Intelligent monitoring and alarming method and system for abnormal vibration of IO equipment connector
By combining miniature vibration sensors and probabilistic neural network models, real-time monitoring and accurate identification of abnormal vibrations in train I/O device connectors are achieved. This solves the problem of poor detection results in traditional methods, improves the accuracy of fault diagnosis and equipment operating efficiency, and reduces maintenance costs.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CRRC ZHUZHOU ELECTRIC LOCOMOTIVE RESEARCH INSTITUTE CO LTD
- Filing Date
- 2024-10-23
- Publication Date
- 2026-04-24
AI Technical Summary
Existing technologies struggle to accurately identify minute abnormal vibrations in train I/O device connectors, resulting in poor fault detection, increased unnecessary troubleshooting costs and misjudgments, and traditional vibration analysis methods lack flexibility and cannot effectively distinguish between normal and abnormal vibration modes.
By employing a miniature vibration sensor combined with a probabilistic neural network model, abnormal vibrations of I/O device connectors can be identified in real time through monitoring, signal conditioning, data preprocessing, and pattern matching, and timely alarms can be provided through an abnormality alarm platform.
It improves the accuracy of fault detection, reduces system downtime and maintenance costs, provides data-driven fault analysis support, and is highly adaptable and scalable, suitable for different types of I/O device connectors.
Smart Images
Figure CN121917047A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of analog input / output and digital input / output of trains in the rail transit industry, specifically to an intelligent monitoring and alarm method and system for abnormal vibration of I / O device connectors. Background Technology
[0002] In modern train systems, the stability of I / O (input / output) devices is crucial to the reliable operation of the entire vehicle. These devices are typically connected to other parts of the train via connectors. However, trains inevitably experience various vibrations during operation, especially at high speeds or when traversing uneven tracks. These vibrations can cause connectors to loosen temporarily, resulting in brief anomalies in digital or analog input / output signals.
[0003] Traditional fault detection methods typically rely on circuit testing or manual inspection. These methods have limited effectiveness in detecting minor, irregular vibration faults and often fail to accurately determine whether such anomalies are caused by a temporary loosening of the connector. This can lead to misdiagnosis as a hardware failure, thereby increasing unnecessary troubleshooting costs. The lack of an effective way to monitor and identify abnormal vibrations of I / O devices in real time limits the ability to prevent faults and maintain them promptly.
[0004] Furthermore, traditional vibration analysis methods often lack the ability to identify complex vibration modes and cannot effectively distinguish between normal and abnormal vibration modes. Moreover, manually setting thresholds or standards is not flexible enough for changing operating environments and may lead to false alarms or missed alarms. Summary of the Invention
[0005] The following provides a brief overview of one or more aspects to offer a basic understanding of them. This overview is not an exhaustive summary of all conceived aspects, nor is it intended to identify key or decisive elements of all aspects, nor to define the scope of any or all aspects. Its sole purpose is to present some concepts of one or more aspects in a simplified form to prepare for the more detailed descriptions that follow.
[0006] The purpose of this invention is to solve the above-mentioned problems and provide an intelligent monitoring and alarm method and system for abnormal vibration of IO device connectors. This not only improves the accuracy of fault diagnosis, but also greatly reduces the cost and time for designers to troubleshoot faults, thereby improving the efficiency and reliability of the entire train system.
[0007] The technical solution of this invention is as follows: This invention discloses an intelligent monitoring and alarm system for abnormal vibration of I / O device connectors. The system includes a hardware signal acquisition circuit, an abnormal vibration identification system, and an abnormal alarm platform, wherein:
[0008] The hardware signal acquisition circuit is used to monitor and capture the vibration signal of the monitored connector and convert the vibration signal into input data for the abnormal vibration identification system.
[0009] An abnormal vibration identification system is used to convert the voltage data output by the hardware signal acquisition circuit into waveform data and then decompose it into corresponding feature vectors, which are then input into a neural network model. The neural network model outputs the vibration state of the monitored connector, which includes abnormal vibration information.
[0010] An abnormal alarm platform is used to display abnormal vibration information and process alarm information.
