A method and device for detecting electrical components of a shield machine

Abstract

The present application discloses a method and device for detecting electrical components of a shield machine, and the present application belongs to the field of detection technology. The method includes: acquiring temperature data, vibration data, and current data of the electrical components of the shield machine; calculating the dynamic mean value, dynamic covariance matrix, and dynamic state deviation degree of the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data acquisition time point; and generating a warning message if there is a dynamic state deviation degree exceeding the preset dynamic state deviation degree threshold. This solution regularly acquires temperature, vibration, and current data, can timely capture the state changes of electrical components, and thus quickly respond to potential faults. Analyzing the data using the dynamic mean value and covariance matrix makes the fault detection more accurate and reduces the probability of false alarms and missed alarms. Once an anomaly is detected, the target fault data can be quickly determined, helping maintenance personnel quickly locate the problem and reducing the fault troubleshooting time.

Description

Technical Field

[0001] This application relates to the field of detection technologies, and particularly to a method and device for detecting electrical components of a shield machine. Background Art

[0002] As an important device for tunnel construction, the electrical devices of a shield machine are responsible for controlling, monitoring, and driving various working components. The stability of electrical components directly affects the overall performance and safety of the shield machine. Due to the complex working environment of the shield machine, which is easily affected by factors such as humidity, temperature, and vibration, electrical component failures may lead to equipment shutdown, construction delays, and even safety accidents. Therefore, it is crucial to regularly detect electrical components in order to timely identify potential faults, optimize maintenance strategies, and ensure the efficiency and safety of the construction process.

[0003] Currently, technicians manually inspect the shield machine and its electrical components at regular intervals according to a predetermined schedule, recording the operating status of the equipment and any visible abnormalities. During the inspection, technicians manually measure the performance parameters of each electrical component using portable instruments (such as thermometers, vibration meters, and ammeters). These measurement data are usually recorded on paper or input into a computer device. After the inspection, technicians analyze the collected data to determine whether the equipment is normal. This usually involves comparing with historical data to detect trends or abnormalities.

[0004] However, relying on manual monitoring and regular inspections cannot timely identify changes in the equipment status, resulting in handling after a failure occurs, which will increase the downtime and maintenance costs. Summary of the Invention

[0005] An embodiment of this application provides a method for detecting electrical components of a shield machine, which solves the problem in the prior art that relying on manual monitoring and regular inspections of electrical components of a shield machine cannot timely identify changes in the equipment status, resulting in handling after a failure occurs, which will increase the downtime and maintenance costs.

[0006] In a first aspect, an embodiment of this application provides a method for detecting electrical components of a shield machine, the method including:

[0007] Obtain temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data acquisition time point, and transmit the temperature data, vibration data, and current data to an edge computing device;

[0008] Calculate the dynamic mean values of the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data acquisition time point through the edge computing device, and calculate the dynamic covariance matrix of the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data acquisition time point;

[0009] According to the dynamic mean, the dynamic covariance matrix, and a preset dynamic state deviation degree calculation formula, calculate the dynamic state deviation degrees of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point;

[0010] If there is a dynamic state deviation degree exceeding the preset dynamic state deviation degree threshold, determine the target fault data, generate a warning message according to the target fault data, and transmit the warning message to the control center.

[0011] Further, the preset dynamic state deviation degree calculation formula is:

[0012] ;

[0013] Wherein, is the dynamic state deviation degree; is the temperature data, vibration data, and current data collected at the preset data acquisition time point t; is the transposed data of; is the dynamic mean of the temperature data, vibration data, and current data of the shield machine's electrical components at the preset data acquisition time point t.

[0014] Further, calculating the dynamic covariance matrix of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point includes:

[0015] Through the edge computing device, according to the preset dynamic mean calculation formula and the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point, calculate the dynamic mean of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point; wherein, the preset dynamic mean calculation formula is:

[0016] ;

[0017] Wherein, is the preset smoothing coefficient, and its value range is usually 0 < < 1; is the dynamic mean of the temperature data, vibration data, and current data at the preset data acquisition time point t - 1.

[0018] Further, calculating the dynamic covariance matrix of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point includes:

[0019] The edge computing device calculates the dynamic covariance matrix of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point according to a preset dynamic covariance matrix calculation formula; wherein, the preset dynamic covariance matrix calculation formula is:

[0020] ;

[0021] Wherein, is the dynamic covariance matrix at the preset data acquisition time point t; is the data vector of the temperature data, vibration data, and current data at the preset data acquisition time point t; is the dynamic covariance matrix at the preset data acquisition time point , that is, the dynamic covariance matrix at the previous preset data acquisition time point of the preset data acquisition time point t; is the difference between the data vector of the temperature data, vibration data, and current data at the preset data acquisition time point t and the dynamic mean; is the transpose of the difference between the data vector of the temperature data, vibration data, and current data at the preset data acquisition time point t and the dynamic mean.

[0022] Further, after transmitting the warning information to the control center, the method further includes:

[0023] Obtain historical fault records, and determine historical fault data, historical fault types, and historical influence ranges according to the historical fault records;

[0024] Label the fault type label and influence range label of the historical fault data according to the historical fault type and historical influence range;

[0025] Construct a fault detection model, and train the fault detection model according to the historical fault data, fault type labels, and influence range labels until the preset fault detection model training standard is reached.

[0026] Further, after training the fault detection model according to the historical fault data, fault type labels, and influence range labels until the preset model training standard is reached, the method further includes:

[0027] Input the target fault data into the fault detection model to determine the target fault type and target influence range;

[0028] Correspondingly, generating warning information according to the target fault data and transmitting the warning information to the control center includes:

[0029] Generate a warning message based on the target fault data, target fault type, and target impact range, and transmit the warning message to the control center.

