicles. The railway vehicle traction filter screen detection method comprises the steps that working condition parameters of a traction system in a railway train during normal operation are obtained; preprocessing the working condition parameters according to a preset data cleaning sequence to obtain a preprocessed data set; executing a preset normalization processing operation on the data set; and carrying out multiple times of sampling from the processed data set by using a density-based clustering algorithm, and training a density-based clustering model. The method solves the problems that in current actual subway operation and maintenance work, blockage of a filter screen of the traction converter is an important reason for abnormal temperature rise of the traction converter, and due to the fact that different line conditions are greatly different, and different operation environments such as tunnels and viaducts exist, a traditional pressure or temperature threshold value judgment method is poor in effect; and automatic learning cannot be carried out according to changes of line conditions.

Description

Technical Field

[0001] The present invention relates to the technical field of rail vehicles, and in particular to a method, device, equipment and storage medium for detecting a traction filter of a rail vehicle. Background Art

[0002] The traction system is the core component of vehicle operation. Currently, the traction inverter will experience abnormal temperature rise during train operation. In severe cases, the traction inverter may even stop working, causing partial traction loss in the traction system and affecting operations.

[0003] At present, in the actual subway operation and maintenance work, the blockage of the traction inverter filter is an important reason for the abnormal temperature rise of the traction inverter. Due to the great differences in conditions of different lines and different operating environments such as tunnels and elevated roads, the traditional pressure or temperature threshold judgment method is not effective and cannot automatically learn according to changes in line conditions. No effective solution has been proposed yet. Summary of the invention

[0004] Purpose of the invention: To provide a rail vehicle traction filter detection method, device, equipment and storage medium to solve the above-mentioned problems existing in the prior art.

[0005] Technical solution: A method for detecting a traction filter of a rail vehicle, comprising: Obtaining operating parameters of the traction system in a rail train during normal operation; Preprocessing the operating condition parameters according to a preset data cleaning sequence to obtain a preprocessed data set; Performing a preset normalization operation on the data set; Using density-based clustering algorithm to perform multiple sampling from the processed data set, a density-based clustering model is trained; The working condition parameters to be identified are input into the density-based clustering model, the prediction results are visualized, and the state of the traction filter represented by each packet of data is expressed according to the different locations of the points.

[0006] Preferably, obtaining the operating parameters of the traction system in the rail train during normal operation includes: Traction converter temperature, traction system fault signal, traction system cooling system status, traction cooling fan gear, traction cooling fan speed, traction cooling fan current and traction converter temperature rise rate.

[0007] Preferably, the operating condition parameters are preprocessed according to a preset data cleaning sequence to obtain a preprocessed data set, including: Taking the pulling instruction as the first data cleaning sequence bit, determining whether the pulling instruction exists; If yes, the traction system fault signal is used as the second data cleaning sequence bit, and it is determined whether there is a fault; If not, the cooling system status is used as the third data cleaning sequence bit, and it is determined whether it is working; If yes, the gear position of the cooling fan of the traction system is used as the fourth data cleaning sequence position, and it is determined whether it is in the high speed gear position; If so, a preset number of data sets are collected.

[0008] Preferably, after preprocessing the operating condition parameters according to a preset data cleaning sequence to obtain the preprocessed data set, before performing a preset normalization operation on the data set, the method further includes: The obtained data set is labeled and divided according to a preset ratio, and divided into a training set for model training and a test set for model testing; wherein the data set includes: traction converter temperature, traction cooling fan current, and traction converter temperature rise rate.

[0009] Preferably, performing a preset normalization operation on the data set includes: Perform preset target data extraction and data normalization operations on the normalized data set.

[0010] Preferably, a density-based clustering algorithm is used to perform multiple sampling from the processed data set to train a density-based clustering model, including: The fit_predict method is used to cluster the training set to obtain the cluster label to which each data point belongs, and the degree of congestion corresponding to the preset level is calculated based on the congestion status.

[0011] Preferably, a density-based clustering algorithm is used to perform multiple sampling from the processed data set to train a density-based clustering model, and further includes: Dynamically adjust the parameters of DBSCAN to achieve the best clustering results for different data environments by using genetic algorithms or grid search algorithms; or By utilizing the changing characteristics of data distribution and combining it with an adaptive clustering algorithm, the density parameters of DBSCAN are adjusted according to different time periods or data levels to ensure that the level of congestion can be accurately identified under different conditions.