[0011] According to an embodiment of the intelligent monitoring and alarm system for abnormal vibration of I / O device connectors of the present invention, the hardware signal acquisition circuit includes a vibration sensor, a signal conditioning module, and an acquisition board, wherein:
[0012] Vibration sensors are attached to the monitored connector to monitor and record the vibration state of the monitored connector. When the monitored connector vibrates, the vibration sensor will capture the physical vibration and convert it into a voltage signal.
[0013] The input of the signal conditioning module is connected to the vibration sensor, which receives the voltage signal output by the vibration sensor and performs signal conditioning, including amplifying the weak voltage signal and filtering out noise in the voltage signal.
[0014] The input terminal of the acquisition board is connected to the output terminal of the signal conditioning module, which is used to convert the voltage analog signal output by the signal conditioning module into voltage data.
[0015] According to an embodiment of the intelligent monitoring and alarm system for abnormal vibration of I / O device connectors of the present invention, the abnormal vibration identification system includes a data preprocessing module, a vibration feature vector analysis module, and a vibration state judgment module, wherein:
[0016] The data preprocessing module is used to receive voltage data transmitted from the acquisition board, convert it into waveform data after standardization, and then convert the waveform data into feature vectors.
[0017] The vibration feature vector analysis module inputs the preprocessed feature vectors into the neural network model. The input feature vectors are matched with the training samples stored in the model. The model achieves pattern matching by calculating the Euclidean distance or similarity between the input feature vectors and each training sample. Based on the pattern matching results, the model calculates the probability that the input feature vectors belong to each category.
[0018] The vibration state judgment module sets one or more thresholds for judging abnormalities during the model training phase or according to actual application needs. Then, it compares the abnormal probability calculated by the vibration feature vector analysis module with the set thresholds. If the abnormal probability exceeds the set threshold, the current vibration state is judged to be abnormal. If the abnormal probability does not exceed the threshold, the current vibration state is judged to be normal.
[0019] According to one embodiment of the intelligent monitoring and alarm system for abnormal vibration of I / O device connectors of the present invention, the neural network model is a product-based neural network model.
[0020] This invention also discloses an intelligent monitoring and alarm method for abnormal vibration of I / O device connectors, the method comprising:
[0021] Step S1: Monitor and capture the vibration signal of the monitored connector through the hardware signal acquisition circuit and convert the vibration signal into the input data required for abnormal vibration identification;
[0022] Step S2: After converting the voltage data output by the hardware signal acquisition circuit into waveform data through the abnormal vibration identification system, the abnormal vibration identification system decomposes it into corresponding feature vectors and inputs them into the neural network model. The neural network model outputs the vibration state of the monitored connector, which includes abnormal vibration information.
[0023] Step S3: Display abnormal vibration information and process alarm information through the abnormal alarm platform.
[0024] According to an embodiment of the intelligent monitoring and alarm method for abnormal vibration of I / O device connectors of the present invention, step S1 further includes:
[0025] Step S1-1: Attach the vibration sensor to the monitored connector to monitor and record the vibration state of the monitored connector. When the monitored connector vibrates, the vibration sensor will convert the captured physical vibration into a voltage signal.
[0026] Step S1-2: The voltage signal output by the vibration sensor is conditioned by the signal conditioning module. The signal conditioning includes amplifying the weak voltage signal and filtering out noise in the voltage signal.
[0027] Step S1-3: The analog voltage signal output by the signal conditioning module is converted into voltage data by the acquisition board so that it can be processed by the subsequent abnormal vibration identification step.
[0028] According to an embodiment of the intelligent monitoring and alarm method for abnormal vibration of I / O device connectors of the present invention, step S2 further includes:
[0029] Step S2-1: Receive voltage data transmitted from the acquisition board, convert it into waveform data after standardization, and convert the waveform data into feature vectors;
[0030] Step S2-2: Input the preprocessed feature vector into the neural network model. The input feature vector is matched with the training samples stored in the model. The model achieves pattern matching by calculating the Euclidean distance or similarity between the input feature vector and each training sample. Based on the pattern matching results, the model calculates the probability that the input feature vector belongs to each category.