[0030] Further, after inputting the target fault data into the fault detection model to determine the target fault type and target impact range, the method further includes:

[0031] If at least two maintenance plans sent by the control center are received, transmit the at least two maintenance plans to a preset maintenance plan evaluation model to obtain the maintenance time, maintenance cost, and maintenance risk of each maintenance plan;

[0032] Perform normalization processing on the maintenance time, maintenance cost, and maintenance risk of each maintenance plan to obtain the normalized maintenance time, normalized maintenance cost, and normalized maintenance risk of each maintenance plan;

[0033] Calculate the maintenance score of each maintenance plan according to the normalized maintenance time, normalized maintenance cost, and normalized maintenance risk of each maintenance plan and a preset maintenance score calculation formula, determine the maintenance plan with the highest maintenance score as the target maintenance plan, and transmit the target maintenance plan to the control center.

[0034] Further, the preset maintenance score calculation formula is:

[0035] ;

[0036] Wherein, is the maintenance score of each maintenance plan; is the preset weight of the normalized maintenance time; is the normalized maintenance time; is the preset weight of the normalized maintenance cost; is the normalized maintenance cost; is the preset weight of the normalized maintenance risk; is the normalized maintenance risk.

[0037] Further, the training process of the preset maintenance plan evaluation model includes:

[0038] Obtain historical maintenance plans, historical maintenance time, historical maintenance cost, and historical maintenance risk of the historical maintenance plans;

[0039] Label the time tag, cost tag, and risk tag of the historical maintenance plan according to the maintenance time, maintenance cost, and maintenance risk;

[0040] Construct a maintenance plan evaluation model, and train the maintenance plan evaluation model according to the historical maintenance plan, time tags, cost tags, and risk tags until the preset training standard of the maintenance plan evaluation model is reached.

[0041] According to the second aspect of the present application, a shield machine electrical component detection device is provided. The device includes:

[0042] A data acquisition module, configured to obtain temperature data, vibration data, and current data of the shield machine electrical components at each preset data acquisition time point, and transmit the temperature data, vibration data, and current data to the edge computing device;

[0043] A calculation module, configured to calculate the dynamic mean values of the temperature data, vibration data, and current data of the shield machine electrical components at each preset data acquisition time point through the edge computing device, and calculate the dynamic covariance matrix of the temperature data, vibration data, and current data of the shield machine electrical components at each preset data acquisition time point;

[0044] A dynamic state deviation degree determination module, configured to calculate the dynamic state deviation degrees of the temperature data, vibration data, and current data of the shield machine electrical components at each preset data acquisition time point according to the dynamic mean values, dynamic covariance matrix, and a preset dynamic state deviation degree calculation formula;

[0045] An early warning information generation module, configured to determine target fault data if there is a dynamic state deviation degree exceeding a preset dynamic state deviation degree threshold, generate early warning information according to the target fault data, and transmit the early warning information to the control center.

[0046] In the embodiments of the present application, regularly obtaining temperature, vibration, and current data can timely capture the state changes of electrical components, so as to quickly respond to potential faults. Analyzing the data using dynamic mean values and covariance matrices makes the fault detection more accurate and reduces the probability of false alarms and missed alarms. Once an anomaly is detected, the target fault data can be quickly determined, helping maintenance personnel quickly locate the problem and reducing the fault troubleshooting time. Description of the Drawings

[0047] Figure 1 is a schematic flowchart of the shield machine electrical component detection method provided in the first embodiment of the present application;

[0048] Figure 2 is a schematic flowchart of the shield machine electrical component detection method provided in the second embodiment of the present application;

[0049] Figure 3 is a schematic flowchart of the shield machine electrical component detection method provided in the third embodiment of the present application;

[0050] Figure 4 It is a schematic structural diagram of the shield machine electrical component detection device provided in the fourth embodiment of the present application. Specific implementation manners

[0051] In order to make the objectives, technical solutions, and advantages of the present application clearer, the following further describes the specific embodiments of the present application in detail with reference to the accompanying drawings. It can be understood that the specific embodiments described herein are only used to explain the present application, rather than limiting the present application. Additionally, it should be noted that for the sake of convenience of description, only the parts related to the present application are shown in the accompanying drawings rather than all the content. Before discussing the exemplary embodiments in more detail, it should be mentioned that some exemplary embodiments are described as processes or methods depicted as flowcharts. Although the flowcharts describe the operations (or steps) as sequential processes, many of the operations can be implemented in parallel, concurrently, or simultaneously. In addition, the order of the operations can be rearranged. When the operations are completed, the process can be terminated, but there can also be additional steps not included in the drawings. The process can correspond to a method, function, procedure, subroutine, subprogram, and so on.

[0052] The following will clearly describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments in the present application belong to the scope of protection of the present application.

[0053] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, rather than to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of the present application can be implemented in an order other than those illustrated or described herein, and the objects distinguished by "first", "second", etc. generally belong to the same category, and the number of objects is not limited. For example, the first object can be one or multiple. In addition, "and / or" in the specification and claims means at least one of the connected objects, and the character " / " generally indicates an "or" relationship between the associated objects before and after.

[0054] The following will, with reference to the accompanying drawings, explain in detail the shield machine electrical component detection method provided in the embodiments of the present application through specific embodiments and their application scenarios.

[0055] Embodiment 1

[0056] Figure 1 It is a schematic flowchart of the shield machine electrical component detection method provided in the first embodiment of the present application. As Figure 1 shown, it specifically includes the following steps:

[0057] S101. At each preset data acquisition time point, obtain the temperature data, vibration data, and current data of the electrical components of the shield machine, and transmit the temperature data, vibration data, and current data to the edge computing device.

[0058] First, the usage scenario of this solution can be a scenario for real-time monitoring of the electrical components of the shield machine to determine whether there are faults.

[0059] Based on the above usage scenario, it can be understood that the execution subject of this application can be a detection device for the electrical components of the shield machine, and there is no excessive limitation here.

[0060] The preset data acquisition time point can be a specific time interval or a specific moment predefined during the operation of the shield machine, such as the moment of data acquisition every few minutes, every hour, or at specific operation stages of the equipment (such as startup, operation, shutdown).

[0061] The electrical components of the shield machine can include motors, controllers, sensors, frequency converters, and other electrical components, which are responsible for controlling and monitoring the operation status of the shield machine.

