Multi-source sensing fusion subway equipment state monitoring method and system

Abstract

The invention relates to the technical field of subway equipment state monitoring, and discloses a multi-source sensing fusion subway equipment state monitoring method and system, and the system comprises a data collection module, an information fusion module, an equipment state diagnosis module, an early warning execution module and a collaborative management module. The data acquisition module acquires multi-source equipment state original data through a multi-type sensor array; the information fusion module carries out calibration and feature fusion on multi-source equipment state original data to generate fusion equipment state information; the equipment state diagnosis module combines and fuses the equipment state information and the early warning execution data, and adopts a mode matching method to obtain diagnosis data; the early warning execution module controls an early warning device to complete early warning action according to the diagnosis data; and the collaborative management module adjusts parameters of each module to form closed-loop control. According to the system, through multi-source sensing data fusion, accurate diagnosis and collaborative management, comprehensive monitoring of the state of subway equipment is realized, the fault early warning accuracy and processing efficiency are improved, and safe operation of a subway is guaranteed.

Description

Technical Field

[0001] The present invention relates to the technical field of subway equipment status monitoring, and in particular to a subway equipment status monitoring method and system using multi-source sensor fusion. Background Art

[0002] With the rapid development of urban rail transit, subways, as a vital component of urban public transportation, are crucial for the safe and stable operation of their equipment. Subway equipment is diverse, including key components such as traction systems, braking systems, and bogies. These components are susceptible to various complex factors during long-term operation, such as vibration, temperature fluctuations, and current fluctuations, making them prone to failure or performance degradation. Equipment failure not only impacts normal subway operations but can also pose safety risks.

[0003] Traditional subway equipment condition monitoring methods typically use a single sensor for data collection. This approach suffers from limited monitoring range and single-source data, making it difficult to fully and accurately reflect the equipment's actual operating status. For example, monitoring equipment vibration solely through a vibration sensor fails to simultaneously capture critical parameters such as temperature and current, easily leading to missed or misdiagnosed conditions. Furthermore, traditional monitoring systems have relatively weak data processing capabilities, making it difficult to effectively integrate and analyze large amounts of monitoring data, making it difficult to promptly detect potential equipment failures.

[0004] Existing subway equipment condition monitoring systems also have shortcomings in their early warning mechanisms and collaborative management. Warning levels and trigger conditions are not set scientifically and rationally, making it difficult to accurately issue warnings based on the criticality and actual operating status of equipment. Furthermore, there is a lack of effective collaborative management between monitoring modules, preventing the formation of a closed-loop monitoring and control system. This results in a suboptimal monitoring process and low monitoring efficiency.

[0005] With the continuous development of multi-sensor fusion technology, its application in subway equipment condition monitoring has become a trend. Multi-source sensor fusion technology can improve the accuracy and reliability of monitoring by deploying various types of sensors to collect multi-source data during equipment operation and fusing this data. However, current systems that apply multi-source sensor fusion technology to subway equipment condition monitoring still have some problems, such as incomplete data calibration methods, suboptimal feature fusion algorithms, and inaccurate equipment condition diagnosis methods. These issues require further improvement and refinement. Summary of the Invention

[0006] The purpose of the present invention is to provide a subway equipment status monitoring method and system based on multi-source sensor fusion to solve the problems raised in the above background technology.

[0007] To achieve the above objectives, the present invention provides the following technical solution: a subway equipment status monitoring system with multi-source sensor fusion, the system comprising: Data acquisition module, information fusion module, equipment status diagnosis module, early warning execution module and collaborative management module; The data acquisition module is used to acquire a multi-source sensor data set related to the equipment status during the operation of subway equipment. Specifically, it implements data acquisition by deploying a multi-type sensor array to obtain multi-source equipment status raw data, and transmits the multi-source equipment status raw data to the information fusion module; The information fusion module is used to process the multi-source device status raw data and generate device status fusion information, specifically by sequentially performing data calibration and feature fusion operations on the multi-source device status raw data to obtain fused device status information, and sending the fused device status information to the device status diagnosis module and the early warning execution module; The device status diagnosis module is used to analyze the real-time device status during device operation. Specifically, it combines the fused device status information with the warning execution data output by the warning execution module, adopts a pattern matching-based device status diagnosis method to obtain device status diagnosis data, and transmits the device status diagnosis data to the collaborative management module. The warning execution module is used to implement warning operations based on the diagnostic data, specifically to control the operation of multiple types of warning execution devices based on the equipment status diagnostic data, complete the warning action of abnormal equipment status, and feed back the warning execution data to the equipment status diagnostic module; The collaborative management module is used to coordinate the operation of each module and optimize the monitoring process. Specifically, it adjusts the working parameters of each module according to the equipment status diagnostic data, realizes the collaborative management of the multi-source sensor monitoring process, and forms a closed-loop monitoring and control system.

[0008] Preferably, in the data acquisition module, the multi-source equipment status original data specifically includes real-time vibration data, equipment surface temperature data, motor operating current data, key component position data and historical equipment status record data; the historical equipment status record data specifically includes historical abnormal time record data, historical parameter adjustment record data and historical warning effect feedback data.

[0009] Preferably, in the information fusion module, the step of sequentially performing data calibration and feature fusion operations on the multi-source device status raw data to obtain fused device status information includes data calibration processing, feature extraction processing, fusion calculation processing, and information generation processing; The data calibration process is used to remove interference values ​​and correct errors in the original data of the multi-source device status, specifically by screening and correcting the original data of the multi-source device status by setting a data valid range to obtain calibrated data; The feature extraction process specifically extracts key features related to the device status from the calibrated data, including vibration frequency features, temperature change gradient features, and current fluctuation features, and performs correlation calculation on the key features to obtain correlated feature data; The fusion calculation process specifically comprises determining the comprehensive characterization value of vibration intensity, temperature distribution and current stability by using the evidence theory fusion method based on the associated characteristic data; The information generation process specifically converts the comprehensive characterization value into analyzable device status fusion information to form the fused device status information.

[0010] Preferably, in the early warning execution module, the operation of multiple types of early warning execution devices is controlled based on the equipment status diagnostic data to complete early warning actions for abnormal equipment status, specifically including equipment graded early warning, signal output control and execution action synchronization; The device classification warning is specifically to set the warning level and trigger conditions for the corresponding device according to the importance of each device and the device status diagnostic data; The signal output control specifically adjusts the output type of each warning device through the signal conversion interface to ensure that the warning requirements of different devices are met; The execution action synchronization is specifically to coordinate the start and stop time of each early warning execution device through a time synchronization controller to avoid mutual interference during the early warning process.

[0011] Preferably, in the device status diagnosis module, the step of combining the fused device status information with the warning execution data output by the warning execution module and adopting a pattern matching-based device status diagnosis method to obtain device status diagnosis data includes historical pattern matching, real-time data comparison, abnormal device status determination, and diagnosis result generation; The historical pattern matching is specifically to extract historical equipment status data that is the same or similar to the current equipment working condition from the historical equipment status record data as a reference template; The real-time data comparison specifically involves comparing the fusion device status information with the reference template item by item, and calculating the difference values ​​of vibration, temperature and current; The abnormal device status determination is specifically to determine whether the current device operation is abnormal based on the difference values ​​of vibration, temperature and current by setting a difference threshold range. If the difference value exceeds the threshold, it is marked as an abnormal device status; The diagnostic result is generated by integrating the comparison result and the abnormal device status information to form the device status diagnostic data.