[0012] Preferably, a density-based clustering algorithm is used to perform multiple sampling from the processed data set to train a density-based clustering model, and further includes: The density-based clustering algorithm is combined with a time series analysis algorithm to capture dynamic changes; wherein the time series analysis algorithm includes: an autoregressive integrated moving average algorithm or a long short-term memory network algorithm.

[0013] Preferably, after inputting the working condition parameters to be identified into the density-based clustering model, visualizing the prediction results, and expressing the traction filter state represented by each packet of data according to different point positions, the method further includes: When cluster modeling is completed, the characterization status of the corresponding filter is reported according to the data packets received at the preset time.

[0014] In order to achieve the above-mentioned purpose, according to another aspect of the present application, a rail vehicle traction filter detection device is provided.

[0015] The rail vehicle traction filter detection device according to the present application includes: An acquisition module is used to acquire operating parameters of a traction system in a rail train during normal operation; A preprocessing module, used to preprocess the operating condition parameters according to a preset data cleaning sequence to obtain a preprocessed data set; A normalization module, used to perform a preset normalization processing operation on the data set; A model training module is used to train a density-based clustering model by sampling multiple times from the processed data set using a density-based clustering algorithm; The prediction and presentation result module is used to input the working condition parameters to be identified into the density-based clustering model, visualize the prediction results, and express the traction filter status represented by each package of data according to different point locations.

[0016] In order to achieve the above-mentioned purpose, according to another aspect of the present application, an electronic device is provided, which includes: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the rail vehicle traction filter detection method described in any one of the present invention.

[0017] In order to achieve the above-mentioned purpose, according to another aspect of the present application, a computer-readable storage medium is provided, in which computer instructions are stored, and the computer instructions are used to implement the rail vehicle traction filter detection method described in any one of the present invention when executed by a processor.

[0018] Beneficial effect: In the embodiment of the present application, a method of establishing a deep learning framework based on improved clustering is adopted, by obtaining the operating parameters of the traction system in the rail train during normal operation; preprocessing the operating parameters according to the preset data cleaning order to obtain the preprocessed data set; performing a preset normalization processing operation on the data set; using a density-based clustering algorithm to perform multiple sampling from the processed data set to train a density-based clustering model; inputting the operating parameters to be identified into the density-based clustering model, visualizing the prediction results, and expressing the traction filter state represented by each packet of data according to different point positions, thereby achieving the purpose of automatic detection, diagnosis and analysis of the traction inverter filter, thereby achieving the technical effect of helping maintenance personnel to realize filter state maintenance and providing a basis for fault location when the inverter has an abnormal temperature rise fault, thereby solving the current actual subway operation and maintenance work. The traction inverter filter blockage is an important reason for the abnormal temperature rise of the traction inverter. Due to the great differences in different line conditions and different operating environments such as tunnels and elevated roads, the traditional pressure or temperature threshold judgment method is not effective and cannot automatically learn according to changes in line conditions. Technical problems. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a schematic flow chart of a method for detecting a traction filter for a rail vehicle according to an embodiment of the present application; Figure 2 It is a flow chart of a method for detecting a traction filter of a railway vehicle according to a preferred embodiment of the present application; Figure 3 This is a DBSCAN test set effect diagram of the rail vehicle traction filter detection method according to an embodiment of the present application; Figure 4 is a structural schematic diagram of a rail vehicle traction filter detection device according to an embodiment of the present application; and Figure 5 It is a structural schematic diagram of an electronic device of a rail vehicle traction filter detection method according to an embodiment of the present application. DETAILED DESCRIPTION

[0020] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present application.

[0021] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present application described here. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0022] In addition, the terms "installed", "set", "provided with", "connected", "connected", and "socketed" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral structure; it can be a mechanical connection or an electrical connection; it can be a direct connection, or an indirect connection through an intermediate medium, or it can be an internal connection between two devices, elements, or components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0023] It should be noted that, in the absence of conflict, the embodiments and features in the embodiments of the present application can be combined with each other. The present application will be described in detail below with reference to the accompanying drawings and in combination with the embodiments.

[0024] According to an embodiment of the present invention, a method for detecting a traction filter of a rail vehicle is provided. Figure 1-2 As shown, the method includes the following steps S101 to S105: Step S101, obtaining operating parameters of a traction system in a rail train during normal operation; By collecting the operating parameters of the train during normal operation in real time, as well as the train operation status data, accurate image data collection can be achieved, thus providing a basis for subsequent processing.

[0025] According to an embodiment of the present invention, preferably, obtaining the operating condition parameters of the traction system in the rail train during normal operation includes: Traction converter temperature, traction system fault signal, traction system cooling system status, traction cooling fan gear, traction cooling fan speed, traction cooling fan current and traction converter temperature rise rate.

