Industrial equipment state monitoring and fault early warning method and system based on big data

Abstract

The invention relates to the technical field of industrial equipment monitoring, and discloses an industrial equipment state monitoring and fault early warning method and system based on big data, and the system comprises a data collection and fusion module, an intelligent diagnosis module and an early warning execution module. Through innovative design of a spatio-temporal feature fusion engine, the problem of spatio-temporal mismatch of vibration, temperature, current and other signals is solved, dynamic characteristics of millisecond-level vibration signals are captured through a long-short-term memory network, a spatial topological relation of equipment is modeled through a graph convolutional network, cross-scale alignment of multi-physical-quantity data is achieved, the limitation of physical significance isolation in a traditional method is broken through, and the method has a good application prospect. The composite fault recognition capability is improved, the diagnostic value of unstructured data such as sound emission is released, a full-life-cycle self-adaptive mechanism is constructed through a dynamic threshold algorithm, the working condition fluctuation tolerance is intelligently adjusted through a sensitivity coefficient, an early warning model continuously optimized along with the operation time is formed, the stiffness defect of a static threshold is changed, and the early warning effect is improved. And an accurate intelligent early warning system is established.

Description

Technical Field

[0001] The present invention relates to the technical field of industrial equipment monitoring, and in particular to a method and system for industrial equipment status monitoring and fault early warning based on big data. Background Art

[0002] In modern industrial production, the efficient and stable operation of equipment is the key to ensuring production continuity and safety. Existing methods for predicting equipment failure and health status monitor the physical conditions of the equipment itself, such as vibration, noise, temperature, voltage, current, and power. By selecting appropriate hardware monitoring sensors, the status signals of various mechanical parts of the equipment are continuously and multi-point collected and analyzed, thereby achieving equipment health status assessment and fault prevention. Modern industrial production is highly dependent on the health status of equipment. However, traditional monitoring methods have problems such as slow response and poor accuracy, and cannot meet the needs of massive data and real-time warnings. Especially for critical equipment, failures will lead to high downtime losses.

[0003] The main defects of existing industrial equipment condition monitoring and fault warning methods and systems include: 1. Barriers to collaboration of multi-source heterogeneous data: Industrial equipment monitoring requires simultaneous processing of multiple physical quantity signals such as vibration, temperature, current, and acoustic emission, resulting in triple mismatches: (1) Time scale mismatch: vibration signals need to be sampled in milliseconds, while temperature signals are updated in seconds; (2) Spatial resolution mismatch: The positioning accuracy of vibration sensors reaches the millimeter level, while the resolution of thermal imaging data is only at the centimeter level; (3) Feature correlation mismatch: Current changes reflect electrical loads, and vibration spectra represent mechanical states. Traditional methods cannot establish cross-domain mapping. Existing technologies use a fixed time window interception and channel-by-channel independent processing strategy, resulting in a cross-physical field feature correlation loss rate of more than 40%, an early composite fault missed detection rate of more than 35%, and an unstructured data utilization rate of less than 30%.

[0004] 2. Lack of dynamic adaptability of the early warning mechanism: (1) Performance degradation effect: After 3 years of operation, the vibration amplitude under the same working conditions increased by 40%; (2) Operating condition fluctuation interference: Sudden load change can cause the vibration value to increase by 80% instantaneously; Mainstream technologies rely on static thresholds and fixed statistical models, resulting in false alarm rates exceeding 50% during equipment aging, missed alarm rates exceeding 35% in variable operating conditions, and warning accuracy fluctuating by up to 40% over the entire lifecycle. 3. Insufficient spatial resolution of fault location: (1) Spatial level limitations: faults can only be located at the equipment unit level and cannot be identified at the component level; (2) Weak noise immunity: In industrial sites where the signal-to-noise ratio is less than 10 dB, the positioning error can reach 20% of the equipment size; This resulted in repairs requiring the investigation of four suspected components, with fault diagnosis taking an average of more than two hours and triggering 20% ​​secondary repairs.

[0005] 4. Low efficiency of human-computer interaction collaboration: (1) Fault alarm, equipment status, and location information are scattered on independent interfaces; (2) Operation and maintenance requires manual association of multiple views such as vibration spectrum, temperature cloud map, and 3D model; (3) Decision-making relies on the experience of engineers with an average working experience of more than 8 years; This results in a single fault diagnosis taking more than 30 minutes, the misjudgment rate of emergency conditions exceeding 25%, and the training cycle for new employees exceeding 6 months.