[0062] The temperature data can be the temperature measured on the surface of the electrical component to monitor its working heat.

[0063] The vibration data can be the vibration amplitude and frequency detected by the component to evaluate the operation status.

[0064] The current data can be the working current measured by the electrical component to judge its load condition.

[0065] The edge computing device can be a computing device deployed close to the data source (such as the shield machine), responsible for real-time processing and analysis of the collected data, and usually includes sensors, microcontrollers, or embedded devices.

[0066] Appropriate sensors can be installed on the electrical components, including temperature sensors (such as thermocouples or thermistors), vibration sensors (such as accelerometers), and current sensors (such as Hall effect sensors). Configure the sensors to automatically collect data at the preset data acquisition time point. A timer or trigger can be used for scheduling to ensure data reading within a specific time interval (such as every minute, every hour). Convert the collected raw data into a standard format (such as JSON or CSV) for subsequent transmission and processing. Send the formatted data to the edge computing device through a wireless network (such as Wi-Fi, LoRa, or cellular network) or a wired connection (such as Ethernet). During the transmission process, data buffering and error checking mechanisms can be used to ensure data integrity and accuracy. The edge computing device is configured to listen on a specific network port or use a message queue to receive the transmitted data. Once the data is received, the edge computing device will perform preliminary processing and storage of the data for subsequent analysis.

[0067] S102, calculate the dynamic means of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point through the edge computing device, and calculate the dynamic covariance matrix of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point.

[0068] The dynamic mean can refer to, within a certain time window, reflecting the trend of data changing over time by performing weighted averaging on the real-time collected data. For example, the exponentially weighted moving average (EWMA) can be used to dynamically update the mean to better adapt to the changes in the data.

[0069] The dynamic covariance matrix can be a statistic that describes the strength of the relationship and the degree of change between multiple variables. It dynamically reflects the correlation of these data changing over time by calculating the covariance between each data dimension (such as temperature, vibration, current). The dynamic covariance matrix can also be updated using a method similar to that of the dynamic mean to maintain accuracy when the data changes.

[0070] At each preset data acquisition time point, the edge computing device can collect the temperature, vibration, and current data from the sensors. These data will be stored in chronological order for subsequent calculations. Then, the exponentially weighted moving average method is used to calculate the dynamic mean. The dynamic covariance matrix is calculated using the update formula of the dynamic covariance matrix.

[0071] Based on the above technical solution, optionally, calculating the dynamic means of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point through the edge computing device includes:

[0072] Calculating the dynamic means of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point through the edge computing device according to the preset dynamic mean calculation formula and the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point; wherein, the preset dynamic mean calculation formula is:

[0073] ;

[0074] Wherein, is the preset smoothing coefficient, and its value range is usually 0 < < 1. It controls the influence degree of the current data on the mean update. A larger will make the model more sensitive to the latest data; is the dynamic mean of the temperature data, vibration data, and current data at the preset data acquisition time point t - 1.

[0075] In this solution, during the first execution, if there is no historical mean value , it is necessary to initialize the dynamic mean. The initial mean can be set to the first data point , or a certain reasonable initial value. At the preset data acquisition time point t, temperature data, vibration data, and current data of the shield machine's electrical components are obtained from the sensor and recorded as . Then, the current data and the dynamic mean value at the previous time point are substituted into the formula for calculation . After the calculation , this value is stored as the input for the next time point (i.e., will become ).

[0076] Based on the above technical solution, optionally, calculate the dynamic covariance matrix of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point, including:

[0077] The edge computing device calculates the dynamic covariance matrix of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point according to the preset dynamic covariance matrix calculation formula and the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point; where the preset dynamic covariance matrix calculation formula is:

[0078] ;

[0079] Among them, is the dynamic covariance matrix at the preset data acquisition time point t; is the data vector of the temperature data, vibration data, and current data at the preset data acquisition time point t; is the dynamic covariance matrix at the preset data acquisition time point , that is, the dynamic covariance matrix of the previous preset data acquisition time point of the preset data acquisition time point t; is the difference between the data vector of the temperature data, vibration data, and current data at the preset data acquisition time point t and the dynamic mean; is the transpose of the difference between the data vector of the temperature data, vibration data, and current data at the preset data acquisition time point t and the dynamic mean.

[0080] In this solution, during the first execution, if there is no historical covariance matrix , it is necessary to initialize the covariance matrix. The initial value can be set as a zero matrix or a matrix calculated from the first set of data. Then, at the preset data acquisition time point t, temperature data, vibration data, and current data of the shield machine's electrical components are obtained from the sensors to form a vector at the current time point . According to the dynamic mean vector at the current time point t calculated in the previous steps . Then calculate the difference between the current data vector and the dynamic mean vector . This step generates a difference vector, representing the deviation of the current temperature, vibration, and current data from their means. Then calculate the outer product of the difference vector (i.e., the transpose of the difference vector multiplied by itself) . This step generates a 3×3 matrix, representing the deviation covariance between temperature, vibration, and current. Use the current data and the covariance matrix at the previous time point and substitute them into the formula to calculate the dynamic covariance matrix. Finally, after the calculation is completed , store this value as the covariance matrix input for the next time point . At each preset data acquisition time point t, repeat the steps to continuously update the covariance matrix.

[0081] S103. According to the dynamic mean, dynamic covariance matrix, and the preset dynamic state deviation degree calculation formula, calculate the dynamic state deviation degrees of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point.

[0082] The preset dynamic state deviation degree calculation formula can be used to measure the deviation degree of the currently collected temperature, vibration, and current data from their dynamic means.

[0083] The dynamic state deviation degree can refer to the deviation degree of the current state data (such as temperature, vibration, and current) of the current electrical component from the dynamic mean at a specific time point. Through this metric, it can be determined whether the device is in a normal working state.

[0084] The current state data (temperature data, vibration data, and current data) at the time point can be obtained, and the current state data, dynamic mean, and covariance matrix are substituted into the dynamic state deviation degree calculation formula to obtain the result.