[0031] Step S2-3: During the model training phase or according to actual application needs, set one or more thresholds for judging abnormalities. Then compare the abnormal probability calculated in the vibration feature vector analysis module with the set thresholds. If the abnormal probability exceeds the set threshold, the current vibration state is judged to be abnormal. If the abnormal probability does not exceed the threshold, the current vibration state is judged to be normal.
[0032] According to one embodiment of the intelligent monitoring and alarm method for abnormal vibration of I / O device connectors of the present invention, the neural network model is a product-based neural network model.
[0033] Compared with the prior art, the present invention has the following beneficial effects: In the solution of the present invention, the combination of micro vibration sensor and neural network provides a new possibility for real-time and accurate monitoring of the vibration status of IO device connector. The micro vibration sensor can accurately capture tiny physical vibrations, while the neural network model, especially the probabilistic neural network (PNN), can efficiently process and analyze a large amount of vibration data and accurately distinguish between normal and abnormal vibration modes.
[0034] In detail, the present invention has the following advantages:
[0035] 1. Improved fault detection accuracy: By integrating miniature vibration sensors and advanced neural network analysis, abnormal vibrations of I / O device connectors can be identified more accurately, significantly improving the accuracy of fault detection compared to traditional methods.
[0036] 2. Real-time monitoring and rapid response: It can monitor the status of IO device connectors in real time. Once abnormal vibration is detected, the alarm mechanism is activated immediately to prevent potential failures, reduce system downtime, and improve equipment operating efficiency.
[0037] 3. Reduce maintenance costs: Accurate fault prediction and timely alarms can reduce unnecessary maintenance and repair work, thereby reducing long-term maintenance costs.
[0038] 4. Data-driven fault analysis: It not only provides real-time monitoring, but also records historical vibration data, providing data support for maintenance personnel to help them better understand the causes and patterns of equipment failures and guide future maintenance strategies.
[0039] 5. High adaptability and scalability: It easily adapts to different types of I / O device connectors, and the system's scalability means that it can be integrated into larger monitoring systems to support overall device health management. Attached Figure Description
[0040] The above-described features and advantages of the present invention will be better understood after reading the following detailed description of embodiments of the present disclosure in conjunction with the accompanying drawings. In the drawings, components are not necessarily drawn to scale, and components having similar related properties or features may have the same or similar reference numerals.
[0041] Figure 1 A block diagram illustrating the overall principle of an embodiment of the intelligent monitoring and alarm system for abnormal vibration of I / O device connectors according to the present invention is shown.
[0042] Figure 2 It shows Figure 1 The schematic diagram of the hardware signal acquisition circuit in the system embodiment shown.
[0043] Figure 3 It shows Figure 1 The diagram shows the principle block diagram of the abnormal vibration identification system in the system embodiment shown.
[0044] Figure 4 It shows Figure 1 The flowchart of model training in the abnormal vibration identification system shown in the system embodiment is illustrated.
[0045] Figure 5 A flowchart of an embodiment of the intelligent monitoring and alarm method for abnormal vibration of I / O device connectors according to the present invention is shown.
[0046] Figure 6 It shows Figure 5 The flowchart illustrates the abnormal vibration identification process in the method embodiment shown. Detailed Implementation
[0047] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should be noted that the aspects described below with reference to the accompanying drawings and specific embodiments are merely exemplary and should not be construed as limiting the scope of protection of the present invention in any way.
[0048] Figure 1 The overall principle of an embodiment of the intelligent monitoring and alarm system for abnormal vibration of I / O device connectors according to the present invention is shown. Please refer to... Figure 1The system in this embodiment includes a hardware signal acquisition circuit, an abnormal vibration identification system, and an abnormal alarm platform.
[0049] The output of the hardware signal acquisition circuit is connected to the input of the abnormal vibration identification system, and the output of the abnormal vibration identification system is connected to the input of the abnormal alarm platform.
[0050] The hardware signal acquisition circuit is used to monitor and capture the vibration signal of the monitored connector and convert the vibration signal into input data for the abnormal vibration identification system.
[0051] like Figure 2 As shown, the hardware signal acquisition circuit includes a vibration sensor, a signal conditioning module, and an acquisition board. The vibration sensor is generally a miniature vibration sensor, attached to the monitored connector, used to monitor and record the vibration state of the monitored connector. When the monitored connector vibrates, the vibration sensor converts the captured physical vibration into a voltage signal.