[0026] It is important to know that voltage sensors and current transformers are used to monitor the voltage and current changes of the traction system. The sensors are installed at the input and output ends of the traction converter. The data is converted from the analog signal of the sensor into a digital signal and transmitted to the data acquisition system (DAQ).

[0027] After monitoring electrical parameters simultaneously through voltage sensors and current sensors, the power analyzer is used to calculate the power factor and traction power of the traction system in real time. The power factor is obtained by the ratio of active power to reactive power, and the traction power can be directly calculated from the voltage and current.

[0028] At the output end of the traction inverter, voltage sensors and current transformers are installed to monitor the AC power parameters output by the inverter. The data is also transmitted to the data acquisition system or the built-in monitoring module of the inverter.

[0029] The torque sensor is installed on the shaft of the traction motor and directly measures the output torque of the motor.

[0030] Thermocouples or thermistors are embedded in key components of traction motors and inverters to monitor temperature in real time, and the data is transmitted to the control system through the sensor interface or the built-in temperature monitoring module of the inverter.

[0031] Temperature sensors similar to thermocouples are installed at key locations in the cooling system, such as the coolant inlet, outlet, and around the fan, to monitor temperature changes in the cooling system.

[0032] The operating current of the cooling fan is monitored using a current sensor, usually measured through the electrical circuit connected to the fan.

[0033] Speed ​​sensors or tachometers are used to monitor the speed of cooling fans and are usually connected directly to the fan shaft or obtain speed information through non-contact optical sensors.

[0034] Step S102, preprocessing the operating condition parameters according to a preset data cleaning sequence to obtain a preprocessed data set; It can achieve accurate data cleaning effects, thereby obtaining the required data, and provide a data basis for subsequent accurate model training.

[0035] According to an embodiment of the present invention, preferably, preprocessing the operating condition parameters according to a preset data cleaning sequence to obtain a preprocessed data set includes: Taking the pulling instruction as the first data cleaning sequence bit, determining whether the pulling instruction exists; If yes, the traction system fault signal is used as the second data cleaning sequence bit, and it is determined whether there is a fault; If not, the cooling system status is used as the third data cleaning sequence bit, and it is determined whether it is working; If yes, the gear position of the cooling fan of the traction system is used as the fourth data cleaning sequence position, and it is determined whether it is in the high speed gear position; If so, a preset number of data sets are collected.

[0036] Specifically, the pull instruction is used as the first data cleaning sequence bit Determine whether the traction instruction exists: First, 1. Check the traction command in the traction system to confirm whether a valid traction command has been issued. The traction command is the starting condition for the system operation. If the command does not exist or is invalid, the data will be directly discarded or marked as invalid.

[0037] If the traction instruction exists (the system is operating normally), continue to the next step of cleaning; otherwise, stop processing this data.

[0038] 2. Use the traction system fault signal as the second data cleaning sequence Determine whether there is a fault in the traction system: Check whether there are any fault signals in the traction system, usually through the fault detection module in the system (such as overvoltage, overcurrent or short circuit faults).

[0039] If there is a fault signal, it means that the system is operating abnormally and the related data may not have reference value. In this case, the data is marked as a fault state and discarded.

[0040] If there is no fault signal, it indicates that the system is in normal operation and continues to process data.

[0041] 3. Use the cooling system status as the third data cleaning order Determine whether the cooling system is working: The cooling system is a critical part of the proper functioning of the traction system.

[0042] Check the status signal of the cooling system to confirm whether the cooling system is in working condition. This can be determined by monitoring indicators such as the cooling fan, motor temperature, and coolant flow data.

[0043] If the cooling system works normally, proceed to the next step of cleaning; otherwise, discard the data or mark it as invalid.

[0044] 4. Use the cooling fan gear position of the traction system as the fourth data cleaning sequence position Determine whether the cooling fan is in high speed gear: If the cooling system is working properly, further check the gear of the cooling fan. The cooling fan usually has a multi-speed adjustment function to adjust the fan speed according to temperature requirements.

[0045] If the cooling fan is in high speed gear, it means that the system is under high load and has the maximum cooling effect. If the cooling fan is not in high speed gear, the data is marked as low cooling state, or the data is discarded to avoid interference from inefficient cooling.

[0046] If the fan is in high speed gear, it means that the system is working in a normal and efficient state, and proceed to the next step of data collection.