[0085] On the basis of the above technical solution, optionally, the preset dynamic state deviation degree calculation formula is:

[0086] ;

[0087] Wherein, is the dynamic state deviation degree; are the temperature data, vibration data, and current data collected at the preset data acquisition time point t; is the transposed data of the dynamic mean value of the temperature data, vibration data, and current data at the preset data acquisition time point t.

[0088] In this solution, at each preset data acquisition time point, the temperature data , vibration data and current data can be obtained. According to the data collected in the previous step, the dynamic mean value is calculated using a preset dynamic mean value update formula . Then, using the current data and the dynamic mean value, the dynamic covariance matrix is calculated, and then its inverse matrix is calculated. Then, matrix multiplication is performed. First, is calculated, and then the result is multiplied by to obtain the dynamic state deviation degree .

[0089] S104. If there is a dynamic state deviation degree exceeding the preset dynamic state deviation degree threshold, determine the target fault data, generate a warning message according to the target fault data, and transmit the warning message to the control center.

[0090] The preset dynamic state deviation degree threshold can be a value determined through statistical analysis and is used to judge whether the current state data is abnormal. It is usually set based on the distribution characteristics of historical data. Exceeding this value means that there may be a fault in the operating state of the device.

[0091] The target fault data can refer to the specific data determined to be abnormal in the calculation of the dynamic state deviation degree. Specifically, it can include temperature data (such as too high temperature), vibration data (such as abnormal vibration frequency or amplitude), and current data (such as abnormal current fluctuation).

[0092] The warning message can be a notification generated based on the target fault data. Specifically, it can include the fault type (such as too high temperature, abnormal vibration, etc.), the time and specific location where the fault occurred.

[0093] The control center can be a device or person responsible for monitoring and managing the device state and can include a data monitoring device and a fault response device. By receiving and processing the warning message, the control center can take corresponding measures in a timely manner to ensure the safe operation of the device.

[0094] Once the deviation degree of a certain dynamic state exceeds the threshold, the device will lock the specific data type with anomalies. For example, if the deviation degree of temperature data exceeds the threshold, the temperature data is determined as the target fault data. If the deviation degrees of vibration and current data exceed the threshold simultaneously, both of these two data may become the target fault data. Based on the determined target fault data, the device can generate a warning message, the content of which may include the fault type (such as too high temperature, abnormal vibration, unstable current, etc.), the time and location of the fault occurrence, and the relevant fault data (specific temperature, vibration or current value). Then, through Internet of Things communication or edge computing devices, the warning message will be automatically transmitted to the control center, and the operator or device in the control center will receive the warning and then perform corresponding maintenance or adjustment operations.

[0095] In the embodiments of the present application, temperature data, vibration data, and current data of the electrical components of the shield machine are obtained at each preset data collection time point, and the temperature data, vibration data, and current data are transmitted to the edge computing device; the edge computing device calculates the dynamic mean values of the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data collection time point, and calculates the dynamic covariance matrix of the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data collection time point; according to the dynamic mean values, dynamic covariance matrix, and a preset dynamic state deviation degree calculation formula, the dynamic state deviation degrees of the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data collection time point are calculated; if there is a dynamic state deviation degree that exceeds the preset dynamic state deviation degree threshold, the target fault data is determined, a warning message is generated according to the target fault data, and the warning message is transmitted to the control center. Through the above shield machine electrical component detection method, temperature, vibration, and current data are regularly obtained, and the state changes of the electrical components can be captured in a timely manner, so as to quickly respond to potential faults. The use of dynamic mean values and covariance matrices to analyze the data makes the fault detection more accurate and reduces the probability of false alarms and missed alarms. Once an anomaly is detected, the target fault data can be quickly determined, helping maintenance personnel quickly locate the problem and reducing the fault troubleshooting time.

[0096] Embodiment 2

[0097] Figure 2 is a flowchart of the shield machine electrical component detection method provided by the second embodiment of the present application. As Figure 2 shown, the specific method includes the following steps:

[0098] S201, obtain temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data collection time point, and transmit the temperature data, vibration data, and current data to the edge computing device.

[0099] S202. Calculate the dynamic mean values of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point through the edge computing device, and calculate the dynamic covariance matrix of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point.

[0100] S203. Calculate the dynamic state deviation degrees of the temperature data, vibration data, and current data of the shield machine's electrical components at each preset data acquisition time point according to the dynamic mean values, the dynamic covariance matrix, and a preset dynamic state deviation degree calculation formula.

[0101] S204. If there is a dynamic state deviation degree exceeding the preset dynamic state deviation degree threshold, determine the target fault data, generate a warning message according to the target fault data, and transmit the warning message to the control center.

[0102] S205. Obtain historical fault records, and determine historical fault data, historical fault types, and historical influence scopes according to the historical fault records.

[0103] The historical fault record can be a device record of past fault events, including information such as the time, location, cause, duration, and impact degree of the fault occurrence.

[0104] The historical fault data can be specific data related to historical faults, including sensor readings (such as temperature, vibration, current, etc.), operating conditions at the time of fault occurrence, equipment status, etc.

[0105] The historical fault type can be the classification of faults, such as electrical faults, mechanical faults, software faults, etc., indicating the nature or category of the faults.

[0106] The historical influence scope can be the degree of impact on equipment, devices, or operations at the time of fault occurrence, which may include the impact on production, the degree of equipment damage, the difficulty of maintenance, etc.

[0107] The data sources for collecting historical fault records can include maintenance logs: recording repair and maintenance activities, including fault descriptions, repair times, and treatment results. Sensor data: real-time monitoring data (temperature, vibration, current, etc.) at the time of fault occurrence. Operating records: including the operating conditions and status of the equipment, which can help analyze the environment in which the fault occurs. Historical fault records can be obtained by means such as database queries and file retrievals, and relevant records are summarized, and then key data in the historical fault records are extracted, including sensor readings, equipment status, external conditions, etc. at the time of fault occurrence. According to the descriptions in the fault records or expert opinions, the faults are classified into specific types (such as electrical faults, mechanical faults, etc.). Analyze the impact of each fault event on equipment, devices, or operations to determine the influence scope (such as whether it affects a single component or the entire device shuts down).