[0052] The input of the signal conditioning module is connected to the vibration sensor to receive the voltage signal output by the vibration sensor. These voltage signals are usually very weak and need to be precisely conditioned for subsequent processing. The signal conditioning module's processing includes amplifying the weak voltage signal and filtering out noise in the voltage signal to ensure the clarity and accuracy of the signal.
[0053] The input terminal of the acquisition board is connected to the output terminal of the signal conditioning module. The signal conditioning module outputs a voltage analog signal, which the acquisition board uses to convert into voltage data so that it can be processed by the abnormal vibration identification system.
[0054] The abnormal vibration identification system is used to convert the voltage data output by the hardware signal acquisition circuit into waveform data and then decompose it into corresponding feature vectors, which are then input into a neural network model. The neural network model outputs the vibration state of the monitored connector, which includes abnormal vibration information.
[0055] like Figure 3 As shown, the abnormal vibration identification system further includes a data preprocessing module, a vibration feature vector analysis module, and a vibration state judgment module. The output of the data preprocessing module is connected to the input of the vibration feature vector analysis module, and the output of the vibration feature vector analysis module is connected to the input of the vibration state judgment module.
[0056] In the data preprocessing module, voltage data transmitted from the acquisition board is received, and after standardization, it is converted into waveform data. The waveform data is then converted into feature vectors, which typically include key parameters such as the frequency, amplitude, and phase of the vibration, as well as other possible time-domain or frequency-domain features.
[0057] The vibration feature vector analysis module inputs the preprocessed feature vectors into the neural network model. In this embodiment, the neural network model is, for example, a probabilistic neural network, including a PNN model (Product-based Neural Network). Then, in the PNN model, the input feature vectors are pattern-matched with the training samples (including normal and abnormal vibration feature vectors) stored within the model. PNN achieves pattern matching by calculating the Euclidean distance or similarity between the input feature vector and each training sample. Finally, based on the pattern matching results, the PNN model calculates the probability that the input feature vector belongs to each category (normal or abnormal).
[0058] During the PNN model training phase or based on actual application requirements, the vibration state judgment module sets one or more thresholds for judging abnormalities. These thresholds determine the probability at which an input feature vector is classified as an abnormal vibration state, at which point the system should consider it abnormal. The abnormal probability calculated in the vibration feature vector analysis module is then compared with the set thresholds. If the abnormal probability exceeds the set threshold, the current vibration state is determined to be abnormal. If the abnormal probability does not exceed the threshold, the current vibration state is determined to be normal.
[0059] In addition, the abnormal vibration identification system also has data recording and historical analysis functions. It can store historical vibration data and provide time series analysis to help maintenance personnel understand the trend of vibration changes over time. This is very important in preventive maintenance and fault analysis, and can help identify potential problems.
[0060] In an abnormal vibration identification system, the training process of the neural network model is as follows: Figure 4 As shown.
[0061] Step 1: Extract multiple vibration feature vectors as sample data.
[0062] Step 1 further includes the following processing:
[0063] Data collection: Collect vibration data from multiple vibration sensors, which should be installed on critical equipment or components, such as train I / O device connectors.
[0064] Data preprocessing: The collected raw vibration data is preprocessed, including noise reduction, filtering, and normalization, to improve data quality.
[0065] Feature extraction: Extract multiple vibration features from the preprocessed data. These features should be able to comprehensively reflect the state of vibration, such as frequency, amplitude, phase, and energy.
[0066] Sample construction: The extracted feature vectors are organized into a sample dataset, with each sample including a feature vector and a corresponding label (normal or abnormal).
[0067] Step 2: Select a specific neural network model, such as the PNN model.
[0068] Step 2 further includes the following processing:
[0069] Model selection: Choose a suitable neural network model based on the application scenario and data characteristics. In this example, the PNN (Product-based Neural Network) model is selected as the model for abnormal vibration identification.
[0070] Model configuration: Configure the parameters of the PNN model, such as the number of neurons in the input layer, hidden layer, and output layer, the choice of activation function, learning rate, etc.
