Oil-free air compressor bearing temperature alarm method and device

Abstract

The present invention provides a bearing temperature alarm method and device for an oil-free air compressor, the method comprising: collecting temperature rise rate, heat flux density and environmental parameter data under different working conditions through a micro-thermocouple array in a sealing structure to obtain a heat flow characteristic data set; calculating the temperature gradient change between adjacent measuring points based on the data set, analyzing the heat accumulation trend in combination with the geometric characteristics of the sealing structure and the thermal physical parameters of the material, generating a three-dimensional hot spot distribution map and heat accumulation rate data; constructing a thermal path network model of the sealing structure and the bearing system according to the hot spot distribution map and the heat accumulation rate data, determining the location of the monitoring point of the potential thermal anomaly zone of the bearing; installing a temperature monitoring device in the potential thermal anomaly zone of the bearing according to the monitoring point arrangement plan, collecting temperature data and extracting the characteristics of the thermal distribution change, and identifying the abnormal heat flow pattern in the sealing structure. The present invention can accurately capture the local thermal anomaly zone of the bearing caused by the sealing structure, and prevent the hot spot blind area from causing early failure of the bearing.

Description

Technical Field

[0001] The invention relates to the technical field of temperature alarm, and in particular to a bearing temperature alarm method and device for an oil-free air compressor. Background Art

[0002] Oil-free air compressors are widely used in the medical, food, pharmaceutical and other industries because of their oil-free gas production characteristics. As a key component of oil-free air compressors, the temperature state of bearings directly affects the operating safety and service life of the equipment. The existing bearing temperature alarm method for oil-free air compressors mainly adopts the method of installing temperature sensors at fixed positions on the bearing housing or using infrared temperature measuring devices for regular detection. However, due to the complex heat transfer path formed by the unique anti-pollution sealing structure of the oil-free air compressor, heat is easy to form local high-temperature areas in the sealing structure, and these areas are often not covered by traditional fixed-position temperature monitoring points. When heat accumulates near the sealing maze structure, the bearing may already be in a dangerous temperature range and the monitoring system still fails to detect it in time, resulting in the risk of early bearing damage being concealed, seriously affecting the reliability and safety of the equipment. Summary of the invention

[0003] The main purpose of the present invention is to solve the technical problem that the existing oil-free air compressor bearing temperature alarm method cannot accurately capture the local thermal abnormal area of ​​the bearing caused by the anti-pollution sealing structure;

[0004] A first aspect of the present invention provides a bearing temperature alarm method for an oil-free air compressor, the bearing temperature alarm method for an oil-free air compressor comprising:

[0005] The temperature rise rate data, heat flux density data and environmental parameter data under different working conditions are collected through the micro-thermocouple array pre-arranged in the sealing structure of the oil-free air compressor to obtain the heat flux characteristic data set of the sealing structure;

[0006] Calculate the temperature gradient change between adjacent measuring points in the micro-thermocouple array according to the heat flow characteristic data set, and analyze the accumulation trend of heat in the sealing structure in combination with the geometric characteristics of the sealing structure and the material thermophysical parameters to obtain a three-dimensional hot spot distribution map and heat accumulation rate data;

[0007] A thermal path network model of the sealing structure and the bearing system is constructed according to the three-dimensional hot spot distribution map and the heat accumulation rate data, and the positions of the fixed monitoring points and the dynamic monitoring points in the potential thermal anomaly area of ​​the bearing are determined by the thermal path network model to obtain a monitoring point arrangement plan;

[0008] According to the monitoring point arrangement plan, a temperature monitoring device is installed in the potential thermal anomaly area of ​​the bearing to collect bearing temperature data, and the thermal distribution change characteristics are extracted from the bearing temperature data. Based on the thermal distribution change characteristics, the abnormal heat flow pattern in the sealing structure is identified to obtain the bearing thermal anomaly diagnosis result, and a temperature alarm is issued based on the bearing thermal anomaly diagnosis result.