[0006] Therefore, in response to the above problems, the present invention provides an industrial equipment status monitoring and fault warning method and system based on big data, which can integrate the spatiotemporal feature fusion engine, dynamic threshold algorithm, gradient back propagation positioning technology, and three-dimensional warning map to solve the collaborative analysis of multi-source data, realize threshold self-adjustment, improve to component-level accuracy, and achieve diagnosis in seconds. Summary of the Invention

[0007] (1) Technical problems solved In response to the deficiencies of the existing technology, the present invention provides a method and system for industrial equipment status monitoring and fault warning based on big data, which solves the problems raised in the above background technology.

[0008] (2) Technical solution To achieve the above objectives, the present invention provides the following technical solutions: a method and system for industrial equipment status monitoring and fault warning based on big data, the method comprising the following steps: S1. Vibration, temperature, current, and acoustic emission data of industrial equipment are collected through a distributed sensor network, and the original equipment operation data set is generated using a timestamp alignment algorithm. S2. Build a spatiotemporal feature fusion model, combine the long short-term memory network and the graph convolutional network to extract features from the original operation data set of the device, and generate a device spatiotemporal feature matrix; S3. Processing the device spatiotemporal feature matrix based on an improved deep forest algorithm, and generating a device fault feature vector through a dynamic weighted random forest and XGBoost sub-model; S4. Calculate the similarity between the equipment fault feature vector and the standard fault mode in the historical fault database, and select the maximum value as the fault mode matching data; S5. Use a dynamic threshold adjustment algorithm to process real-time data streams and generate dynamic fault warning thresholds based on the historical matching mean, standard deviation, and time decay factor; S6. When the fault pattern matching degree exceeds the dynamic fault warning threshold, a fault warning signal is triggered and the fault source device component is located through gradient back propagation to generate fault location information; S7. Integrate fault warning signals, fault location information, and real-time equipment operating status to construct a three-dimensional fault warning map, which is then visualized and output through the industrial supervision platform.

[0009] Preferably, said S1 comprises the following steps: S11. Collect vibration data of industrial equipment through distributed sensor networks , temperature data ; S12. Use the timestamp alignment algorithm to synchronize multi-source data according to the equipment operation cycle to generate the original operation data set aligned in time and space. .

[0010] Preferably, said S2 comprises the following steps: S21. Build a spatiotemporal feature fusion model that integrates the temporal feature extraction capabilities of the long short-term memory network and the spatial feature extraction capabilities of the graph convolutional network. S22, the original running data set Input the spatiotemporal feature fusion model and output the device spatiotemporal feature matrix: in is the spatiotemporal feature matrix, is the time step, is the fusion feature dimension.

[0011] Preferably, said S3 comprises the following steps: S31. Build an improved deep forest model whose cascade structure uses a dynamic weighting mechanism to fuse random forest and XGBoost sub-models; S32, the device spatiotemporal feature matrix Input the improved deep forest model and output the device fault feature vector

[0012]

[0013] where is the fault feature vector, is an m-dimensional real vector, is the dimension of fault characteristics.

[0014] Preferably, said S4 comprises the following steps: S41. Extract standard failure mode matrix from historical failure database ; S42. Calculate the device fault feature vector With each standard failure mode The similarity of is the standard characteristic vector of the jth type of fault; S43, select the highest similarity value as the fault mode matching data .

[0015] Preferably, S5 comprises the following steps: S51. The dynamic threshold adjustment algorithm is defined as: in is the historical matching mean, is the standard deviation, is the sensitivity coefficient, is the time decay factor is a natural constant; is the dynamic fault warning threshold, Accumulate the running time of the equipment; S52, based on real-time data stream update and ,Generate dynamic fault warning threshold.

[0016] Preferably, the S6 comprises the following steps: S61, when When the fault occurs, a level 3 severity fault warning signal is generated; S62. Locate the fault source using the gradient back propagation algorithm: in is the fault source location index, Smatch is the fault mode matching data, is the partial derivative of the matching degree with respect to the sensor data, To obtain the sensor index that maximizes the partial derivative, For the i-th sensor data source, output fault location information.