[0047] 5. Collect a preset number of data sets After the previous cleaning steps, all states of the system meet normal operating conditions, that is, the traction command is valid, there is no fault, the cooling system is normal and the fan is working at high speed.

[0048] In this case, a preset amount of operating data is collected from the system, such as traction current, voltage, cooling fan current, temperature, fan speed, traction power, axle speed, etc. The data is used for subsequent model training or anomaly detection analysis.

[0049] Step S103, performing a preset normalization operation on the data set; Normalize the data samples, that is, map the data to (0, 1), improve the convergence speed of the next residual network, and reduce the impact of different ranges of data on the model. The normalization formula is as follows: .

[0050] According to an embodiment of the present invention, preferably, after preprocessing the operating condition parameters according to a preset data cleaning sequence, obtaining the preprocessed data set, and before performing a preset normalization operation on the data set, the method further includes: The obtained data set is labeled and divided according to a preset ratio, and divided into a training set for model training and a test set for model testing; wherein the data set includes: traction converter temperature, traction cooling fan current, and traction converter temperature rise rate.

[0051] Specifically, the data set is annotated with images and divided into training and test sets, and 900 training images and 300 test images are prepared for each data set. Of course, the actual ratio of the test set to the training set can be set according to actual usage requirements, and is not limited in this application.

[0052] According to an embodiment of the present invention, preferably, after performing a preset normalization operation on the pantograph image, multiple sampling is performed from the processed pantograph image using a residual network algorithm, and before training a residual network model, the method further includes: Perform preset data cleaning, target data extraction and data normalization operations on the normalized data set.

[0053] Specifically, the purpose of data cleaning is to remove invalid or abnormal data to ensure the high quality and consistency of the data set. The preset cleaning steps include: 1.1 Missing value processing, check whether there are missing values ​​in the data set (such as null values, NaN, etc.).

[0054] Common processing methods include: Deleting rows with missing values: If the amount of data is sufficient, you can delete rows with missing values. Filling with mean, median, previous data, etc.: If the amount of data is small or there are few missing values, use mean or median to fill missing values.

[0055] 1.2 Outlier processing: Check for outliers or unreasonable data points, such as data with values ​​such as temperature and current that are outside the normal operating range of the device. Use rule-based methods to remove outliers, such as data with temperatures that are too high or too low, or fan currents that are outside the reasonable range. Statistical methods (such as z-score, IQR) can be used to detect and remove outliers, or thresholds can be set based on historical experience to define normal and abnormal data.

[0056] 1.3 Duplicate data processing, check whether there are duplicate rows in the data set, especially during the data collection process, due to sensor errors or duplicate records may cause data duplication. Delete duplicate data rows to ensure that each data is independent and meaningful.

[0057] 2. Target data extraction: extract key features for modeling and analysis from the cleaned data set. The specific steps are as follows: 2.1 Determine the target features and extract relevant features from the data set according to the model requirements and actual working conditions. For example, the data set of the traction system may contain multiple working condition parameters, such as traction converter temperature, cooling fan current, traction converter temperature rise rate, etc.

[0058] From these features, parameters related to the prediction task are selected as target data. For example, the temperature rise rate and fan current can be used as key variables to predict the cooling effect of the traction system.

[0059] 2.2 Time window segmentation: split the data of continuous time periods into a package with each 500 milliseconds as a time window. Ensure that the data is arranged in order and extract the target features in each package of data for further processing.

[0060] Each packet of data can contain multiple features, ensuring that the extracted features match the time dimension well.

[0061] 2.3 Label generation: If classification prediction is required (for example, predicting the degree of filter blockage), the data can be labeled based on preset rules or models (such as 1-4 levels of blockage).

[0062] The generation of labels is usually based on business experience or the numerical range of certain characteristics. For example, the rate of change of fan current may be directly related to the degree of dirtiness and blockage.

[0063] Step S104: using a density-based clustering algorithm to perform multiple sampling from the processed data set to train a density-based clustering model; It can achieve accurate model training results, thus providing a basis and guarantee for subsequent predictions.

[0064] According to an embodiment of the present invention, preferably, a density-based clustering algorithm is used to perform multiple sampling from the processed data set to train a density-based clustering model, including: The fit_predict method is used to cluster the training set to obtain the cluster label to which each data point belongs, and the degree of congestion corresponding to the preset level is calculated based on the congestion status.