[0071] Model initialization: The weights and biases of the PNN model are initialized, usually using a random initialization method.
[0072] Step 3: Select training samples from the sample data.
[0073] Step 3 further includes the following processing:
[0074] Data partitioning: The sample dataset is divided into a training set and a test set, with the training set usually making up the majority and the test set making up the minority.
[0075] Training sample selection: Randomly select training samples from the training set or select them according to certain rules to ensure the diversity and representativeness of the samples.
[0076] Data augmentation: If the number of training samples is insufficient, data augmentation techniques can be used, such as adding noise, rotation, scaling, etc., to increase the diversity of training samples.
[0077] Step 4: Input the training samples into the neural network model for model training.
[0078] Step 4 further includes the following processing:
[0079] Input data: Input the feature vectors of the training samples into the input layer of the PNN model.
[0080] Forward propagation: Calculates the predicted values of the output layer through the hidden layers of the PNN model.
[0081] Loss calculation: Calculate the loss function based on the predicted value and the true label, such as cross-entropy loss, mean squared error, etc.
[0082] Backpropagation: Based on the gradient of the loss function, the weights and biases of the PNN model are updated using the backpropagation algorithm.
[0083] Iterative training: Repeat the above steps until the preset number of iterations is reached or the loss function converges.
[0084] Step 5: Select test samples from the sample data.
[0085] Step 6: Input the test samples into the trained model to test the model.
[0086] Step 6 further includes the following processing:
[0087] Input data: Input the feature vectors of the test samples into the trained PNN model.
[0088] Predicted output: The predicted value of the output layer is calculated through the hidden layers of the PNN model.
[0089] Performance evaluation: Calculate the model's performance metrics, such as accuracy, recall, and F1 score, based on the predicted values and the true labels.
[0090] Step 7: Determine the difference between the training error and the preset threshold. If the training error is less than the preset threshold, the model training is complete; otherwise, proceed to step 3.
[0091] The abnormal vibration alarm platform is used to display and process abnormal vibration information, including displaying abnormal vibration information through a human-machine interface. Once the abnormal vibration identification system determines that an abnormal vibration state has occurred, the abnormal alarm platform immediately triggers the alarm mechanism. According to the preset alarm strategy, the abnormal alarm platform can issue alarms through various means, such as visual (e.g., warning messages on the display screen), sound (e.g., alarm sounds), or by sending alarm information to maintenance personnel's mobile devices.
[0092] The abnormal alarm platform records abnormal vibration events and related information (such as time, location, vibration characteristics, etc.) and may provide them to maintenance personnel as a reference for troubleshooting.
[0093] Figure 5 The flowchart of an embodiment of the intelligent monitoring and alarm method for abnormal vibration of I / O device connectors of the present invention is shown.
[0094] Please see Figure 5 The method in this embodiment includes:
[0095] Step S1: The vibration signal of the monitored connector is monitored and captured by the hardware signal acquisition circuit and converted into input data for the abnormal vibration identification system.
[0096] Step S1 further includes:
[0097] Step S1-1: Attach the vibration sensor (usually a miniature vibration sensor) to the connector being monitored, monitor and record the vibration state of the connector. When the connector vibrates, the vibration sensor converts the captured physical vibration into a voltage signal.
[0098] Step S1-2: The voltage signal output by the vibration sensor is conditioned by the signal conditioning module. The signal conditioning includes amplifying the weak voltage signal and filtering out noise in the voltage signal to ensure the clarity and accuracy of the signal.
[0099] Step S1-3: The analog voltage signal output by the signal conditioning module is converted into voltage data by the acquisition board so that it can be processed by the subsequent abnormal vibration identification step.
[0100] Step S2: After converting the voltage data output by the hardware signal acquisition circuit into waveform data through the abnormal vibration identification system, the abnormal vibration identification system decomposes it into corresponding feature vectors and inputs them into the neural network model. The neural network model outputs the vibration state of the monitored connector, which includes abnormal vibration information.
[0101] Figure 6 It shows Figure 5 The abnormal vibration identification process in the illustrated method embodiment further includes step S2:
[0102] Step S2-1: Receive voltage data transmitted from the acquisition board, convert it into waveform data after standardization, and convert the waveform data into feature vectors. These feature vectors usually include key parameters such as the frequency, amplitude, and phase of the vibration, as well as other possible time-domain or frequency-domain features.