[0017] 10. Preferably, said S7 comprises the following steps: S71. Construct a three-dimensional fault warning map in It is a three-dimensional fault warning map. It is a fault warning signal. Fault location information. Real-time operating status data of the equipment; S72. Dynamic rendering of three-dimensional maps is achieved through the industrial supervision platform, and the color gradient mapping function is used to visualize the severity of faults: in is the color mapping function, is the fault mode matching data, Generates red and yellow gradient colors based on the matching degree.

[0018] Preferably, the system comprises: Data acquisition and fusion module: includes distributed sensor array, multi-source data synchronization unit and spatiotemporal feature fusion engine; Intelligent diagnosis module: includes an improved deep forest processor, a fault pattern matching unit, and a dynamic threshold generator; Early warning execution module: includes fault location engine, 3D map builder and industrial platform interface unit.

[0019] Preferably, including: The data acquisition and fusion module is deployed on the industrial edge computing gateway and supports the 5G industrial Internet of Things communication protocol; The intelligent diagnosis module adopts FPGA hardware acceleration architecture, and the fault feature extraction delay is less than 50 milliseconds; The early warning execution module integrates the WebGL visualization engine to achieve real-time rendering of the three-dimensional fault map on the browser side.

[0020] (3) Beneficial effects Compared with the existing technology, the present invention provides a method and system for industrial equipment status monitoring and fault early warning based on big data, which has the following beneficial effects: 1. Deep collaboration of multi-source data: Through the innovative design of the spatiotemporal feature fusion engine, the spatiotemporal mismatch problem of vibration, temperature, current and other signals is solved. The long short-term memory network captures the dynamic characteristics of millisecond-level vibration signals, and the graph convolutional network models the spatial topological relationship of equipment to achieve cross-scale alignment of multi-physical quantity data, breaking through the limitations of physical isolation in traditional methods, improving the ability to identify complex faults, and releasing the diagnostic value of unstructured data such as acoustic emissions.

[0021] 2. Dynamic early warning system innovation: Through the dynamic threshold algorithm, a full life cycle adaptive mechanism is constructed. The time decay factor automatically compensates for the aging effect of equipment, and the sensitivity coefficient intelligently adjusts the tolerance for operating condition fluctuations, forming an early warning model that is continuously optimized with operating time, changing the rigid defects of static thresholds, avoiding misjudgments in the equipment degradation stage, accurately identifying real faults in changing operating conditions, and establishing an accurate intelligent early warning system.

[0022] 3. Breakthrough in component-level positioning: Gradient back-propagation positioning technology is used to achieve a substantial improvement in spatial resolution. By analyzing the partial derivatives of fault matching with respect to sensor data, the physical propagation path of abnormal signals can be traced, improving positioning accuracy from the device level to the component level. Stable performance is maintained in strong noise environments, and microscopic fault sources such as bearing damage and gear tooth breakage are targeted, avoiding the waste of resources in troubleshooting multiple components.

[0023] 4. 3D fusion decision reconstruction: The early warning map entity integrates fault alarms, physical locations, and real-time operation data into a unified spatial view. Color mapping is used to intuitively display fault evolution trends, eliminate the operational burden of switching between multiple interfaces, establish an accurate mapping between physical entities and digital information, reduce the dependence of diagnostic decisions on personal experience, and build a standardized intelligent operation and maintenance framework. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a schematic diagram of the overall system architecture of the present invention.

[0025] Figure 2 Flow chart of the method of the present invention. DETAILED DESCRIPTION

[0026] 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.

[0027] See also Figure 1 The invention provides a method and system for industrial equipment status monitoring and fault early warning based on big data, and the method comprises the following steps: S1. Vibration, temperature, current, and acoustic emission data of industrial equipment are collected through a distributed sensor network, and the original equipment operation data set is generated using a timestamp alignment algorithm. S2. Build a spatiotemporal feature fusion model, combine the long short-term memory network and the graph convolutional network to extract features from the original operation data set of the device, and generate a device spatiotemporal feature matrix; S3. Processing the device spatiotemporal feature matrix based on an improved deep forest algorithm, and generating a device fault feature vector through a dynamic weighted random forest and XGBoost sub-model; S4. Calculate the similarity between the equipment fault feature vector and the standard fault mode in the historical fault database, and select the maximum value as the fault mode matching data; S5. Use a dynamic threshold adjustment algorithm to process real-time data streams and generate dynamic fault warning thresholds based on the historical matching mean, standard deviation, and time decay factor; S6. When the fault pattern matching degree exceeds the dynamic fault warning threshold, a fault warning signal is triggered and the fault source device component is located through gradient back propagation to generate fault location information; S7. Integrate fault warning signals, fault location information, and real-time equipment operating status to construct a three-dimensional fault warning map, which is then visualized and output through the industrial supervision platform.