[0012] Preferably, in the collaborative management module, the operating parameters of each module are adjusted according to the device status diagnostic data to achieve collaborative management of the multi-source sensor monitoring process, specifically including data reception and processing, management strategy generation, instruction issuance and execution, and feedback optimization and adjustment; The data receiving and processing is specifically receiving the device status diagnostic data and analyzing the abnormal information and comparison results therein; The management strategy generation is specifically to generate an adjustment strategy for abnormal equipment or parameters based on the abnormal information and the comparison results. The adjustment strategy includes a vibration parameter correction strategy, a temperature control compensation strategy, or a current limit adjustment strategy; The instructions are sent for execution, specifically converting the adjustment strategy into a management instruction and sending it to the corresponding module, wherein the information fusion module receives the parameter correction instruction, and the early warning execution module receives the action adjustment instruction; The feedback optimization and adjustment specifically involves collecting the adjusted fusion device status information and diagnostic data, verifying the adjustment effect and further optimizing the management strategy to form a continuously optimized collaborative management process.

[0013] Preferably, the multi-type sensor array deployed in the data acquisition module specifically includes a piezoelectric vibration sensor for collecting vibration data, an infrared temperature sensor for collecting temperature data, a Hall-type current sensor for collecting current data, and a laser position sensor for collecting position data; each sensor collects data according to a preset period, and the preset period includes collecting data once every 3 seconds, collecting data once every 6 seconds, and collecting data once every 15 seconds, and the original data of the multi-source device status is obtained through the collection methods of different periods.

[0014] Preferably, the data calibration processing performed in the information fusion module specifically includes noise suppression processing and missing value filling processing; the noise suppression processing specifically adopts a median filtering algorithm to smooth the high-frequency noise data; the missing value filling processing specifically uses a spline interpolation method to supplement the missing sensor data within the acquisition period, thereby ensuring the integrity and continuity of the original data of the multi-source device status.

[0015] Preferably, the device status diagnostic data obtained in the device status diagnostic module specifically includes abnormality probability data, device status category data, and diagnostic index data; the abnormality probability data is used to represent the probability value of abnormality occurring in the current device operation, and the value range is 0 to 100%; the device status category data is used to represent the identifiable device operation status types, specifically including stable operation state, normal operation state, sub-healthy operation state, and abnormal operation state; the diagnostic index data is used to provide a basis for device status discrimination, and specifically includes vibration difference recording data, temperature difference recording data, and current difference recording data; The management strategies generated in the collaborative management module specifically include parameter optimization strategies for sub-healthy operating states, emergency intervention strategies for abnormal operating states, and parameter fine-tuning strategies for normal operating states.

[0016] Preferably, the present invention further includes a multi-source sensor fusion subway equipment status monitoring method, which is applied to the above-mentioned multi-source sensor fusion subway equipment status monitoring system, and the method includes the following steps: Step 1: Deploy a multi-type sensor array to implement data collection, obtain multi-source device status raw data related to the device status during the operation of subway equipment, and transmit the multi-source device status raw data to the information fusion module; Step 2: The information fusion module sequentially performs data calibration and feature fusion operations on the multi-source device status raw data to obtain fused device status information, and sends the fused device status information to the device status diagnosis module and the early warning execution module; Step 3: The device status diagnosis module combines the fused device status information and the early warning execution data, adopts a pattern matching-based device status diagnosis method, obtains device status diagnosis data, and transmits the device status diagnosis data to the collaborative management module; Step 4: The early warning execution module controls the operation of multiple types of early warning execution devices based on the equipment status diagnosis data, completes the early warning action of the abnormal equipment status, and feeds back the early warning execution data to the equipment status diagnosis module; In step 5, the collaborative management module adjusts the working parameters of each module according to the device status diagnostic data, realizes the collaborative management of the multi-source sensor monitoring process, and forms a closed-loop monitoring and control system.

[0017] Compared with the prior art, the present invention has the following beneficial effects: In terms of data collection, a multi-type sensor array, including piezoelectric vibration sensors, infrared temperature sensors, Hall-effect current sensors, and laser position sensors, is deployed to comprehensively collect raw data from multiple sources, including real-time vibration data, surface temperature data, motor current data, key component location data, and historical equipment status records during subway equipment operation. Different sensors collect data according to preset cycles, ensuring access to equipment status information across different time dimensions. This provides rich and comprehensive data support for subsequent equipment status analysis, effectively addressing the limitations of traditional single-sensor monitoring, which often results in limited data and coverage.

[0018] In terms of data processing and fusion, the information fusion module sequentially performs data calibration and feature fusion operations on the raw data of multi-source equipment status. Data calibration uses a median filter algorithm to suppress noise, and spline interpolation fills missing values ​​to ensure data integrity and continuity. Feature extraction extracts key features such as vibration frequency, temperature gradient, and current fluctuation from the calibrated data and performs correlation calculations. Evidence theory fusion methods are then used to determine the comprehensive characterization value, ultimately generating analyzable equipment status fusion information. This series of operations effectively improves data quality and usability. Through the fusion of multi-source data, it can more accurately reflect the actual operating status of the equipment and reduce the impact of data errors and interference on monitoring results.

[0019] For equipment status diagnosis, the module integrates equipment status information and early warning execution data, employing a pattern-matching approach. This module extracts reference templates from historical equipment status records, compares real-time data with these templates, and sets a discrepancy threshold to identify abnormal equipment states. Ultimately, it generates equipment status diagnostic data containing abnormality probability data, equipment status category data, and diagnostic indicator data. This approach leverages historical data insights, improving the accuracy and reliability of equipment status diagnosis and enabling timely detection of potential equipment failures and abnormalities.

[0020] In terms of early warning execution, the early warning execution module implements tiered early warnings, signal output control, and synchronized execution based on device status diagnostic data. Different warning levels and trigger conditions are set based on the device's criticality and diagnostic data. The output type of the early warning device is adjusted through a signal conversion interface, and the action timing of each early warning device is coordinated using a time synchronization controller. This ensures the accuracy and effectiveness of early warnings, avoids mutual interference during the early warning process, and ensures timely and accurate early warning signals when equipment anomalies occur, providing ample time for equipment maintenance and troubleshooting.