[0103] Step S2-2: Input the preprocessed feature vector into the neural network model. In this embodiment, the neural network model is, for example, a probabilistic neural network, including a PNN model (Product-based Neural Network). Then, in the PNN model, the input feature vector is pattern-matched with the training samples (including normal and abnormal vibration feature vectors) stored within the model. PNN achieves pattern matching by calculating the Euclidean distance or similarity between the input feature vector and each training sample. Finally, based on the pattern matching results, the PNN model calculates the probability that the input feature vector belongs to each category (normal or abnormal).
[0104] Step S2-3: During the PNN model training phase or according to actual application needs, set one or more thresholds for judging anomalies. These thresholds determine the probability at which the input feature vector is classified as an abnormal vibration state, at which point the system should consider it an anomaly. Then, compare the anomaly probability calculated in step S2-2 with the set thresholds. If the anomaly probability exceeds the set threshold, the current vibration state is determined to be abnormal. If the anomaly probability does not exceed the threshold, the current vibration state is determined to be normal.
[0105] The training process of the neural network model has been described in detail in the foregoing embodiments and will not be repeated here.
[0106] Step S3: Display and process abnormal vibration information through the abnormal alarm platform, including displaying abnormal vibration information through the human-machine interface.
[0107] Once the abnormal vibration identification process determines that an abnormal vibration state has occurred, the abnormal alarm platform immediately triggers the alarm mechanism. According to the preset alarm strategy, the abnormal alarm platform can issue alarms through various means, such as visual (e.g., warning messages on the display screen), audible (e.g., alarm sounds), or by sending alarm information to maintenance personnel's mobile devices.
[0108] The abnormal alarm platform records abnormal vibration events and related information (such as time, location, vibration characteristics, etc.) and may provide them to maintenance personnel as a reference for troubleshooting.
[0109] Although the methods described above are illustrated and depicted as a series of actions for the sake of simplicity, it should be understood and appreciated that these methods are not limited by the order of the actions, as some actions may occur in a different order and / or concurrently with other actions from the illustrations and descriptions herein or not illustrated and described herein but which may be understood by those skilled in the art, according to one or more embodiments.
[0110] Those skilled in the art will further appreciate that the various illustrative logic blocks, modules, circuits, and algorithm steps described in conjunction with the embodiments disclosed herein can be implemented as electronic hardware, computer software, or a combination of both. To clearly illustrate this interchangeability between hardware and software, the various illustrative components, blocks, modules, circuits, and steps are described above in a generalized manner in terms of their functionality. Whether such functionality is implemented as hardware or software depends on the specific application and the design constraints imposed on the overall system. Those skilled in the art may implement the described functionality in different ways for each specific application, but such implementation decisions should not be construed as departing from the scope of the invention.
[0111] The prior description of this disclosure is provided to enable any person skilled in the art to make or use this disclosure. Various modifications to this disclosure will be apparent to those skilled in the art, and the general principles defined herein may be applied to other variations without departing from the spirit or scope of this disclosure. Therefore, this disclosure is not intended to be limited to the examples and designs described herein, but should be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims
1. An intelligent monitoring and alarm system for abnormal vibration of I / O device connectors, characterized in that, The system includes a hardware signal acquisition circuit, an abnormal vibration identification system, and an abnormal alarm platform, among which: The hardware signal acquisition circuit is used to monitor and capture the vibration signal of the monitored connector and convert the vibration signal into input data for the abnormal vibration identification system. An abnormal vibration identification system is used to convert the voltage data output by the hardware signal acquisition circuit into waveform data and then decompose it into corresponding feature vectors, which are then input into a neural network model. The neural network model outputs the vibration state of the monitored connector, which includes abnormal vibration information. An abnormal alarm platform is used to display abnormal vibration information and process alarm information.