[0108] S206. Label the fault type label and the impact range label of the historical fault data according to the historical fault type and the historical impact range.

[0109] The fault type label can be a label used to identify and classify historical fault data, which is convenient for classification and analysis in a machine learning model. For example, a certain fault data may be labeled as "electrical fault" or "overload fault".

[0110] The impact range label can be a label for the degree of impact of historical fault data, indicating the specific impact of the fault on equipment, devices, or production processes. For example, "slight impact", "moderate impact", or "severe impact".

[0111] Match each piece of historical fault data with its corresponding fault type and assign it the corresponding label (such as "electrical fault", "mechanical fault"). According to the analysis of the historical impact range, assign an impact range label (such as "severe", "medium", "slight") to each fault data.

[0112] S207. Build a fault detection model and train the fault detection model according to the historical fault data, the fault type label, and the impact range label until the preset fault detection model training standard is reached.

[0113] The fault detection model can be a machine learning or statistical model. By using historical fault data and related labels and learning the patterns in the historical data, it can identify and predict faults in equipment or devices. Common models include decision trees, support vector machines, neural networks, etc.

[0114] The preset fault detection model training standard can be a standard for evaluating the performance of the model, which may include indicators such as model accuracy, recall rate, F1 score, etc. The training process will end when the model reaches these standards to ensure that the model has reliable fault detection capabilities.

[0115] The labeled historical fault data, fault type labels, and impact range labels can be organized into a training set and a test set. Select a suitable machine learning algorithm (such as decision tree, random forest, support vector machine, etc.) or a deep learning model (such as a neural network) for fault detection. Use the training set to train the model and adjust the hyperparameters to improve the accuracy of the model. Use the test set to evaluate the performance of the model and calculate indicators such as accuracy, recall rate, and F1 score. If the performance of the model reaches the preset standard (for example, the accuracy exceeds 95%), stop training; otherwise, continue to optimize the model and adjust the data and algorithms.

[0116] In this embodiment, by using historical data and machine learning algorithms, the model can more accurately identify the types of faults, thereby reducing the situations of missed detection and misdetection. Real-time monitoring and analysis can issue early warnings at the initial stage of fault occurrence, helping the maintenance team respond promptly and reducing equipment downtime and maintenance costs.

[0117] Based on the above technical solution, optionally, after training the fault detection model according to the historical fault data, fault type labels, and impact scope labels until the preset model training standard is reached, the method further includes:

[0118] Input the target fault data into the fault detection model to determine the target fault type and the target impact scope;

[0119] Correspondingly, generating a warning message according to the target fault data and transmitting the warning message to the control center includes:

[0120] Generating a warning message according to the target fault data, the target fault type, and the target impact scope, and transmitting the warning message to the control center.

[0121] In this solution, the target fault type may refer to the specific fault type identified during the fault detection process, such as overheating of electrical components, abnormal vibration, or unstable current, etc.

[0122] The target impact scope may refer to the device part or equipment that the fault may affect, such as may affect the overall operation efficiency or cause downtime, etc.

[0123] The target fault data (such as temperature, vibration, current, etc.) collected from the monitoring device can be input into the fault detection model. The pre-trained fault detection model analyzes the input data. The model will use historical data and feature extraction algorithms to identify possible fault features in the data. The model outputs the identification result, clarifying the specific type of the fault. This may include feature matching similar to historical faults to determine the most likely fault type. According to the fault type, the model analyzes the potential impact scope of the fault on the device and determines the equipment or operations that the fault may affect. According to the identified target fault type and impact scope, a warning message is automatically generated, including fault description, impact analysis, recommended measures, etc. The generated warning message is transmitted to the control center through the network so that relevant personnel can receive the information in time and take actions.

[0124] In this solution, the fault type and impact scope can be identified in real time, corresponding measures can be taken quickly, and potential losses can be reduced. Through early warning, equipment downtime and serious faults can be avoided, and maintenance and repair costs can be reduced.

[0125] Embodiment III

[0126] Figure 3 It is a schematic flowchart of the method for detecting electrical components of a shield machine provided in Embodiment 3 of the present application. As Figure 3 shown, the specific method includes the following steps:

[0127] S301, Obtain the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data acquisition time point, and transmit the temperature data, vibration data, and current data to the edge computing device.

[0128] S302, Calculate the dynamic mean values of the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data acquisition time point through the edge computing device, and calculate the dynamic covariance matrix of the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data acquisition time point.

[0129] S303, Calculate the dynamic state deviation degrees of the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data acquisition time point according to the dynamic mean values, dynamic covariance matrix, and the preset dynamic state deviation degree calculation formula.

[0130] S304, If there is a dynamic state deviation degree exceeding the preset dynamic state deviation degree threshold, determine the target fault data, generate a warning message according to the target fault data, and transmit the warning message to the control center.

[0131] S305, Obtain the historical fault records, and determine the historical fault data, historical fault types, and historical influence ranges according to the historical fault records.

[0132] S306, Mark the fault type labels and influence range labels of the historical fault data according to the historical fault types and historical influence ranges.

[0133] S307, Construct a fault detection model, and train the fault detection model according to the historical fault data, fault type labels, and influence range labels until the preset fault detection model training standard is reached.

[0134] Input the target fault data into the fault detection model to determine the target fault type and target influence range.

[0135] S308, If at least two maintenance plans sent by the control center are received, transmit the at least two maintenance plans to the preset maintenance plan evaluation model to obtain the maintenance time, maintenance cost, and maintenance risk of each maintenance plan.