[0009] Optionally, in a first implementation of the first aspect of the present invention, the temperature gradient change between adjacent measuring points in the micro-thermocouple array is calculated according to the heat flow characteristic data set, and the accumulation trend of heat in the sealing structure is analyzed in combination with the geometric characteristics of the sealing structure and the material thermophysical parameters, and the three-dimensional hot spot distribution map and heat accumulation rate data are obtained, including:

[0010] Eliminate noise from the temperature rise rate data by smoothing and filtering the time series data to obtain filtered temperature rise field data;

[0011] Calculating the temperature gradient vectors and the change rates of adjacent measuring points in the micro-thermocouple array according to the temperature rise field data and the heat flux density data, and performing temperature field correction in combination with the environmental parameter data to obtain a preliminary temperature distribution model;

[0012] A heat transfer differential equation is established in combination with the preliminary temperature distribution model and the geometric characteristics of the sealing structure, and the heat transfer differential equation is solved by finite element analysis to obtain a continuous temperature field inside the sealing structure;

[0013] The temperature gradient mutation area and the heat flux density abnormal area are identified from the continuous temperature field and heat flux density data, and the heat accumulation index of each temperature gradient mutation area and heat flux density abnormal area is calculated in combination with the material thermophysical parameters to obtain a three-dimensional hot spot distribution map and heat accumulation rate data.

[0014] Optionally, in a second implementation of the first aspect of the present invention, the temperature gradient vectors and the rate of change of adjacent measuring points in the micro-thermocouple array are calculated according to the temperature rise field data and the heat flux density data, and the temperature field is corrected in combination with the environmental parameter data to obtain a preliminary temperature distribution model, including:

[0015] According to the temperature rise field data, the temperature value of each measuring point is spatially mapped in the three-dimensional space coordinate system of the micro-thermocouple array to obtain an initial temperature spatial distribution diagram;

[0016] Calculate the temperature gradient vector for the temperature values ​​of adjacent measuring points in the initial temperature spatial distribution diagram, and use the central difference method to calculate the temperature change rate of each measuring point along the three directions of x, y, and z to obtain the temperature gradient vector field;

[0017] According to the corresponding relationship between the heat flux density data and the temperature gradient vector field, the equivalent thermal conductivity of each area of ​​the sealing structure is calculated by using the least square method to establish a heat flux-temperature relationship model;

[0018] The heat flow-temperature relationship model is modified using the environmental parameter data to obtain a preliminary temperature distribution model.

[0019] Optionally, in a third implementation of the first aspect of the present invention, the thermal path network model of the sealing structure and the bearing system is constructed according to the three-dimensional hot spot distribution map and the heat accumulation rate data, and the positions of the fixed monitoring points and the dynamic monitoring points in the potential thermal anomaly zone of the bearing are determined by the thermal path network model, and the monitoring point arrangement scheme is obtained, including:

[0020] Performing cluster analysis on the hot spots in the sealing structure according to the three-dimensional hot spot distribution map, merging similar hot spots into hot zones based on the results of the cluster analysis, and calculating the heat flow propagation direction and rate of each hot zone according to the heat accumulation rate data to obtain a hot zone influence map;

[0021] Constructing a thermal path network model in the form of a directed graph according to the thermal zone influence map and the geometric relationship between the sealing structure and the bearing system, wherein the nodes represent the thermal zones, the edges represent the heat transfer paths, and the weights represent the heat transfer coefficients;

[0022] Performing a critical path analysis on the thermal path network model, calculating the heat propagation time and thermal attenuation coefficient from each thermal zone to the key parts of the bearing, and ranking each thermal zone according to the degree of influence on the bearing temperature based on the heat propagation time and thermal attenuation coefficient, to obtain a ranking table of thermal zone importance;

[0023] The positions of fixed monitoring points and dynamic monitoring points in the potential thermal anomaly zone of the bearing are determined according to the hot zone importance ranking table, and a monitoring point arrangement plan is obtained.