[0021] In terms of collaborative management, the collaborative management module adjusts the operating parameters of each module based on equipment status diagnostic data, achieving collaborative management of the multi-source sensor monitoring process and forming a closed-loop monitoring and control system. Through data reception and processing, management strategy generation, command issuance and execution, and feedback optimization and adjustment, corresponding management strategies are generated for different equipment states, such as parameter optimization strategies for sub-healthy equipment states and emergency intervention strategies for abnormal equipment states. These adjustment strategies are then sent to the corresponding modules, and feedback information is collected and optimized after adjustment. This enables continuous optimization and dynamic adjustment of the monitoring system, improving the operating efficiency and reliability of the entire monitoring system and ensuring that subway equipment is always in good operating condition. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a working principle diagram of the subway equipment status monitoring system with multi-source sensor fusion according to the present invention; Figure 2 This is the design diagram of the information fusion module processing flow; Figure 3 A design diagram of the workflow of the early warning execution module; Figure 4 Design diagram of the diagnostic process of the equipment status diagnosis module; Figure 5 This is a design diagram of the collaborative process of the collaborative management module. DETAILED DESCRIPTION

[0023] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0024] See also Figure 1-Figure 5 The present invention relates to a multi-source sensor fusion subway equipment status monitoring system, which includes: a data acquisition module, an information fusion module, an equipment status diagnosis module, an early warning execution module and a collaborative management module. The specific implementation method is as follows: The data acquisition module implements data acquisition by deploying a multi-type sensor array to obtain the original data of multi-source device status and transmits the original data of multi-source device status to the information fusion module.

[0025] The information fusion module performs data calibration and feature fusion operations on the original data of multi-source equipment status in turn to obtain fused equipment status information, and then sends the information to the equipment status diagnosis module and the early warning execution module.

[0026] The equipment status diagnosis module combines the fusion equipment status information and early warning execution data, adopts the equipment status diagnosis method based on pattern matching, obtains the equipment status diagnosis data, and transmits it to the collaborative management module.

[0027] The early warning execution module controls the operation of multiple types of early warning execution devices based on the equipment status diagnosis data, completes the early warning action of abnormal equipment status, and feeds back the early warning execution data to the equipment status diagnosis module.

[0028] The collaborative management module adjusts the working parameters of each module according to the equipment status diagnostic data, realizes the collaborative management of the multi-source sensor monitoring process, and forms a closed-loop monitoring and control system.

[0029] Example 1: In the multi-source sensor fusion subway equipment status monitoring system, the data acquisition module undertakes the important task of acquiring multi-source equipment status raw data. The core function of this module is to implement data acquisition by deploying a multi-type sensor array, thereby obtaining multi-source equipment status raw data related to the operating status of subway equipment and transmitting this multi-source equipment status raw data to the information fusion module. The multi-source equipment status raw data mentioned here specifically covers information from multiple dimensions, including real-time vibration data, equipment surface temperature data, motor operating current data, key component location data, and historical equipment status record data. Among them, historical equipment status record data is further refined into historical abnormal time record data, historical parameter adjustment record data, and historical warning effect feedback data.

[0030] The multi-source sensor data collection emphasizes the overall collection property of all data related to the device status obtained through a multi-type sensor array; The multi-source device status raw data refers to the unprocessed raw data in the set; In terms of hardware deployment, the data acquisition module incorporates various sensor types, each performing a distinct function and collectively forming a complete sensor array. Piezoelectric vibration sensors are used to collect vibration data. These sensors can sensitively detect vibration signals generated during equipment operation and convert them into electrical signals. Infrared temperature sensors collect surface temperature data by sensing infrared radiation emitted from the surface. Hall-effect current sensors are used to collect motor current data. Based on the Hall effect principle, they accurately measure current magnitude. Laser position sensors collect position data for key components, using laser ranging to determine component location information.

[0031] When operating, these sensors do not collect data randomly, but regularly according to preset cycles. Preset cycles include collecting data every 3 seconds, every 6 seconds, and every 15 seconds. Different collection cycles are used to adapt to the characteristics and monitoring needs of different types of data. For parameters that change relatively quickly and require real-time attention, such as vibration data, a shorter collection cycle, such as every 3 seconds, may be used to ensure that subtle changes can be captured in a timely manner. For parameters that change relatively slowly, such as the position data of certain key components, a longer collection cycle, such as every 15 seconds, may be used. This can not only meet monitoring requirements, but also reduce the amount of data and improve the operating efficiency of the system.

[0032] During data collection, each sensor operates continuously according to its own collection cycle. For example, a piezoelectric vibration sensor monitors the vibration of the device in real time at each scheduled collection moment, converting the sensed vibration intensity, frequency, and other information into a corresponding electrical signal. Similarly, an infrared temperature sensor regularly checks the device's surface temperature, a Hall-effect current sensor monitors the motor's current in real time, and a laser position sensor periodically measures the position of key components. These sensors perform preliminary processing on the raw data collected from multiple sources of device status before transmitting it to the information fusion module via a specific transmission channel.

[0033] It is important to ensure data accuracy and integrity during data transmission to avoid data loss or distortion. This requires the transmission channel to have good stability and anti-interference capabilities, ensuring that the raw data of multi-source device status can smoothly reach the information fusion module from the data acquisition module, providing a reliable foundation for subsequent data analysis and processing.

[0034] In addition, the collection and storage of historical equipment status record data is also a key task of the data acquisition module. The system records in real time the time points when anomalies occur during equipment operation, forming historical anomaly time record data. At the same time, every adjustment to equipment parameters, including the time, parameters adjusted, and adjustment range, is recorded in detail to generate historical parameter adjustment record data. After each warning action is executed, feedback on the warning effect, such as whether the warning was timely and the warning method was appropriate, is also recorded to form historical warning effect feedback data. This historical equipment status record data is of great reference value for subsequent equipment status analysis and diagnosis. It can help the system better understand the historical operation of the equipment and provide a basis for judging the current equipment status.

[0035] Example 2: In the multi-source sensor fusion subway equipment status monitoring system, the information fusion module primarily processes the multi-source equipment status raw data transmitted by the data acquisition module to generate fused equipment status information. This module's workflow involves sequentially performing data calibration and feature fusion on the multi-source equipment status raw data, ultimately generating fused equipment status information that is then sent to the equipment status diagnosis module and the early warning execution module. The entire processing process encompasses several key steps: data calibration, feature extraction, fusion calculation, and information generation.

[0036] Data calibration is the first step in the information fusion module's processing of multi-source device status raw data. Its purpose is to remove interference and correct errors. This process is divided into two parts: noise suppression and missing value filling. In noise suppression, a median filter algorithm is used to smooth high-frequency noise in the data. For the collected multi-source device status raw data, a window is selected, centered on a data point, encompassing that point and several adjacent points. The data within the window is sorted, and the median value is taken as the new value for that point. This method effectively reduces the impact of high-frequency noise on the data and smooths the data. Missing value filling, on the other hand, uses spline interpolation to supplement missing sensor data within an acquisition cycle. If sensor data is missing due to various reasons within a certain acquisition cycle, the spline interpolation method constructs a smooth curve based on the distribution of adjacent data points to estimate the value of the missing point, thereby ensuring the integrity and continuity of the multi-source device status raw data.

[0037] After data calibration is complete, the feature extraction phase begins. This phase primarily involves extracting key features related to the device's status from the calibrated data. These include vibration frequency features, temperature gradient features, and current fluctuation features. For example, the vibration frequency of a device often exhibits different characteristics during normal and abnormal operation. By analyzing the calibrated vibration data and extracting the frequency components, the vibration frequency features can be derived. The temperature gradient feature is obtained by calculating the rate of change of the device's surface temperature over a specific period of time, reflecting the device's temperature trend. The current fluctuation feature analyzes and extracts the fluctuations in the motor's operating current data. After extracting these key features, they must be correlated. Because the device's operating status is often the result of the interaction of multiple features, correlation calculations can reveal the inherent connections between these features, providing more valuable information for subsequent fusion calculations.