[0136] A maintenance plan can be a detailed scheme for maintaining equipment, devices or components, which may include maintenance objectives: clarifying the problems to be solved or the maintenance effects to be achieved, such as improving the operating efficiency of equipment or extending its service life. Maintenance tasks: specific operation steps, such as replacing parts, lubricating equipment, checking electrical devices, etc. Resource allocation: a list of required personnel, tools and materials, including the skill requirements of technicians. Time arrangement: the start and end times of maintenance, as well as the time nodes of each task, to ensure that normal production is not affected.

[0137] A preset maintenance plan evaluation model can be a tool for evaluating and comparing different maintenance schemes, which may include historical data analysis: using past maintenance records and failure data to analyze the effects and success rates of various maintenance plans. Evaluation criteria: setting a series of criteria to measure the effectiveness of maintenance plans, such as maintenance time, cost and risk, etc. Decision-making algorithm: using mathematical models or machine learning algorithms to output evaluation results based on the input maintenance plan information to help decision-makers select the optimal plan.

[0138] Maintenance time can refer to the time required to complete a specific maintenance task, which may include preparation time: the time required for maintenance preparation, including tool preparation and personnel deployment. Execution time: the time for actually carrying out maintenance operations. Inspection time: the time for detecting and confirming the equipment after maintenance is completed.

[0139] Maintenance cost can refer to the financial expenditure required to execute a maintenance plan, which may include direct costs: including material costs, outsourcing service costs and labor wages, etc. Indirect costs: such as production losses and time waste caused by equipment downtime. Forecast cost: the expected expenditure for future maintenance activities to facilitate better budget management.

[0140] Maintenance risk can refer to the potential problems associated with maintenance activities and their possible consequences, which may include equipment risk: the risk of causing equipment failure or damage during maintenance. Safety risk: factors that may pose a threat to the safety of operators during maintenance activities. Economic risk: additional costs or economic losses that may be caused by improper maintenance.

[0141] Details of at least two maintenance plans can be received from the control center, including task content, schedule, required resources, etc. for each plan. Structure the received maintenance plans, extract key data such as maintenance objectives, tasks, required materials, estimated time, and cost. Transmit the prepared maintenance plan data to a preset maintenance plan evaluation model. This usually involves using an API or data transfer protocol to ensure information security and integrity. In the preset maintenance plan evaluation model, apply historical data and evaluation criteria to evaluate each maintenance plan. Among them, maintenance time: calculate the estimated execution time for each plan. Maintenance cost: evaluate the required direct and indirect costs. Maintenance risk: analyze the risks that each plan may face, and judge its potential problems in combination with historical maintenance records. The model outputs the maintenance time, maintenance cost, and maintenance risk of each maintenance plan. This information is usually presented in the form of a table or report for easy analysis and decision-making.

[0142] S309, normalize the maintenance time, maintenance cost, and maintenance risk of each maintenance plan to obtain the normalized maintenance time, normalized maintenance cost, and normalized maintenance risk of each maintenance plan.

[0143] The normalized maintenance time can be adjusted to a value between 0 and 1 for the maintenance time of each maintenance plan.

[0144] The normalized maintenance cost can be adjusted to a value between 0 and 1 for the maintenance cost of each maintenance plan.

[0145] The normalized maintenance risk can be adjusted to a value between 0 and 1 for the maintenance cost of each maintenance risk.

[0146] The maintenance time, maintenance cost, and maintenance risk of each plan can be sorted to screen out the maximum maintenance time, minimum maintenance time, maximum maintenance cost, minimum maintenance cost, maximum maintenance risk, and minimum maintenance risk. Then use the following formula to calculate the normalized maintenance time, normalized maintenance cost, and normalized maintenance risk:

[0147] ;

[0148] S310, calculate the maintenance score of each maintenance plan according to the normalized maintenance time, normalized maintenance cost, and normalized maintenance risk of each maintenance plan and the preset maintenance score calculation formula, determine the maintenance plan with the highest maintenance score as the target maintenance plan, and transmit the target maintenance plan to the control center.

[0149] A preset maintenance score calculation formula can be a mathematical expression used to quantify and compare different maintenance plans. Its main purpose is to comprehensively evaluate the key indicators (such as maintenance time, maintenance cost, maintenance risk) of each maintenance plan to help decision-makers select the optimal maintenance plan.

[0150] The maintenance score can be a comprehensive indicator used to measure the overall quality of a maintenance plan. The higher the score, the better the performance of the maintenance plan in terms of time, cost, and risk.

[0151] The maintenance score can be a comprehensive indicator used to measure the overall quality of a maintenance plan. The higher the score, the better the performance of the maintenance plan in terms of time, cost, and risk.

[0152] The target maintenance plan can be the one with the highest maintenance score. After comparing all maintenance plans, select the optimal plan and execute it.

[0153] The maintenance score of each maintenance plan can be calculated. Compare the maintenance scores of all maintenance plans. Determine the plan with the highest maintenance score as the target maintenance plan. Transmit the detailed information of the target maintenance plan to the control center for further execution or arrangement.

[0154] In this embodiment, by quantifying the indicators of the maintenance plan, different options can be evaluated more scientifically, thus making a better decision. The device-based evaluation model can speed up the selection process of the maintenance plan, reduce the time of manual judgment, and improve the overall work efficiency.

[0155] Based on the above technical solution, optionally, the preset maintenance score calculation formula is:

[0156] ;

[0157] Where is the maintenance score of each maintenance plan; is the preset weight of the normalized maintenance time; is the normalized maintenance time; is the preset weight of the normalized maintenance cost; is the normalized maintenance cost; is the preset weight of the normalized maintenance risk; is the normalized maintenance risk.

[0158] In this solution, 、 and can be read from the database, and then the normalized maintenance time, normalized maintenance cost, normalized maintenance risk, 、 and Substitute into the formula to calculate the maintenance score for each maintenance plan.

[0159] Based on the above technical solution, optionally, the training process of the preset maintenance plan evaluation model includes:

[0160] Obtain historical maintenance plans, historical maintenance times of the historical maintenance plans, historical maintenance costs, and historical maintenance risks;

[0161] Label the time tag, cost tag, and risk tag of the historical maintenance plan according to the maintenance time, maintenance cost, and maintenance risk;

[0162] Construct a maintenance plan evaluation model, and train the maintenance plan evaluation model according to the historical maintenance plan, time tag, cost tag, and risk tag until the preset maintenance plan evaluation model training standard is reached.