[0024] Optionally, in a fourth implementation of the first aspect of the present invention, determining the positions of fixed monitoring points and dynamic monitoring points in the potential thermal anomaly zone of the bearing according to the hot zone importance ranking table to obtain a monitoring point arrangement plan includes:

[0025] The hot zones are graded according to the hot zone importance ranking table, and the coverage radius of each level of hot zone is calculated to obtain a hot zone classification table;

[0026] Performing position stability analysis on the advanced hot zones in the hot zone classification table, calculating the standard deviation of the position offset of each advanced hot zone under different working conditions, and dividing the advanced hot zones into stable hot zones and variable hot zones according to the standard deviation of the position offset;

[0027] According to the location of the stable hot zone, fixed monitoring points are set, and the number of fixed monitoring points is optimized by using a set covering algorithm so that each stable hot zone is covered by at least one fixed monitoring point, while minimizing the total number of fixed monitoring points, thereby obtaining a fixed monitoring point arrangement plan;

[0028] Dynamic monitoring points are set for the variable hot zone according to the moving range and the changing frequency of the variable hot zone to obtain a dynamic monitoring point arrangement plan, and the fixed monitoring point arrangement plan and the dynamic monitoring point arrangement plan are integrated to form a monitoring point arrangement plan.

[0029] Optionally, in a fifth implementation of the first aspect of the present invention, the temperature monitoring device is installed in the potential thermal anomaly area of ​​the bearing according to the monitoring point arrangement scheme to collect bearing temperature data, the heat distribution change characteristics are extracted from the bearing temperature data, and the abnormal heat flow pattern in the sealing structure is identified based on the heat distribution change characteristics, and the bearing thermal anomaly diagnosis result is obtained, which includes:

[0030] According to the monitoring point arrangement plan, a temperature monitoring device is installed in the potential thermal anomaly area of ​​the bearing to collect bearing temperature data;

[0031] Perform multi-scale decomposition on the bearing temperature data by wavelet transform, extract the frequency domain characteristics and time domain characteristics of the bearing temperature data, and obtain the bearing temperature change feature vector;

[0032] A heat flow state matrix of the sealing structure is constructed according to the bearing temperature change characteristic vector, and a similarity calculation is performed with a preset normal heat flow pattern template to obtain a heat flow abnormality index;

[0033] A multivariate analysis is performed on the heat flow anomaly index in combination with the equipment operating condition parameters, the abnormal mode is classified using a fault pattern recognition algorithm based on the sealing structure characteristics, and the bearing thermal damage risk assessment index is calculated to obtain the bearing thermal anomaly diagnosis result, and a temperature alarm is issued based on the bearing thermal anomaly diagnosis result.

[0034] Optionally, in a sixth implementation of the first aspect of the present invention, the heat flow abnormality index is combined with the equipment operating condition parameters for multivariate analysis, the abnormal mode is classified using a fault mode recognition algorithm based on the sealing structure characteristics, and the bearing thermal damage risk assessment index is calculated, and the bearing thermal abnormality diagnosis result is obtained, including:

[0035] The three-dimensional hot spot distribution map and the heat accumulation rate data are combined with the operating parameters under different operating conditions to construct a high-dimensional feature space, and the data points in the feature space are pre-classified according to the structural characteristics of the sealing structure to form a benchmark thermal flow anomaly pattern library related to the operating conditions;

[0036] After combining the heat flow anomaly index with the current operating condition parameters, the index is located in the constructed high-dimensional feature space, and the Mahalanobis distance is calculated with each mode in the reference heat flow anomaly mode library, and the anomaly mode with the closest distance is selected as the anomaly mode classification result;

[0037] Based on the abnormal pattern classification results and the heat flow path network model, the process of heat transfer to the bearing along the heat flow path is simulated, the temperature change trend of each part of the bearing is predicted, and the thermal damage risk assessment index is calculated according to the thermal tolerance characteristics of the bearing to obtain the bearing thermal anomaly diagnosis result, and a temperature alarm is issued based on the bearing thermal anomaly diagnosis result.