[0038] The correlation calculation involves extracting key features from the calibrated data, including vibration frequency, temperature gradient, and current fluctuation. The purpose of performing the correlation calculation on these key features is to capture the inherent connections between these different features. For example, an abnormal change in vibration frequency may be associated with an increase in the temperature gradient or an increase in current fluctuation. By analyzing the temporal or numerical correspondence between these features, the mutual influence and degree of correlation between them are quantified, forming correlated feature data. This provides a foundation for subsequently determining the comprehensive characterization values ​​of vibration intensity, temperature distribution, and current stability using evidence theory fusion methods.

[0039] The result of the correlation calculation after extracting the key features is the correlated feature data. The correlated feature data refers to the key features such as vibration frequency features, temperature change gradient features, current fluctuation features extracted from the calibration data, which are quantitatively processed by analyzing their mutual influence and correlation degree during the operation of the equipment. These data are no longer isolated single feature information, but a comprehensive feature embodiment of the intrinsic connection between different features. They can more comprehensively reflect the state correlation of multiple features under the joint action of multiple features in the operation of the equipment, and provide more valuable basic data for the subsequent determination of the comprehensive characterization value of vibration intensity, temperature distribution and current stability based on the evidence theory fusion method.

[0040] Next comes the fusion calculation and processing phase. Based on the correlated feature data, an evidence theory fusion method is employed to determine a comprehensive representation value for vibration intensity, temperature distribution, and current stability. This method is an effective method for processing uncertain information, integrating information provided by multiple sensors to produce a more reliable result. In specific applications, the credibility of each feature data set must first be determined. Then, a fusion calculation is performed according to the rules of evidence theory. The final result is a single value that comprehensively reflects vibration intensity, temperature distribution, and current stability. This value is the comprehensive representation value.

[0041] During the condition monitoring process for subway traction motors, the information fusion module, after data calibration and feature extraction, generates correlated data for vibration frequency, temperature gradient, and current fluctuation characteristics. The vibration frequency characteristic indicates that the motor vibration amplitude is slightly above the normal range. Combined with historical data, this characteristic reflects a minor equipment anomaly with a confidence level of 0.6. The temperature gradient characteristic indicates that the motor temperature rise rate is within the normal range, with a confidence level of 0.8 indicating normal equipment operation. The current fluctuation characteristic indicates that the instantaneous current peak value slightly exceeds the standard value, with a confidence level of 0.5 indicating a minor equipment anomaly. Based on the evidence theory fusion method, the credibility of each feature data is first determined. Considering the correlation between abnormal vibration and current characteristics, the credibility of both is comprehensively considered in the fusion calculation. The high-confidence normal result of the temperature characteristic provides support for the overall condition. These credibility values ​​are fused using evidence combination rules to ultimately generate a comprehensive value that comprehensively represents the traction motor's vibration intensity, temperature distribution, and current stability. This value is then used to generate analyzable fused equipment condition information.

[0042] The information generation and processing phase converts the comprehensive characterization value into analyzable fused device status information, forming fused device status information. Since the comprehensive characterization value is typically a numerical value, it needs to be converted into a more intuitive and understandable form to facilitate subsequent device status diagnosis and early warning. For example, the comprehensive characterization value can be compared with a preset device status level to determine the device's current status level. The relevant information is then organized and formatted to form the final fused device status information.

[0043] In the subway brake system's condition monitoring, the information fusion module calculates and synthesizes comprehensive characterization values ​​for the brake system's vibration intensity, temperature distribution, and current stability. The vibration intensity characterization value is 0.7 (corresponding to a slightly abnormal range), the temperature distribution characterization value is 0.3 (corresponding to a normal range), and the current stability characterization value is 0.4 (corresponding to a basically stable range). The information fusion module organizes the corresponding status descriptions for these comprehensive characterization values, clearly labeling the brake system's device name as "brake system assembly," the vibration intensity status as "slightly abnormal (vibration frequency fluctuation exceeds the normal range by 5%)," the temperature distribution status as "normal (temperature gradient within the standard range)," and the current stability status as "basically stable (current fluctuation amplitude does not exceed the safety threshold)." The module also adds the acquisition time of each characteristic data as "2025-07-25 10:30:00" and the data source sensor type as "piezoelectric vibration sensor, infrared temperature sensor, Hall-effect current sensor." This information is structured and integrated according to the device status feature classification, status description, and collected information format to form the final fused device status information, which is then sent to the device status diagnosis module and the early warning execution module for subsequent processing.

[0044] Throughout the information fusion module's operation, each processing link is closely interconnected and influences each other. The effectiveness of data calibration processing directly affects the accuracy of feature extraction, which in turn significantly impacts the results of the fusion calculations and, ultimately, the reliability of the generated fused device status information. Therefore, each link must strictly adhere to prescribed methods and processes to ensure that the information fusion module can effectively integrate multi-source data and provide accurate and reliable fused device status information to the device status diagnosis module and the early warning execution module. This enables the entire monitoring system to more accurately determine the device's operating status, promptly detect abnormalities, and issue early warnings.

[0045] The information fusion module's processing of multi-source raw equipment status data is a complex and sophisticated process, requiring not only advanced algorithms and methods but also a deep understanding of the equipment's operating characteristics. Through multi-source data calibration, feature extraction, fusion calculations, and information generation, the information fusion module transforms the previously scattered and disorganized raw data from multiple sources into valuable fused equipment status information. This provides strong support for equipment status monitoring and fault diagnosis for subway equipment, enabling the system to more accurately grasp the equipment's operating status and providing a crucial basis for equipment maintenance and management.

[0046] Example 3: In the multi-source sensor fusion subway equipment status monitoring system, the early warning execution module plays a crucial role in implementing early warning operations based on equipment status diagnostic data. Its core function is to control the operation of multiple early warning execution devices based on equipment status diagnostic data, complete early warning actions for equipment status anomalies, and feed back early warning execution data to the equipment status diagnostic module.

[0047] Equipment-level warnings set specific warning levels and trigger conditions based on the criticality of each device. In subway systems, different pieces of equipment vary in importance. For example, traction motors, as critical power equipment, have failures that could potentially cause train disruptions, while failures in some auxiliary equipment have relatively minor impacts. Therefore, higher warning levels and stricter trigger conditions are set for critical equipment such as traction motors, gearboxes, and braking systems. For example, a level 1 warning may be triggered when diagnostic data for a traction motor's vibration frequency characteristics shows a difference of 80% from historical normal data. For auxiliary equipment like air conditioning systems, a level 2 warning may be triggered only when the difference reaches 100% of the threshold. Warning levels are typically categorized into multiple levels, such as level 1, level 2, and level 3, each corresponding to different warning methods and response processes. Trigger conditions should comprehensively consider factors such as the equipment's operating conditions and historical fault data to ensure accurate and timely warnings and avoid false alarms and missed alerts.