[0163] In this solution, the historical maintenance plan can be a detailed record of past implemented maintenance activities, including the measures and operation steps taken.

[0164] The historical maintenance time can be the actual time required to execute each maintenance plan.

[0165] The historical maintenance cost can be the financial resources consumed by each maintenance plan, including labor, materials, and equipment costs.

[0166] The historical maintenance risk can be the probability and impact of potential problems or failures that each maintenance plan may face during implementation.

[0167] The time tag can be an identifier marked for the historical maintenance time of each maintenance plan, usually used for classification and analysis.

[0168] The cost tag can be an identifier of the historical cost information of the maintenance plan.

[0169] The risk tag can be an identifier of the risks associated with each maintenance plan, usually involving the type and probability of failures.

[0170] The preset maintenance plan evaluation model training standard can include specific performance indicators, such as accuracy, recall rate, or F1 score, to determine when to stop training the model and consider it to have reached a qualified level.

[0171] Historical maintenance plans, maintenance times, costs, and risk data can be obtained from the maintenance record database. For the collected historical data, analyze and label the corresponding time, cost, and risk tags one by one. Machine learning labeling methods or manual labeling can be used. Then select a suitable algorithm (such as decision tree, random forest, or neural network) to construct a maintenance plan evaluation model. Input the historical maintenance plan, time tags, cost tags, and risk tags into the model for training. Use the preset training criteria to evaluate the performance of the model. During the training process, adjust the parameters according to the performance of the model on the validation set until the preset training criteria are met. Use an independent test set to verify the accuracy and stability of the model to ensure its effectiveness in practical applications.

[0172] In this solution, by analyzing historical data through instrumentation, the effectiveness of the maintenance plan can be quickly evaluated, which can reduce the time of manual judgment. It can help identify the best maintenance strategy, reduce maintenance time and costs, and improve the efficiency of resource use.

[0173] Embodiment 4

[0174] Figure 4 is a schematic structural diagram of a shield machine electrical component detection device provided in Embodiment 4 of the present application, as Figure 4 shown. This device is used to implement the method of a shield machine electrical component detection device provided in Embodiments 1, 2, and 3. Specifically, this device includes the following:

[0175] A data acquisition module 401, which is used to obtain the temperature data, vibration data, and current data of the shield machine electrical components at each preset data acquisition time point, and transmit the temperature data, vibration data, and current data to the edge computing device;

[0176] A calculation module 402, which is used to calculate the dynamic mean values of the temperature data, vibration data, and current data of the shield machine electrical components at each preset data acquisition time point through the edge computing device, and calculate the dynamic covariance matrix of the temperature data, vibration data, and current data of the shield machine electrical components at each preset data acquisition time point;

[0177] A dynamic state deviation degree determination module 403, which calculates the dynamic state deviation degree of the temperature data, vibration data, and current data of the shield machine electrical components at each preset data acquisition time point according to the dynamic mean value, dynamic covariance matrix, and a preset dynamic state deviation degree calculation formula;

[0178] An early warning information generation module 404, which is used to determine target fault data if there is a dynamic state deviation degree exceeding the preset dynamic state deviation degree threshold, generate early warning information according to the target fault data, and transmit the early warning information to the control center.

[0179] In the embodiment of the present application, the data acquisition module is used to obtain the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data acquisition time point, and transmit the temperature data, vibration data, and current data to the edge computing device; the calculation module is used to calculate the dynamic mean values of the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data acquisition time point through the edge computing device, and calculate the dynamic covariance matrix of the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data acquisition time point; the dynamic state deviation degree determination module calculates the dynamic state deviation degree of the temperature data, vibration data, and current data of the electrical components of the shield machine at each preset data acquisition time point according to the dynamic mean value, dynamic covariance matrix, and a preset dynamic state deviation degree calculation formula; the warning information generation module is used to determine target fault data if there is a dynamic state deviation degree exceeding the preset dynamic state deviation degree threshold, generate warning information according to the target fault data, and transmit the warning information to the control center. Through the above shield machine electrical component detection device, the temperature, vibration, and current data are regularly obtained, and the state changes of the electrical components can be captured in time, so as to quickly respond to potential faults. The data is analyzed using the dynamic mean value and covariance matrix, making the fault detection more accurate and reducing the probability of false alarms and missed alarms. Once an anomaly is detected, the target fault data can be quickly determined, helping maintenance personnel quickly locate the problem and reducing the fault troubleshooting time.

[0180] The above is only the preferred embodiment of the present application and the applied technical principle. The present application is not limited to the specific embodiments described herein. Various obvious changes, re-adjustments, and substitutions that can be made by those skilled in the art will not depart from the protection scope of the present application. Therefore, although the present application has been described in detail through the above embodiments, the present application is not limited to the above embodiments. Without departing from the concept of the present application, more other equivalent embodiments can be included, and the scope of the present application is determined by the scope of the claims.