[0028] S1 includes the following steps: S11. Collect vibration data of industrial equipment through distributed sensor networks , temperature data ; S12. Use the timestamp alignment algorithm to synchronize multi-source data according to the equipment operation cycle to generate the original operation data set aligned in time and space. .

[0029] S2 includes the following steps: S21. Build a spatiotemporal feature fusion model that integrates the temporal feature extraction capabilities of the long short-term memory network and the spatial feature extraction capabilities of the graph convolutional network. S22, the original running data set Input spatiotemporal feature fusion model, output device spatiotemporal feature matrix

[0030] in is the spatiotemporal feature matrix, is the time step, is the fusion feature dimension.

[0031] S3 includes the following steps: S31. Build an improved deep forest model whose cascade structure uses a dynamic weighting mechanism to fuse random forest and XGBoost sub-models; S32, the device spatiotemporal feature matrix Input the improved deep forest model and output the device fault feature vector

[0032] in is the fault feature vector, is an m-dimensional real vector, is the dimension of fault characteristics.

[0033] S4 includes the following steps: S41. Extract standard failure mode matrix from historical failure database ; S42. Calculate the device fault feature vector With each standard failure mode The similarity of is the standard characteristic vector of the jth type of fault; S43, select the highest similarity value as the fault mode matching data .

[0034] S5 includes the following steps: S51. The dynamic threshold adjustment algorithm is defined as: in is the historical matching mean, is the standard deviation, is the sensitivity coefficient, is the time decay factor is a natural constant; is the dynamic fault warning threshold, Accumulate the running time of the equipment; S52, based on real-time data stream update and ,Generate dynamic fault warning threshold.

[0035] S6 includes the following steps: S61, when When the fault occurs, a level 3 severity fault warning signal is generated; S62. Locate the fault source using the gradient back propagation algorithm: in is the fault source location index, Smatch is the fault mode matching data, is the partial derivative of the matching degree with respect to the sensor data, To obtain the sensor index that maximizes the partial derivative, For the i-th sensor data source, output fault location information.

[0036] S7 includes the following steps: S71. Construct a three-dimensional fault warning map in It is a three-dimensional fault warning map. It is a fault warning signal. Fault location information. Real-time operating status data of the equipment; S72. Dynamic rendering of three-dimensional maps is achieved through the industrial supervision platform, and the color gradient mapping function is used to visualize the severity of faults: in is the color mapping function, is the fault mode matching data, Generates red and yellow gradient colors based on the matching degree.

[0037] The system includes: Data acquisition and fusion module: includes distributed sensor array, multi-source data synchronization unit and spatiotemporal feature fusion engine; Intelligent diagnosis module: includes an improved deep forest processor, a fault pattern matching unit, and a dynamic threshold generator; Early warning execution module: includes fault location engine, 3D map builder and industrial platform interface unit.

[0038] include: The data acquisition and fusion module is deployed on the industrial edge computing gateway and supports the 5G industrial Internet of Things communication protocol; The intelligent diagnosis module adopts FPGA hardware acceleration architecture, and the fault feature extraction delay is less than 50 milliseconds; The early warning execution module integrates the WebGL visualization engine to achieve real-time rendering of three-dimensional fault maps on the browser side.

[0039] Example 1: Wind turbine gearbox monitoring: Step 1: Multi-source data collection: High-precision vibration sensors are deployed on the input and output shafts of the gearbox, temperature sensors are installed in the planetary gear train, and the three-phase current of the generator is collected through current transformers. The vibration signal is sampled at a frequency of 10 kHz, the temperature signal is sampled at a frequency of 1 Hz, and the current signal is sampled at a frequency of 10 kHz to generate an original data set containing three-dimensional vibration data, temperature distribution data, and three-phase current data.

[0040] Step 2: Spatiotemporal feature fusion: Construct a spatiotemporal feature fusion model: First, use a long short-term memory network to process the temporal characteristics of the vibration signal. Then, adopt a graph convolutional network to model the spatial topological relationship of the temperature sensor. Finally, a unified spatiotemporal feature matrix is ​​generated through weighted fusion, where the spatial topology matrix is ​​predefined according to the physical structure of the gearbox.