[0048] Equipment-level warnings are based on the importance of each device in the subway system, with corresponding warning levels and trigger conditions set for different devices. Key equipment, such as traction motors and gearboxes, have higher warning levels and stricter trigger conditions due to the greater impact of failures, while auxiliary equipment, such as air conditioners, have relatively looser conditions. Different warning levels correspond to different warning methods and response processes. The signal output control link is based on these graded warning results. The signal conversion interface adjusts the output type of each warning device. For example, when the traction motor triggers a level 1 warning, the warning control signal is converted into a switching signal for the sound and light alarm, a digital signal for the display screen, and a communication signal for the SMS notification module. When auxiliary equipment, such as the air conditioner, triggers the corresponding warning level, the signal conversion interface is also used to adapt the required warning device output type. This ensures that when equipment of different importance triggers the corresponding warning level, the warning information can be effectively transmitted through the corresponding warning device, meeting the warning needs of different equipment.

[0049] The signal output control link adjusts the output type of each warning device through the signal conversion interface to meet the warning requirements of different devices. A subway equipment status monitoring system may include multiple warning devices, such as audible and visual alarms, display screens, and SMS notification modules. Different warning devices require different types of signal inputs. For example, audible and visual alarms require digital signals to control their sound and light output; display screens require digital signals to display warning information; and SMS notification modules require signals supported by communication protocols to send SMS messages. The signal conversion interface converts the warning control signal into the output type required by each warning device based on the equipment status diagnostic data received by the warning execution module and the preset warning strategy. For example, when the traction motor triggers a level 1 warning, the signal conversion interface converts the warning control signal into a digital signal for the audible and visual alarm, causing it to emit a strong audible and visual alarm; converts it into a digital signal for the display screen, displaying the words "Traction motor abnormality, level 1 warning" on the train monitoring screen; and converts it into a communication signal for the SMS notification module, sending a warning SMS message to relevant maintenance personnel. This ensures that the warning requirements of different devices are met, allowing relevant personnel to obtain warning information promptly through multiple channels.

[0050] The synchronization of execution actions requires a time synchronization controller to coordinate the start and stop times of each warning execution device to avoid interference during the warning process. In actual warning processes, multiple warning devices, such as audible and visual alarms, display screens, and SMS notification modules, may need to be activated simultaneously. If the start and stop times of these devices are not synchronized, this can lead to confusion in warning information, hindering personnel's interpretation and handling of the warning. Based on the warning strategy and device status diagnostic data, the time synchronization controller generates a unified time reference and control signals to control each warning execution device to start warning actions at the same time and stop them synchronously after a specified time. For example, when a device triggers a level 2 warning, the time synchronization controller will simultaneously activate the audible and visual alarms, display screens, and SMS notification modules within one second of receiving the warning signal. The audible and visual alarms will continue to sound for 10 minutes, and the display screen will continue to display the warning message until the fault is resolved. The SMS notification module will send a warning SMS message upon activation. After 10 minutes, the time synchronization controller will control the audible and visual alarms to stop sounding, but the display screen and SMS notification module will remain in the warning device state until the device returns to normal or receives a new control signal. Through this time synchronization control, it is ensured that each early warning execution device works in coordination, forming an orderly early warning process, avoiding mutual interference, and improving the effectiveness and reliability of the early warning.

[0051] Taking a subway train gearbox as an example, when the data acquisition module detects abnormal vibration data, after processing by the information fusion module, the equipment status diagnosis module determines that the gearbox is in an abnormal operating state. It then generates corresponding equipment status diagnostic data and transmits it to the early warning execution module. Upon receiving the diagnostic data, the early warning execution module first initiates a graded equipment early warning. Because the gearbox is a critical piece of equipment, its early warning level is set to level 2, triggered when the vibration difference exceeds a preset threshold. Combined with the diagnostic data, it determines the need for an early warning and generates an early warning control signal to control the early warning execution device. The signal output control component then converts the early warning control signal, via a signal conversion interface, into a high-frequency flashing and beeping signal for the audible and visual alarm, a text prompt for the onboard monitoring display, and a text message for the maintenance personnel's handheld terminal. Simultaneously, the execution action synchronization component, through a time synchronization controller, ensures that the audible and visual alarm activates within 3 seconds and continues to sound for 15 minutes. The display refreshes in real time to display the gearbox abnormality information, and a text message is sent to the designated maintenance personnel within 10 seconds. During the early warning execution process, the early warning execution module feeds back early warning execution data, such as the alarm activation time and the operating status of each device, to the equipment status diagnosis module, providing reference for subsequent equipment status analysis and diagnosis. When the gearbox equipment status returns to normal, the early warning execution module receives new diagnostic data and controls each early warning device to stop the early warning action through the same process, completing a complete early warning operation.

[0052] Through the coordinated efforts of three key components—equipment-level early warning, signal output control, and synchronized execution—the early warning execution module accurately activates the appropriate early warning device based on the device's status and criticality, issuing warnings in an appropriate manner and at the appropriate time. This allows personnel to promptly identify equipment anomalies and take appropriate measures, thereby ensuring the safe operation of subway equipment. The entire early warning process is seamlessly linked, with each step strictly following pre-set rules and procedures to ensure the accuracy, timeliness, and effectiveness of early warnings.

[0053] Example 4: In the multi-source sensor fusion subway equipment status monitoring system, the equipment status diagnosis module is primarily responsible for analyzing real-time equipment status during operation. Its core function is to combine the fused equipment status information output by the information fusion module with the early warning execution data fed back by the early warning execution module. Using a pattern-matching-based equipment status diagnosis method, it generates equipment status diagnostic data and transmits this data to the collaborative management module. This process specifically includes four key steps: historical pattern matching, real-time data comparison, abnormal equipment status identification, and diagnostic result generation. Each step is closely linked to achieve accurate equipment status diagnosis.

[0054] The historical pattern matching process needs to extract historical equipment status data that is identical or similar to the current equipment operating condition from the historical equipment status record data as a reference template. During the operation of subway equipment, the operating status of the equipment varies under different operating conditions such as starting, constant speed driving, braking, etc. Therefore, historical pattern matching needs to first determine the current equipment operating condition. For example, when the subway train is in a constant speed driving condition, the equipment status diagnosis module will first filter out historical data under the same operating condition from the historical equipment status record data. The historical equipment status record data contains information such as vibration data, temperature data, current data of the equipment under different operating conditions, as well as historical abnormal time records, historical parameter adjustment records, etc. During the screening process, matching will be performed based on factors such as operating condition type, operating time, and ambient temperature to find the historical equipment status data that is closest to the current operating condition. Assuming the train is currently running at a constant speed of 60 km / h at an ambient temperature of 30°C, the equipment status diagnosis module will search historical data for equipment status data when running at similar ambient temperatures (such as 28°C-32°C) and the same speed range (such as 55 km / h-65 km / h), and integrate this data as a reference template.