[0065] According to an embodiment of the present invention, preferably, a density-based clustering algorithm is used to perform multiple sampling from the processed data set to train a density-based clustering model, and the method further includes: Dynamically adjust the parameters of DBSCAN to achieve the best clustering results for different data environments by using genetic algorithms or grid search algorithms; or By utilizing the changing characteristics of data distribution and combining it with an adaptive clustering algorithm, the density parameters of DBSCAN are adjusted according to different time periods or data levels to ensure that the level of congestion can be accurately identified under different conditions.

[0066] According to an embodiment of the present invention, preferably, a density-based clustering algorithm is used to perform multiple sampling from the processed data set to train a density-based clustering model, and the method further includes: The density-based clustering algorithm is combined with a time series analysis algorithm to capture dynamic changes; wherein the time series analysis algorithm includes: an autoregressive integrated moving average algorithm or a long short-term memory network algorithm.

[0067] Step S105: input the working condition parameters to be identified into the density-based clustering model, visualize the prediction results, and express the traction filter status represented by each packet of data according to different point positions.

[0068] It can achieve accurate result prediction effect and at the same time, achieve good visualization presentation effect.

[0069] According to an embodiment of the present invention, preferably, after inputting the working condition parameters to be identified into the density-based clustering model, visualizing the prediction results, and expressing the traction filter state represented by each packet of data according to different point positions, the method further includes: When cluster modeling is completed, the characterization status of the corresponding filter is reported according to the data packets received at the preset time.

[0070] From the above description, it can be seen that the present application achieves the following technical effects: In the embodiment of the present application, a deep learning framework is established based on improved clustering, by obtaining the operating parameters of the traction system in the rail train during normal operation; preprocessing the operating parameters according to a preset data cleaning order to obtain a preprocessed data set; performing a preset normalization operation on the data set; using a density-based clustering algorithm to perform multiple sampling from the processed data set to train a density-based clustering model; inputting the operating parameters to be identified into the density-based clustering model, visualizing the prediction results, and expressing the traction filter state represented by each packet of data according to different point positions, thereby achieving the purpose of automatic detection, diagnosis and analysis of the traction inverter filter, thereby achieving the technical effect of helping maintenance personnel to achieve filter state maintenance and providing a basis for fault location when the inverter has an abnormal temperature rise fault, thereby solving the current actual subway operation and maintenance work. The traction inverter filter blockage is an important reason for the abnormal temperature rise of the traction inverter. Due to the great differences in different line conditions and different operating environments such as tunnels and elevated roads, the traditional pressure or temperature threshold judgment method is not effective and cannot automatically learn according to changes in line conditions.

[0071] This application only needs to build a machine learning framework based on DBSCAN (density-based clustering algorithm) through traction cooling fan current, traction converter temperature, traction cooling system status, and traction cooling fan gear. Through data cleaning and combined with the self-learning ability of the clustering algorithm, the detection and diagnosis of the traction converter filter can be realized. The filter status is analyzed every day, which is respectively level I, II, III, and IV, among which level I is completely free of dirt and blockage, and level IV is completely blocked. It helps maintenance personnel to realize filter status maintenance and provide fault location basis when the converter has abnormal temperature rise failure.

[0072] According to actual data analysis, during the operation of the train, when the traction converter filter screen gradually becomes dirty and clogged, the traction converter temperature rise rate will change to a certain extent. According to the data, the traction converter temperature rise rate and temperature upper limit are increased. At the same time, the current value of the cooling fan will also change according to different gears. The higher the degree of dirtiness and blockage, the higher the current value of the cooling fan at the same speed will be.

[0073] like Figure 3 As shown in the figure, three-dimensional data is put into the DBSCAN framework for learning and two-dimensional visualization is performed. It is obvious that the data feature differences of four different levels of dirtiness and congestion can be seen.

[0074] The workflow of this application is as follows: First, data cleaning is performed, and data packets under train traction conditions are selected to filter out data sets without traction system failures. Data packets with normal cooling systems are filtered out, and data on the high-speed gear status of the cooling fan is obtained. The data is divided into 500ms packets, and 1200 available data are sorted out, including traction inverter temperature, cooling fan current, and traction inverter temperature rise rate. The data set is divided into 900 training sets and 300 test sets, and the data is normalized. Initialize the object dbscan, and use the fit_predict method to cluster the data training set to obtain the cluster label clusters to which each data point belongs. According to the clustering density, 1-4 levels of dirty and blocked points are formed, and a two-dimensional visualization is performed ( Figure 2 ). According to the test set data, the corresponding 1-4 level clusters are used to determine what level the current filter status belongs to and report it.