Claims

1. A method for detecting electrical components of a shield machine, characterized in that: The method comprises: Acquire temperature data, vibration data, and current data of electrical components of the shield machine at each preset data collection time point, and transmit the temperature data, vibration data, and current data to the edge computing device; The edge computing device calculates the dynamic mean of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point, and calculates the dynamic covariance matrix of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point; wherein the edge computing device calculates the dynamic mean of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point, including: The edge computing device calculates the dynamic mean of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point according to the preset dynamic mean calculation formula and the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point; wherein the preset dynamic mean calculation formula is: ； in, is the preset smoothing coefficient, and its value range is usually 0< <1; is the dynamic mean of the temperature data, vibration data and current data at the preset data collection time point t-1; Calculate the dynamic covariance matrix of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point, including: The edge computing device calculates the dynamic covariance matrix of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point according to the preset dynamic covariance matrix calculation formula and the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point; wherein the preset dynamic covariance matrix calculation formula is: ； in, The dynamic covariance matrix at the preset data collection time point t; is a data vector of temperature data, vibration data and current data at a preset data collection time point t; At the preset data collection time point The dynamic covariance matrix of , that is, the dynamic covariance matrix of the previous preset data collection time point of the preset data collection time point t; is the difference between the data vector and the dynamic mean of the temperature data, vibration data and current data at the preset data collection time point t; is the transposition of the difference between the data vector of the temperature data, vibration data and current data at the preset data collection time point t and the dynamic mean; According to the dynamic mean, the dynamic covariance matrix and the preset dynamic state deviation calculation formula, the dynamic state deviation of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point is calculated; wherein the preset dynamic state deviation calculation formula is: ； in, is the dynamic state deviation; The temperature data, vibration data and current data collected at the preset data collection time point t; for The transposed data of is the dynamic mean of temperature data, vibration data and current data at the preset data collection time point t; Dynamic Covariance Matrix ; If there is a dynamic state deviation that exceeds a preset dynamic state deviation threshold, target fault data is determined, warning information is generated according to the target fault data, and the warning information is transmitted to a control center.

2. The shield machine electrical component detection method according to claim 1, characterized in that: After transmitting the warning information to the control center, the method further includes: Obtain historical fault records, and determine historical fault data, historical fault types, and historical impact ranges based on the historical fault records; Marking the historical fault data with a fault type label and an impact range label according to the historical fault type and the historical impact range; A fault detection model is constructed, and the fault detection model is trained according to the historical fault data, fault type labels, and impact range labels until a preset fault detection model training standard is reached.

3. The shield machine electrical component detection method according to claim 2, characterized in that: After training the fault detection model according to the historical fault data, the fault type label and the impact range label until a preset model training standard is reached, the method further includes: Inputting the target fault data into a fault detection model to determine the target fault type and target impact range; Accordingly, generating warning information according to the target fault data and transmitting the warning information to a control center includes: Generate warning information according to the target fault data, target fault type and target impact range, and transmit the warning information to a control center.

4. The shield machine electrical component detection method according to claim 3, characterized in that: After inputting the target fault data into the fault detection model to determine the target fault type and the target impact range, the method further includes: If at least two maintenance plans are received from the control center, the at least two maintenance plans are transmitted to a preset maintenance plan evaluation model to obtain maintenance time, maintenance cost, and maintenance risk of each maintenance plan; The maintenance time, maintenance cost and maintenance risk of each maintenance plan are normalized to obtain the normalized maintenance time, normalized maintenance cost and normalized maintenance risk of each maintenance plan; The maintenance score of each maintenance plan is calculated according to the normalized maintenance time, normalized maintenance cost and normalized maintenance risk of each maintenance plan and a preset maintenance score calculation formula, and the maintenance plan with the highest maintenance score is determined as the target maintenance plan, and the target maintenance plan is transmitted to the control center.

5. The shield machine electrical component detection method according to claim 4, characterized in that: The preset maintenance score calculation formula is: ； in, Give each maintenance plan a maintenance score; The preset weight for normalized maintenance time; To normalize the maintenance time; The preset weights for normalizing maintenance costs; is the normalized maintenance cost; Preset weights for maintaining risks for normalization; Maintain risk for normalization.

6. The shield machine electrical component detection method according to claim 4, characterized in that: The training process of the preset maintenance plan evaluation model includes: Obtain historical maintenance plans, historical maintenance time, historical maintenance costs, and historical maintenance risks of historical maintenance plans; Marking the historical maintenance plan with a time label, a cost label, and a risk label according to the maintenance time, maintenance cost, and maintenance risk; A maintenance plan judgment model is constructed, and the maintenance plan judgment model is trained according to the historical maintenance plan, time label, cost label, and risk label until a preset maintenance plan judgment model training standard is reached.

7. A shield machine electrical component detection device, characterized in that: The device comprises: A data acquisition module, used to obtain temperature data, vibration data and current data of electrical components of the shield machine at each preset data acquisition time point, and transmit the temperature data, vibration data and current data to the edge computing device; A calculation module is used to calculate the dynamic mean of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point through the edge computing device, and calculate the dynamic covariance matrix of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point; wherein the dynamic mean of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point is calculated by the edge computing device, including: The edge computing device calculates the dynamic mean of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point according to the preset dynamic mean calculation formula and the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point; wherein the preset dynamic mean calculation formula is: ； in, is the preset smoothing coefficient, and its value range is usually 0< <1; is the dynamic mean of the temperature data, vibration data and current data at the preset data collection time point t-1; Calculate the dynamic covariance matrix of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point, including: The edge computing device calculates the dynamic covariance matrix of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point according to the preset dynamic covariance matrix calculation formula and the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point; wherein the preset dynamic covariance matrix calculation formula is: ； in, The dynamic covariance matrix at the preset data collection time point t; is a data vector of temperature data, vibration data and current data at a preset data collection time point t; At the preset data collection time point The dynamic covariance matrix of , that is, the dynamic covariance matrix of the previous preset data collection time point of the preset data collection time point t; is the difference between the data vector and the dynamic mean of the temperature data, vibration data and current data at the preset data collection time point t; is the transposition of the difference between the data vector of the temperature data, vibration data and current data at the preset data collection time point t and the dynamic mean; The dynamic state deviation determination module calculates the dynamic state deviation of the temperature data, vibration data and current data of the electrical components of the shield machine at each preset data collection time point according to the dynamic mean, the dynamic covariance matrix and the preset dynamic state deviation calculation formula; wherein the preset dynamic state deviation calculation formula is: ； in, is the dynamic state deviation; The temperature data, vibration data and current data collected at the preset data collection time point t; for The transposed data of is the dynamic mean of temperature data, vibration data and current data at the preset data collection time point t; Dynamic Covariance Matrix ; The warning information generation module is used to determine target fault data if there is a dynamic state deviation exceeding a preset dynamic state deviation threshold, generate warning information according to the target fault data, and transmit the warning information to the control center.

Images