[0055] The real-time data comparison phase compares the fused device status information against the reference template, calculating the differences in vibration, temperature, and current. The fused device status information includes key characteristic data, such as vibration frequency, temperature gradient, and current fluctuation, processed by the information fusion module. For example, the reference template records the vibration frequency range for normal operation of the device under specific operating conditions, such as 100Hz-200Hz. However, the vibration frequency characteristic in the real-time fused device status information may show a frequency range of 180Hz-250Hz. In this case, the difference between the two must be calculated, meaning the upper limit exceeds 50Hz and the lower limit exceeds 80Hz. For the temperature gradient characteristic, the reference template may specify that the hourly gradient of the device surface temperature under specific operating conditions must not exceed 5°C, while the real-time data shows a temperature gradient of 8°C / hour, resulting in a difference of 3°C / hour. The current fluctuation characteristic comparison compares the real-time current fluctuation range with the normal fluctuation range in the reference template and calculates the difference in fluctuation amplitude. During the comparison process, each characteristic data item must be carefully compared to ensure that no information that may reflect the device status is missed.

[0056] The abnormal equipment status determination phase uses vibration, temperature, and current variances within set thresholds to determine if the equipment is operating abnormally. If the variance exceeds the threshold, the equipment is flagged as abnormal. Setting the variance threshold requires comprehensive consideration of factors such as the equipment type, criticality, and historical fault data. For critical equipment such as traction motors, the vibration variance threshold might be set at ±10% of the reference template value, the temperature variance threshold at ±5°C, and the current variance threshold at ±15%. For auxiliary equipment such as air conditioning systems, the thresholds might be more relaxed, such as ±20% for vibration, ±8°C for temperature, and ±25% for current. For example, in a traction motor, if the vibration frequency variance exceeds the ±10% threshold during real-time data comparison—for example, a calculated variance of +12%—the vibration characteristic is considered abnormal and flagged as abnormal. A temperature gradient variance of +6°C, exceeding the ±5°C threshold, is also flagged as abnormal. When multiple features exhibit abnormalities simultaneously, a comprehensive assessment is needed to determine whether the overall equipment operating status is abnormal.

[0057] The diagnostic result generation phase integrates the comparison results and abnormal equipment status information to form equipment status diagnostic data. Equipment status diagnostic data specifically includes abnormality probability data, equipment status category data, and diagnostic indicator data. Abnormality probability data represents the probability of an abnormality occurring in the current equipment operation, ranging from 0 to 100%. It is calculated by combining the difference values ​​of each feature and their weights. The weights are determined based on the importance of each feature. For example, the vibration feature of a traction motor might have a weight of 40%, the temperature feature 35%, and the current feature 25%. Assuming that an abnormal vibration feature contributes 40% to the abnormality probability, an abnormal temperature feature 30%, and a normal current feature, the abnormality probability data is 70%. Equipment status category data represents identifiable equipment operating status types, including stable, normal, sub-healthy, and abnormal. Equipment status is categorized based on the abnormality probability data and the abnormality of each feature. For example, an abnormality probability of 0-20% indicates stable operation, 21-40% indicates normal operation, 41-70% indicates sub-healthy operation, and 71-100% indicates abnormal operation. Diagnostic indicator data provides the basis for determining equipment status, including vibration difference recording data, temperature difference recording data, and current difference recording data. This data details the actual value of each characteristic, the reference template value, the difference value, and whether it exceeds the threshold.

[0058] Taking the subway train braking system as an example, when the train is operating under braking conditions, the data acquisition module collects multiple-source raw device status data, including vibration, temperature, and current data, from a deployed sensor array. This data is then transmitted to the information fusion module. The information fusion module calibrates, extracts features, and performs fusion calculations on the multiple-source raw device status data to generate fused device status information, which is then sent to the device status diagnosis module. The device status diagnosis module first performs historical pattern matching, searching for historical data from the historical device status records that is identical or similar to the current braking condition. For example, this module uses brake system device status data from previous periods under the same braking intensity and ambient temperature as a reference template. The module then compares the vibration frequency, temperature gradient, and current fluctuation characteristics in the fused device status information against the reference template, calculating the difference. Assume that the vibration frequency reference template for the braking system is 80Hz-150Hz, the real-time vibration frequency is 170Hz, and the variance is +20Hz, exceeding the preset ±15% threshold (i.e., ±22.5Hz). This is within the specified range. The temperature gradient reference template is for a temperature rise of no more than 10°C per braking cycle, while the real-time temperature gradient is 12°C, resulting in a variance of +2°C, exceeding the preset ±15% threshold (i.e., ±1.5°C). The current fluctuation reference template is for a normal fluctuation range of ±5%, while the real-time current fluctuation is +8%, resulting in a variance of +3%, exceeding the preset ±5% threshold. Based on these variances, the abnormal device status determination process identifies any temperature and current characteristic variances that exceed the threshold and indicates an abnormal device status. The diagnostic result generation phase calculates abnormality probability data. Assuming a 40% weight for temperature, a 30% weight for current, and a 30% weight for vibration, the temperature abnormality contributes 40% × (2 / 1.5) ≈ 53.3%, and the current abnormality contributes 30% × (3 / 5) = 18%, resulting in a total abnormality probability of approximately 71.3%. Therefore, the device status category data is determined to indicate an abnormal operating state, and the diagnostic indicator data details the differences in each characteristic. The device status diagnostic module transmits the generated device status diagnostic data to the collaborative management module, providing a basis for subsequent collaborative management.

[0059] Through the orderly implementation of four steps—historical pattern matching, real-time data comparison, abnormal equipment status identification, and diagnostic result generation—the equipment status diagnosis module, based on pattern matching, combines and integrates equipment status information and early warning execution data to comprehensively and accurately analyze and diagnose the equipment's real-time operating status. This generates detailed equipment status diagnostic data, providing strong support for subway equipment maintenance and management, enabling relevant personnel to promptly understand the equipment's operating status and take appropriate measures to ensure the safe and stable operation of subway equipment. The entire diagnostic process strictly adheres to preset rules and procedures, ensuring the accuracy and reliability of diagnostic results and laying a solid foundation for subsequent early warning and collaborative management.

[0060] Example 5: In the multi-source sensor fusion subway equipment status monitoring system, the collaborative management module plays a crucial role in coordinating the operation of various modules and optimizing the monitoring process. Its core function is to adjust the operating parameters of each module based on equipment status diagnostic data, enabling collaborative management of the multi-source sensor monitoring process and forming a closed-loop monitoring and control system. This process specifically includes four key steps: data reception and processing, management strategy generation, command issuance and execution, and feedback optimization and adjustment. These steps work together to ensure the efficient and stable operation of the entire monitoring system.

[0061] The data reception and processing phase primarily involves receiving equipment status diagnostic data and analyzing the anomaly information and comparison results. Equipment status diagnostic data is generated by the equipment status diagnostic module and includes anomaly probability data, equipment status category data, and diagnostic indicator data. Taking the diagnostic data of a subway train traction motor as an example, the diagnostic data received by the collaborative management module may indicate an anomaly probability of 75%, an abnormal operating state for the equipment status category, and diagnostic indicator data indicating vibration differences exceeding a threshold of 20%, temperature differences exceeding a threshold of 15%, and current differences exceeding a threshold of 10%. After receiving the data, the collaborative management module analyzes this data and extracts key anomaly information, such as which characteristic parameters are abnormal, the severity of the anomaly, and the comparison results between real-time data and historical reference templates. This provides a basis for subsequent management strategy generation.