[0041] Step 3: Fault feature extraction: An improved deep forest model was used to process the spatiotemporal feature matrix: the first-level random forest generated 256-dimensional primary features, and the second-level XGBoost model weighted the feature importance, ultimately outputting a 64-dimensional fault feature vector. This example successfully extracted the planetary gear pitting fault features.

[0042] Step 4: Failure pattern matching: The cosine similarity between the fault feature vector and the historical fault library is calculated. In this example, the matching degree with the pitting failure mode reaches 0.91.

[0043] Step 5: Dynamic threshold warning: Update the dynamic threshold based on 1500 hours of device operation. Combining the historical matching mean of 0.85, the standard deviation of 0.07, and the time decay factor, the calculated warning threshold is 0.93. Because the current matching degree of 0.91 is below the threshold, no warning is triggered.

[0044] Example 2: CNC machine tool spindle monitoring system: 1. Hardware architecture: (1) Edge layer equipment includes: A three-axis accelerometer monitors spindle vibration; Infrared thermal imager collects temperature field; The edge computing gateway performs data preprocessing; (2) The cloud platform uses a GPU computing cluster to deploy core modules such as the spatiotemporal fusion engine and the dynamic threshold generator; (3) User terminals include workshop industrial signage, mobile early warning terminals, and VR diagnostic equipment; 2. Operation process: When a spindle tool chipping fault is detected: (1) The gradient back propagation algorithm locates the fault to the tool holder clamping position; (2) The 3D visualization system marks the red warning area in the digital twin model; (3) Maintenance personnel view fault details through VR equipment; Example 3: Mine conveyor belt system verification: Test conditions: The monitoring points are set at the drive roller and the redirecting roller. The fault type is roller bearing failure. The total length of the conveyor belt is 1.2 kilometers, and a large amount of monitoring data is generated every day.

[0045] 2. Implementation effect: (1) Data fusion performance: vibration and temperature signals are accurately aligned; (2) Warning accuracy: The dynamic threshold system automatically adapts to equipment aging and reduces false alarm rates; (3) Positioning accuracy: Gradient backpropagation technology locks the fault position of the drive roller bearing; (4) Operation and maintenance benefits: Fault diagnosis time is shortened to within 4 minutes, and maintenance costs are reduced.

[0046] Example 4: Monitoring of chemical centrifugal pump units: 1. Special configuration: (1) Explosion-proof sensor: Intrinsically safe vibration sensor; Armored thermocouple temperature measuring device; (2) Corrosion-resistant data relay: 316L stainless steel protective shell; Fiber optic transmission replaces cables; 2. Fault detection process: (1) Detection of mechanical seal leakage characteristics: The vibration signal has a high-frequency component of 9.5kHz; The temperature gradient of the pump body increases; (2) Dynamic threshold system identification: The matching degree exceeds the threshold for 3 hours; Triggering the second level warning; 3. Gradient back propagation positioning: Accurately point to the position of the mechanical seal gland; 4. Implementation value: (1) Avoid high-risk media leakage accidents; (2) Reduce losses from unplanned downtime; (3) Maintenance preparation time is shortened by 65%; 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 the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0047] 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 method for industrial equipment status monitoring and fault early warning based on big data, characterized in that: The method comprises the following steps: S1. Vibration, temperature, current, and acoustic emission data of industrial equipment are collected through a distributed sensor network, and the original equipment operation data set is generated using a timestamp alignment algorithm. S2. Build a spatiotemporal feature fusion model, combine the long short-term memory network and the graph convolutional network to extract features from the original operation data set of the device, and generate a device spatiotemporal feature matrix; S3. Processing the device spatiotemporal feature matrix based on an improved deep forest algorithm, and generating a device fault feature vector through a dynamic weighted random forest and XGBoost sub-model; S4. Calculate the similarity between the equipment fault feature vector and the standard fault mode in the historical fault database, and select the maximum value as the fault mode matching data; S5. Use a dynamic threshold adjustment algorithm to process real-time data streams and generate dynamic fault warning thresholds based on the historical matching mean, standard deviation, and time decay factor; S6. When the fault pattern matching degree exceeds the dynamic fault warning threshold, a fault warning signal is triggered and the fault source device component is located through gradient back propagation to generate fault location information; S7. Integrate fault warning signals, fault location information, and real-time equipment operating status to construct a three-dimensional fault warning map, which is then visualized and output through the industrial supervision platform.