2. The intelligent monitoring and alarm system for abnormal vibration of I / O device connectors according to claim 1, characterized in that, The hardware signal acquisition circuit includes a vibration sensor, a signal conditioning module, and an acquisition board, wherein: Vibration sensors are attached to the monitored connector to monitor and record the vibration state of the monitored connector. When the monitored connector vibrates, the vibration sensor will capture the physical vibration and convert it into a voltage signal. The input of the signal conditioning module is connected to the vibration sensor, which receives the voltage signal output by the vibration sensor and performs signal conditioning, including amplifying the weak voltage signal and filtering out noise in the voltage signal. The input terminal of the acquisition board is connected to the output terminal of the signal conditioning module, which is used to convert the voltage analog signal output by the signal conditioning module into voltage data.
3. The intelligent monitoring and alarm system for abnormal vibration of I / O device connectors according to claim 1, characterized in that, The abnormal vibration identification system includes a data preprocessing module, a vibration feature vector analysis module, and a vibration state judgment module, among which: The data preprocessing module is used to receive voltage data transmitted from the acquisition board, convert it into waveform data after standardization, and then convert the waveform data into feature vectors. The vibration feature vector analysis module inputs the preprocessed feature vectors into the neural network model. The input feature vectors are matched with the training samples stored in the model. The model achieves pattern matching by calculating the Euclidean distance or similarity between the input feature vectors and each training sample. Based on the pattern matching results, the model calculates the probability that the input feature vectors belong to each category. The vibration state judgment module sets one or more thresholds for judging abnormalities during the model training phase or according to actual application needs. Then, it compares the abnormal probability calculated by the vibration feature vector analysis module with the set thresholds. If the abnormal probability exceeds the set threshold, the current vibration state is judged to be abnormal. If the abnormal probability does not exceed the threshold, the current vibration state is judged to be normal.
4. The intelligent monitoring and alarm system for abnormal vibration of I / O device connectors according to claim 3, characterized in that, Neural network models are product-based neural network models.
5. A smart monitoring and alarm method for abnormal vibration of I / O device connectors, characterized in that, The methods include: Step S1: Monitor and capture the vibration signal of the monitored connector through the hardware signal acquisition circuit and convert the vibration signal into the input data required for abnormal vibration identification; Step S2: After converting the voltage data output by the hardware signal acquisition circuit into waveform data through the abnormal vibration identification system, the abnormal vibration identification system decomposes it into corresponding feature vectors and inputs them into the neural network model. The neural network model outputs the vibration state of the monitored connector, which includes abnormal vibration information. Step S3: Display abnormal vibration information and process alarm information through the abnormal alarm platform.
6. The intelligent monitoring and alarm method for abnormal vibration of I / O device connectors according to claim 5, characterized in that, Step S1 further includes: Step S1-1: Attach the vibration sensor to the monitored connector to monitor and record the vibration state of the monitored connector. When the monitored connector vibrates, the vibration sensor will convert the captured physical vibration into a voltage signal. Step S1-2: The voltage signal output by the vibration sensor is conditioned by the signal conditioning module. The signal conditioning includes amplifying the weak voltage signal and filtering out noise in the voltage signal. Step S1-3: The analog voltage signal output by the signal conditioning module is converted into voltage data by the acquisition board so that it can be processed by the subsequent abnormal vibration identification step.
7. The intelligent monitoring and alarm method for abnormal vibration of I / O device connectors according to claim 5, characterized in that, Step S2 further includes: Step S2-1: Receive voltage data transmitted from the acquisition board, convert it into waveform data after standardization, and convert the waveform data into feature vectors; Step S2-2: Input the preprocessed feature vector into the neural network model. The input feature vector is matched with the training samples stored in the model. The model achieves pattern matching by calculating the Euclidean distance or similarity between the input feature vector and each training sample. Based on the pattern matching results, the model calculates the probability that the input feature vector belongs to each category. Step S2-3: During the model training phase or according to actual application needs, set one or more thresholds for judging abnormalities. Then compare the abnormal probability calculated in the vibration feature vector analysis module with the set thresholds. If the abnormal probability exceeds the set threshold, the current vibration state is judged to be abnormal. If the abnormal probability does not exceed the threshold, the current vibration state is judged to be normal.
8. The intelligent monitoring and alarm method for abnormal vibration of I / O device connectors according to claim 7, characterized in that, Neural network models are product-based neural network models.