[0062] The management strategy generation phase involves generating adjustment strategies for abnormal equipment or parameters based on abnormal information and comparison results. Adjustment strategy types include vibration parameter correction, temperature control compensation, and current limit adjustment. Strategies also need to be generated based on the equipment's status, such as parameter optimization for sub-healthy operating conditions, emergency intervention for abnormal operating conditions, and parameter fine-tuning for normal operating conditions. Continuing with the example of an abnormal traction motor, the management strategy generation phase generates specific adjustment strategies based on vibration, temperature, and current anomalies. For vibration anomalies, a vibration parameter correction strategy might be generated, such as adjusting the hardness of the shock-absorbing pads on the motor mounting base to reduce vibration intensity. For temperature anomalies, a temperature control compensation strategy might be generated, such as increasing the cooling fan speed or activating the backup cooling system to reduce motor temperature. For current anomalies, a current limit adjustment strategy might be generated, such as temporarily reducing the motor's output power to keep the current within a safe range. Furthermore, due to the abnormal equipment operating state, an emergency intervention strategy might be generated, such as triggering a level 1 alert in the early warning execution module, notifying maintenance personnel for immediate inspection and repair.

[0063] The instruction issuance and execution link is to convert the adjustment strategy into a management instruction and send it to the corresponding module. Among them, the information fusion module receives the parameter correction instruction, and the early warning execution module receives the action adjustment instruction. For example, for the vibration parameter correction strategy of the traction motor, the collaborative management module will generate an instruction to adjust the vibration data calibration threshold in the information fusion module and send it to the information fusion module, so that the information fusion module will use the new calibration threshold when processing the vibration data in the subsequent process to improve the accuracy of data processing; at the same time, the emergency intervention strategy that triggers the first-level early warning is converted into an instruction and sent to the early warning execution module, requiring the early warning execution module to activate the sound and light alarm, send SMS notifications and other first-level early warning actions. During the instruction issuance process, it is necessary to ensure the accuracy and timeliness of the instructions to avoid further deterioration of the equipment status due to instruction errors or delays.

[0064] The feedback optimization and adjustment phase collects adjusted, fused device status information and diagnostic data, verifies the effectiveness of the adjustments, and further optimizes the management strategy, forming a continuously optimized collaborative management process. After the adjustment strategy has been implemented for a period of time, the data acquisition module recollects the device's operating data. The information fusion module then processes it to generate new fused device status information. The device status diagnosis module generates new diagnostic data based on the new fused device status information and the early warning execution data, and feeds this data back to the collaborative management module. The collaborative management module analyzes the feedback data to evaluate the effectiveness of the adjustment strategy. For example, after implementing the adjustment strategy for the traction motor, the new diagnostic data shows that the abnormality probability has dropped to 40%, the device status category has changed to sub-healthy operation, and the vibration, temperature, and current differences have all decreased, indicating that the adjustment strategy has achieved some results. However, the vibration difference still exceeds the threshold of 5%, the temperature difference exceeds the threshold of 3%, and the current difference has returned to normal. Based on this feedback, the collaborative management module optimizes the original adjustment strategy, such as further adjusting the hardness of the shock absorber pad to further reduce vibration or adjusting the cooling fan operating time to optimize temperature control. A new management strategy is generated and re-issued for execution until the device status returns to normal.

[0065] For example, a subway train gearbox's equipment status diagnostic data indicates a subhealthy operating state, with an abnormality probability of 60%, a vibration variance exceeding a threshold of 10%, and a temperature variance exceeding a threshold of 8%, while current data is normal. The collaborative management module receives this diagnostic data and analyzes the abnormal vibration and temperature parameters. The management strategy generation phase generates parameter optimization strategies tailored to this subhealthy operating state. For example, for vibration anomalies, a strategy adjusts the gearbox lubricant viscosity to reduce vibration during gear meshing; for temperature anomalies, a strategy increases the frequency of cleaning the gearbox heat sink to improve heat dissipation efficiency. The instruction issuance and execution phase converts these strategies into instructions and sends them to the information fusion module and the early warning execution module. The information fusion module receives parameter correction instructions related to lubricant viscosity and considers the impact of lubricant viscosity changes on temperature when processing temperature data. The early warning execution module receives notification of the parameter optimization strategy and may initiate a secondary alert to alert maintenance personnel to the gearbox's equipment status. After a period of time, the feedback optimization and adjustment phase collects new diagnostic data showing that the abnormality probability has dropped to 30%, the vibration variance has exceeded a threshold of 5%, and the temperature variance has returned to normal. After evaluation, the collaborative management module concluded that the temperature adjustment strategy was effective and continued to optimize the vibration adjustment strategy, such as increasing the fine-tuning of the gear meshing clearance and generating new instructions for execution until the gearbox equipment status returned to normal operation.

[0066] Through a repetitive cycle of four steps—data reception and processing, management strategy generation, command issuance and execution, and feedback optimization and adjustment—the collaborative management module achieves dynamic collaborative management of the multi-source sensor monitoring process. It promptly adjusts the operating parameters and strategies of each module based on the equipment's real-time status and diagnostic results, continuously optimizing the monitoring process. This allows the entire system to better adapt to changes in equipment operation, improves monitoring accuracy and reliability, and ensures the safe and stable operation of subway equipment. The entire collaborative management process forms a closed loop, with each step closely linked, ensuring continuous system improvement and optimization, providing strong support for equipment status monitoring in subways.

[0067] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "comprises," "includes," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that includes a list of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, method, article, or apparatus.

[0068] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A multi-source sensor fusion subway equipment status monitoring system, characterized by: It includes data acquisition module, information fusion module, equipment status diagnosis module, early warning execution module and collaborative management module; The data acquisition module is used to acquire a multi-source sensor data set related to the equipment status during the operation of subway equipment. Specifically, it implements data acquisition by deploying a multi-type sensor array to obtain multi-source equipment status raw data, and transmits the multi-source equipment status raw data to the information fusion module; The information fusion module is used to process the multi-source device status raw data and generate device status fusion information, specifically by sequentially performing data calibration and feature fusion operations on the multi-source device status raw data to obtain fused device status information, and sending the fused device status information to the device status diagnosis module and the early warning execution module; The device status diagnosis module is used to analyze the real-time device status during device operation. Specifically, it combines the fused device status information with the warning execution data output by the warning execution module, adopts a pattern matching-based device status diagnosis method to obtain device status diagnosis data, and transmits the device status diagnosis data to the collaborative management module. The warning execution module is used to implement warning operations based on the equipment status diagnosis data, specifically to control the operation of multiple types of warning execution devices based on the equipment status diagnosis data, complete the warning action of the abnormal equipment status, and feed back the warning execution data corresponding to the warning action to the equipment status diagnosis module; The collaborative management module is used to coordinate the operation of each module and optimize the monitoring process. Specifically, it adjusts the working parameters of each module according to the equipment status diagnostic data, realizes the collaborative management of the multi-source sensor monitoring process, and forms a closed-loop monitoring and control system.