[0075] Specifically, 1. Data preparation and preprocessing Data collection and segmentation: The operation data of the traction filter is collected at a time interval of 500ms, and a total of 1,200 valid data are collected. The core features of these data include: Traction inverter temperature: reflects the actual operating temperature of the traction inverter.

[0076] Cooling fan current: The fan current can reflect the system load. Excessive current may indicate system blockage.

[0077] Traction inverter temperature rise rate: The temperature rise rate is a measure of the rate of temperature change. A faster temperature rise rate may mean that the system heat dissipation is hindered and the filter is dirty and clogged.

[0078] After the initial data collection, the data was split in a ratio of 9:3, with 900 data items used for training and 300 data items used for testing.

[0079] Data cleaning and normalization To ensure the accuracy of clustering, the data is first cleaned to remove any noise or abnormal data points. Next, each feature is normalized to ensure that all feature values ​​are in the same scale (such as [0, 1] interval). This can avoid the difference in dimensions of different features introducing deviations in distance calculation.

[0080] 2. Application of DBSCAN Clustering Algorithm DBSCAN model initialization The DBSCAN algorithm relies on two main parameters: eps: defines the distance radius at which a sample point is considered a neighborhood point. The appropriate eps value can usually be determined through tools such as K-distance graphs. For traction filter data, a reasonable choice of this parameter can ensure a clear division between dirty and normal states.

[0081] min_samples: Specifies the minimum number of data points in a cluster to prevent isolated points from being mistaken for clusters. This can be set based on empirical data in the system or the minimum number of samples during actual runtime.

[0082] The parameters are adjusted so that clustering can effectively distinguish the operation status of filters in different states.

[0083] DBSCAN model training The preprocessed 900 training data are input into the DBSCAN model and trained by the fit_predict method. The goal of this step is to assign the data to different clusters (cluster labels such as 0, 1, 2, 3, etc.) through density evaluation. In addition, DBSCAN will identify some discrete points, which are marked as noise points (labeled as -1), usually indicating that the data is abnormal or cannot be effectively classified.

[0084] Cluster labels and congestion levels According to the actual working principle of the traction filter and the results of DBSCAN clustering, different clusters can be associated with the degree of dirtiness and blockage. The data points in each cluster represent the operating status of similar feature patterns, which are then mapped to different degrees of dirtiness and blockage: Level 1 dirty blockage (filter clean): corresponds to the densest cluster identified by DBSCAN, usually indicating that the temperature rise rate is low, the cooling fan current is within the normal range, and the system runs stably without blockage. This cluster can be labeled 0.

[0085] Secondary dirty blockage (mild blockage): DBSCAN identifies a relatively low density cluster, with a slightly increased temperature rise rate and a slightly increased cooling fan current, but the system is still within an acceptable range. This cluster can be label 1.

[0086] Level 3 dirty blockage (moderate blockage): The cluster density decreases further, the temperature rise rate is high, and the fan current is close to the upper limit, indicating that the filter begins to be blocked and the cooling effect is reduced. This cluster can be label 2.

[0087] Level 4 dirty blockage (serious blockage): The cluster density is extremely low and even a large number of noise points appear. The temperature rise rate increases sharply, the cooling fan current load is extremely large, and the system is seriously blocked and needs to be processed immediately. The cluster can be labeled 3 and the noise point label is -1.

[0088] 3. Visualization of clustering results In order to understand the clustering effect of DBSCAN more intuitively, a two-dimensional feature map can be used to visualize the distribution of each cluster. Commonly used feature combinations are: X-axis: traction converter temperature Y-axis: cooling fan current. The specific setting can be based on actual conditions and is not limited in this application.

[0089] In two-dimensional space, different clusters are represented by different colors, while noise points are marked in gray or black. This visual display can help operation and maintenance personnel quickly identify whether the system is running within the normal range or whether there is a potential congestion problem in the system.

[0090] From these graphs, it is clear that some clusters are densely packed, indicating good filter condition, while other clusters are more sparsely distributed, indicating the presence of mild to severe clogging issues.

[0091] 4. Determination of the dirty and blocked state of the test set data DBSCAN prediction test set data Input 300 test set data into the trained DBSCAN model. The model will assign a cluster label to each data point, indicating which dirty and congested situation the running status of the data point corresponds to.

[0092] Classification of dirtiness and blockage According to the cluster label of each data point in the test set, the degree of dirtiness and blockage corresponding to the current filter status can be inferred: Cluster label 0: corresponds to the first level of dirty blockage, and the filter is clean.

[0093] Cluster label 1: corresponds to the second level of dirty blockage, the filter is slightly blocked.