2. The method for industrial equipment status monitoring and fault early warning based on big data according to claim 1, characterized in that: Said S1 comprises the following steps: S11. Collect vibration data of industrial equipment through distributed sensor networks , temperature data ; S12. Use the timestamp alignment algorithm to synchronize multi-source data according to the equipment operation cycle to generate the original operation data set aligned in time and space. .

3. The method for industrial equipment status monitoring and fault early warning based on big data according to claim 1 is characterized in that: The S2 comprises the following steps: S21. Build a spatiotemporal feature fusion model that integrates the temporal feature extraction capabilities of the long short-term memory network and the spatial feature extraction capabilities of the graph convolutional network. S22, the original running data set Input the spatiotemporal feature fusion model and output the device spatiotemporal feature matrix: in is the spatiotemporal feature matrix, is the time step, is the fusion feature dimension.

4. The method for industrial equipment status monitoring and fault early warning based on big data according to claim 1, characterized in that: The S3 includes the following steps: S31. Build an improved deep forest model whose cascade structure uses a dynamic weighting mechanism to fuse random forest and XGBoost sub-models; S32, the device spatiotemporal feature matrix Input the improved deep forest model and output the device fault feature vector: in is the fault feature vector, is an m-dimensional real vector, is the dimension of fault characteristics.

5. The method for industrial equipment status monitoring and fault early warning based on big data according to claim 1, characterized in that: The S4 comprises the following steps: S41. Extract standard failure mode matrix from historical failure database ; S42. Calculate the device fault feature vector With each standard failure mode The similarity of is the standard characteristic vector of the jth type of fault; S43, select the highest similarity value as the fault mode matching data .

6. The method for industrial equipment status monitoring and fault early warning based on big data according to claim 1 is characterized in that: Said S5 comprises the following steps: S51. The dynamic threshold adjustment algorithm is defined as: in is the historical matching mean, is the standard deviation, is the sensitivity coefficient, is the time decay factor is a natural constant; is the dynamic fault warning threshold, Accumulate the running time of the equipment; S52, based on real-time data stream update and ,Generate dynamic fault warning threshold.

7. The method for industrial equipment status monitoring and fault early warning based on big data according to claim 1 is characterized in that: The S6 comprises the following steps: S61, when When the fault occurs, a level 3 severity fault warning signal is generated; S62. Locate the fault source using the gradient back propagation algorithm: in is the fault source location index, Smatch is the fault mode matching data, is the partial derivative of the matching degree with respect to the sensor data, To obtain the sensor index that maximizes the partial derivative, For the i-th sensor data source, output fault location information.

8. The method for industrial equipment status monitoring and fault early warning based on big data according to claim 1, characterized in that: The S7 comprises the following steps: S71. Construct a three-dimensional fault warning map in It is a three-dimensional fault warning map. It is a fault warning signal. Fault location information. Real-time operating status data of the equipment; S72. Dynamic rendering of three-dimensional maps is achieved through the industrial supervision platform, and the color gradient mapping function is used to visualize the severity of faults: in is the color mapping function, is the fault mode matching data, Generates red and yellow gradient colors based on the matching degree.

9. A big data-based industrial equipment status monitoring and fault early warning system, used to implement a big data-based industrial equipment status monitoring and fault early warning method according to any one of claims 1 to 8, characterized in that: The system comprises: Data acquisition and fusion module: includes distributed sensor array, multi-source data synchronization unit and spatiotemporal feature fusion engine; Intelligent diagnosis module: includes an improved deep forest processor, a fault pattern matching unit, and a dynamic threshold generator; Early warning execution module: includes fault location engine, 3D map builder and industrial platform interface unit.

10. The industrial equipment status monitoring and fault early warning system based on big data according to claim 1, characterized in that: include: The data acquisition and fusion module is deployed on the industrial edge computing gateway and supports the 5G industrial Internet of Things communication protocol; The intelligent diagnosis module adopts FPGA hardware acceleration architecture, and the fault feature extraction delay is less than 50 milliseconds; The early warning execution module integrates the WebGL visualization engine to achieve real-time rendering of the three-dimensional fault map on the browser side.

Images