2. The multi-source sensor fusion subway equipment status monitoring system according to claim 1 is characterized by: The multi-source equipment status original data specifically includes real-time vibration data, equipment surface temperature data, motor operating current data, key component position data and historical equipment status record data; the historical equipment status record data specifically includes historical abnormal time record data, historical parameter adjustment record data and historical warning effect feedback data.

3. The subway equipment status monitoring system based on multi-source sensor fusion according to claim 2 is characterized by: In the information fusion module, the step of sequentially performing data calibration and feature fusion operations on the multi-source device status raw data to obtain fused device status information includes data calibration processing, feature extraction processing, fusion calculation processing, and information generation processing; The data calibration process is used to remove interference values ​​and correct errors in the original data of the multi-source device status, specifically by screening and correcting the original data of the multi-source device status by setting a data valid range to obtain calibrated data; The feature extraction process specifically extracts key features related to the device status from the calibrated data, including vibration frequency features, temperature gradient features, and current fluctuation features, and performs correlation calculation on the key features to obtain correlated feature data; The fusion calculation process specifically comprises determining the comprehensive characterization value of vibration intensity, temperature distribution and current stability by using the evidence theory fusion method based on the associated characteristic data; The information generation process specifically converts the comprehensive characterization value into analyzable device status fusion information to form the fused device status information.

4. The subway equipment status monitoring system based on multi-source sensor fusion according to claim 3 is characterized by: In the warning execution module, the operation of multiple types of warning execution devices is controlled based on the equipment status diagnostic data to complete the warning action of abnormal equipment status, specifically including equipment graded warning, signal output control and execution action synchronization; The device classification warning is specifically to set the warning level and trigger conditions for the corresponding device according to the importance of each device and the device status diagnostic data; The signal output control specifically adjusts the output types of multiple types of warning execution devices through the signal conversion interface to ensure that the warning requirements of different devices are met; The execution action synchronization is specifically to coordinate the start and stop time of each early warning execution device through a time synchronization controller to avoid mutual interference during the early warning process.

5. The subway equipment status monitoring system based on multi-source sensor fusion according to claim 4 is characterized by: In the device status diagnosis module, the step of combining the fused device status information with the warning execution data output by the warning execution module and adopting a pattern matching-based device status diagnosis method to obtain device status diagnosis data includes historical pattern matching, real-time data comparison, abnormal device status determination, and diagnosis result generation; The historical pattern matching is specifically to extract historical equipment status data that is the same or similar to the current equipment working condition from the historical equipment status record data as a reference template; The real-time data comparison specifically involves comparing the fusion device status information with the reference template item by item, and calculating the difference values ​​of vibration, temperature and current; The abnormal device status determination is specifically to determine whether the current device operation is abnormal based on the difference values ​​of vibration, temperature and current by setting a difference threshold range. If the difference value exceeds the threshold, it is marked as an abnormal device status; The diagnostic result is generated by integrating the comparison result and the abnormal device status information to form the device status diagnostic data.

6. The multi-source sensor fusion subway equipment status monitoring system according to claim 5 is characterized by: In the collaborative management module, the operating parameters of each module are adjusted according to the device status diagnostic data to achieve collaborative management of the multi-source sensor monitoring process, specifically including data reception and processing, management strategy generation, instruction issuance and execution, and feedback optimization and adjustment; The data receiving and processing is specifically receiving the device status diagnostic data and analyzing the abnormal information and comparison results therein; The management strategy generation is specifically to generate an adjustment strategy for abnormal equipment or parameters based on the abnormal information and the comparison results. The adjustment strategy includes a vibration parameter correction strategy, a temperature control compensation strategy, or a current limit adjustment strategy; The instructions are sent for execution, specifically converting the adjustment strategy into a management instruction and sending it to the corresponding module, wherein the information fusion module receives the parameter correction instruction, and the early warning execution module receives the action adjustment instruction; The feedback optimization and adjustment specifically involves collecting the adjusted fusion device status information and diagnostic data, verifying the adjustment effect and further optimizing the management strategy to form a continuously optimized collaborative management process.

7. The multi-source sensor fusion subway equipment status monitoring system according to claim 6 is characterized by: The multi-type sensor array deployed in the data acquisition module specifically includes a piezoelectric vibration sensor for collecting vibration data, an infrared temperature sensor for collecting temperature data, a Hall-type current sensor for collecting current data, and a laser position sensor for collecting position data; each sensor collects data according to a preset period, and the preset period includes collecting data once every 3 seconds, once every 6 seconds, and once every 15 seconds. The original data of the multi-source device status is obtained through the collection methods of different periods.

8. The subway equipment status monitoring system based on multi-source sensor fusion according to claim 7 is characterized by: The data calibration processing performed in the information fusion module specifically includes noise suppression processing and missing value filling processing; the noise suppression processing specifically uses a median filtering algorithm to smooth the high-frequency noise data; the missing value filling processing specifically uses a spline interpolation method to supplement the missing sensor data within the acquisition period to ensure the integrity and continuity of the original data of the multi-source device status.

9. The subway equipment status monitoring system based on multi-source sensor fusion according to claim 8 is characterized by: The device status diagnostic data obtained in the device status diagnostic module specifically includes abnormality probability data, device status category data, and diagnostic index data; the abnormality probability data is used to represent the probability of abnormality occurring in the current device operation, with a value range of 0 to 100%; the device status category data is used to represent the identifiable device operation status types, specifically including stable operation state, normal operation state, sub-healthy operation state, and abnormal operation state; the diagnostic index data is used to provide a basis for device status discrimination, specifically including vibration difference recording data, temperature difference recording data, and current difference recording data; The management strategies generated in the collaborative management module specifically include parameter optimization strategies for sub-healthy operating states, emergency intervention strategies for abnormal operating states, and parameter fine-tuning strategies for normal operating states.

10. A method for monitoring the status of subway equipment using multi-source sensor fusion, applied to the subway equipment status monitoring system using multi-source sensor fusion according to any one of claims 1 to 9, characterized in that: The steps include: Step 1: Deploy a multi-type sensor array to implement data collection, obtain multi-source device status raw data related to the device status during the operation of subway equipment, and transmit the multi-source device status raw data to the information fusion module; Step 2: The information fusion module sequentially performs data calibration and feature fusion operations on the multi-source device status raw data to obtain fused device status information, and sends the fused device status information to the device status diagnosis module and the early warning execution module; Step 3: The device status diagnosis module combines the fused device status information and the early warning execution data, adopts a pattern matching-based device status diagnosis method, obtains device status diagnosis data, and transmits the device status diagnosis data to the collaborative management module; Step 4: The early warning execution module controls the operation of multiple types of early warning execution devices based on the equipment status diagnosis data, completes the early warning action of the abnormal equipment status, and feeds back the early warning execution data to the equipment status diagnosis module; In step 5, the collaborative management module adjusts the working parameters of each module according to the device status diagnostic data, realizes the collaborative management of the multi-source sensor monitoring process, and forms a closed-loop monitoring and control system.

Images