[0094] Cluster label 2: corresponds to level 3 dirty blockage, the filter is moderately blocked.

[0095] Cluster label 3: corresponds to level 4 dirty blockage, the filter is seriously clogged and needs to be dealt with immediately.

[0096] Noise point (labeled as -1): indicates extreme blockage or system abnormality, and the system and filter should be checked immediately.

[0097] Generate filter status report Based on the prediction results, a detailed filter status report can be generated. The report will indicate: Is the filter in normal working condition (first or second level dirty or blocked)? Is there mild to moderate blockage (level 3 dirty blockage)? Or whether the system is already in a serious blockage state (level 4 dirty blockage or noise point).

[0098] The present invention also has the following beneficial effects: The present invention can be extended to all urban rail vehicles according to the application requirements of specific scenarios. It can safely and reliably diagnose the status of the traction inverter filter, can promptly report abnormal filter blockage during operation, can report the specific degree of filter dirtiness and blockage during maintenance, and provide cleaning / replacement suggestions.

[0099] The present application also relates to a rail vehicle traction filter detection device. Figure 4 As shown, the rail vehicle traction filter detection device includes: An acquisition module is used to acquire operating parameters of a traction system in a rail train during normal operation; By collecting the operating parameters of the train during normal operation in real time, as well as the train operation status data, accurate image data collection can be achieved, thus providing a basis for subsequent processing.

[0100] A preprocessing module, used to preprocess the operating condition parameters according to a preset data cleaning sequence to obtain a preprocessed data set; It can achieve accurate data cleaning effects, thereby obtaining the required data, and provide a data basis for subsequent accurate model training.

[0101] A normalization module, used to perform a preset normalization processing operation on the data set; The data samples are normalized, that is, the data is mapped to (0, 1), which improves the convergence speed of the next residual network and reduces the impact of data in different ranges on the model.

[0102] A model training module is used to train a density-based clustering model by sampling multiple times from the processed data set using a density-based clustering algorithm; It can achieve accurate model training results, thus providing a basis and guarantee for subsequent predictions.

[0103] The prediction and presentation result module is used to input the working condition parameters to be identified into the density-based clustering model, visualize the prediction results, and express the traction filter status represented by each package of data according to different point locations.

[0104] It can achieve accurate result prediction effect and at the same time, achieve good visualization presentation effect.

[0105] From the above description, it can be seen that the present application achieves the following technical effects: In the embodiment of the present application, a deep learning framework is established based on improved clustering, by obtaining the operating parameters of the traction system in the rail train during normal operation; preprocessing the operating parameters according to a preset data cleaning order to obtain a preprocessed data set; performing a preset normalization operation on the data set; using a density-based clustering algorithm to perform multiple sampling from the processed data set to train a density-based clustering model; inputting the operating parameters to be identified into the density-based clustering model, visualizing the prediction results, and expressing the traction filter state represented by each packet of data according to different point positions, thereby achieving the purpose of automatic detection, diagnosis and analysis of the traction inverter filter, thereby achieving the technical effect of helping maintenance personnel to achieve filter state maintenance and providing a basis for fault location when the inverter has an abnormal temperature rise fault, thereby solving the current actual subway operation and maintenance work. The traction inverter filter blockage is an important reason for the abnormal temperature rise of the traction inverter. Due to the great differences in different line conditions and different operating environments such as tunnels and elevated roads, the traditional pressure or temperature threshold judgment method is not effective and cannot automatically learn according to changes in line conditions.

[0106] like Figure 5 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, ROM 12 and RAM 13 are connected to each other through a bus 14. The input / output (I / 0) interface 15 is also connected to the bus 14.

[0107] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0108] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a rail vehicle traction filter detection method.

[0109] In some embodiments, the rail vehicle traction filter detection method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the rail vehicle traction filter detection method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to execute the rail vehicle traction filter detection method in any other appropriate manner (e.g., by means of firmware).

[0110] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0111] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0112] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, device, or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices, or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0113] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0114] The systems and techniques described herein may be implemented in a computing system that includes a backend component (e.g., as a data server), or a computing system that includes a middleware component (e.g., an application server), or a computing system that includes a frontend component (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with an implementation of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by digital data communication of any form or medium (e.g., a communication network). Examples of communication networks include: a local area network (LA), a wide area network (WA), a blockchain network, and the Internet.

[0115] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0116] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0117] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for detecting a traction filter for a rail vehicle, characterized in that: include: Obtaining operating parameters of the traction system in a rail train during normal operation; Preprocessing the operating condition parameters according to a preset data cleaning sequence to obtain a preprocessed data set; Performing a preset normalization operation on the data set; Using density-based clustering algorithm to perform multiple sampling from the processed data set, a density-based clustering model is trained; The working condition parameters to be identified are input into the density-based clustering model, the prediction results are visualized, and the state of the traction filter represented by each packet of data is expressed according to the different locations of the points.

2. The method for detecting a traction filter for a rail vehicle according to claim 1, characterized in that: Obtain the operating parameters of the traction system in the rail train during normal operation, including: Traction converter temperature, traction system fault signal, traction system cooling system status, traction cooling fan gear, traction cooling fan speed, traction cooling fan current and traction converter temperature rise rate.

3. The method for detecting a traction filter for a rail vehicle according to claim 2, characterized in that: The operating condition parameters are preprocessed according to a preset data cleaning sequence to obtain a preprocessed data set, including: Taking the pulling instruction as the first data cleaning sequence bit, determining whether the pulling instruction exists; If yes, the traction system fault signal is used as the second data cleaning sequence bit, and it is determined whether there is a fault; If not, the cooling system status is used as the third data cleaning sequence bit, and it is determined whether it is working; If yes, the gear position of the cooling fan of the traction system is used as the fourth data cleaning sequence position, and it is determined whether it is in the high speed gear position; If so, a preset number of data sets are collected.

4. The method for detecting a traction filter of a railway vehicle according to claim 1, characterized in that: After preprocessing the operating condition parameters according to a preset data cleaning sequence to obtain a preprocessed data set, and before performing a preset normalization operation on the data set, the method further includes: The obtained data set is labeled and divided according to a preset ratio, and divided into a training set for model training and a test set for model testing; wherein the data set includes: traction converter temperature, traction cooling fan current, and traction converter temperature rise rate.

5. The method for detecting a traction filter of a railway vehicle according to claim 1, characterized in that: Performing a preset normalization operation on the data set includes: Perform preset target data extraction and data normalization operations on the normalized data set.

6. The method for detecting a traction filter of a railway vehicle according to claim 3, characterized in that: The density-based clustering algorithm is used to perform multiple sampling from the processed data set to train a density-based clustering model, including: The fit_predict method is used to cluster the training set to obtain the cluster label to which each data point belongs, and the degree of congestion corresponding to the preset level is calculated based on the congestion status.

7. The method for detecting a traction filter for a rail vehicle according to claim 3, characterized in that: Using a density-based clustering algorithm to perform multiple sampling from the processed data set, a density-based clustering model is trained, which also includes: Dynamically adjust the parameters of DBSCAN to achieve the best clustering results for different data environments by using genetic algorithms or grid search algorithms; or By utilizing the changing characteristics of data distribution and combining it with an adaptive clustering algorithm, the density parameters of DBSCAN are adjusted according to different time periods or data levels to ensure that the level of congestion can be accurately identified under different conditions.

8. The method for detecting a traction filter of a railway vehicle according to claim 3, characterized in that: Using a density-based clustering algorithm to perform multiple sampling from the processed data set, a density-based clustering model is trained, which also includes: The density-based clustering algorithm is combined with a time series analysis algorithm to capture dynamic changes; wherein the time series analysis algorithm includes: an autoregressive integrated moving average algorithm or a long short-term memory network algorithm.

9. The method for detecting a traction filter of a railway vehicle according to claim 1, characterized in that: After inputting the working condition parameters to be identified into the density-based clustering model, visualizing the prediction results, and expressing the traction filter state represented by each packet of data according to different point positions, it also includes: When cluster modeling is completed, the characterization status of the corresponding filter is reported according to the data packets received at the preset time.

10. A rail vehicle traction filter detection device, characterized in that: include: An acquisition module is used to acquire operating parameters of a traction system in a rail train during normal operation; A preprocessing module, used to preprocess the operating condition parameters according to a preset data cleaning sequence to obtain a preprocessed data set; A normalization module, used to perform a preset normalization processing operation on the data set; A model training module is used to train a density-based clustering model by sampling multiple times from the processed data set using a density-based clustering algorithm; The prediction and presentation result module is used to input the working condition parameters to be identified into the density-based clustering model, visualize the prediction results, and express the traction filter status represented by each package of data according to different point locations.

11. An electronic device, characterized in that The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the rail vehicle traction filter detection method described in any one of claims 1-9.

12. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the rail vehicle traction filter detection method described in any one of claims 1-9 when executed.

Images