[0038] A second aspect of the present invention provides an oil-free air compressor bearing temperature alarm device, the oil-free air compressor bearing temperature alarm device comprising:

[0039] A data acquisition module is used to collect temperature rise rate data, heat flux density data and environmental parameter data under different working conditions through a micro-thermocouple array pre-arranged in the sealing structure of the oil-free air compressor, so as to obtain a heat flux characteristic data set of the sealing structure;

[0040] A hotspot mapping module is used to calculate the temperature gradient change between adjacent measuring points in the micro-thermocouple array according to the heat flow characteristic data set, and analyze the accumulation trend of heat in the sealing structure in combination with the geometric characteristics of the sealing structure and the material thermophysical parameters to obtain a three-dimensional hotspot distribution map and heat accumulation rate data;

[0041] A monitoring arrangement module is used to construct a thermal path network model of the sealing structure and the bearing system according to the three-dimensional hot spot distribution map and the heat accumulation rate data, determine the positions of the fixed monitoring points and the dynamic monitoring points in the potential thermal anomaly area of ​​the bearing through the thermal path network model, and obtain a monitoring point arrangement plan;

[0042] The thermal anomaly diagnosis module is used to install a temperature monitoring device in a potential thermal anomaly area of ​​the bearing according to the monitoring point arrangement plan to collect bearing temperature data, extract thermal distribution change characteristics from the bearing temperature data, and identify abnormal heat flow patterns in the sealing structure based on the thermal distribution change characteristics, obtain bearing thermal anomaly diagnosis results, and issue a temperature alarm based on the bearing thermal anomaly diagnosis results.

[0043] The above-mentioned oil-free air compressor bearing temperature alarm method and device collects temperature rise rate, heat flux density and environmental parameter data under different working conditions through the micro-thermocouple array in the sealing structure to obtain a heat flow characteristic data set; based on the data set, the temperature gradient change between adjacent measuring points is calculated, and the heat accumulation trend is analyzed in combination with the geometric characteristics of the sealing structure and the thermal physical parameters of the material, and a three-dimensional hot spot distribution map and heat accumulation rate data are generated; according to the hot spot distribution map and the heat accumulation rate data, a thermal path network model of the sealing structure and the bearing system is constructed to determine the location of the monitoring point of the potential thermal anomaly area of ​​the bearing; according to the monitoring point layout plan, a temperature monitoring device is installed in the potential thermal anomaly area of ​​the bearing, temperature data is collected and the characteristics of thermal distribution changes are extracted to identify abnormal heat flow patterns in the sealing structure. The present invention can accurately capture the local thermal anomaly area of ​​the bearing caused by the sealing structure, and prevent the hot spot blind area from causing early failure of the bearing.

[0044] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention are realized and obtained by the structures particularly pointed out in the description, claims and drawings.

[0045] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, preferred embodiments are given below and described in detail with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a schematic diagram of a first embodiment of a bearing temperature alarm method for an oil-free air compressor according to an embodiment of the present invention;

[0047] Figure 2 Schematic diagram of an embodiment of an oil-free air compressor bearing temperature alarm device in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are 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 creative work are within the scope of protection of the present invention.

[0049] The terms "including" and "having" and any variations thereof mentioned in the embodiments of the present invention are intended to cover non-exclusive inclusions. For example, a process, method, apparatus, product or device end including a series of steps or units is not limited to the listed steps or units, but may optionally include other steps or units that are not listed, or may optionally include other steps or units that are inherent to these processes, methods, products or device ends.

[0050] To facilitate understanding of this embodiment, firstly, a bearing temperature alarm method for an oil-free air compressor disclosed in an embodiment of the present invention is described in detail. Figure 1 As shown, the method comprises the following steps:

[0051] 101. The temperature rise rate data, heat flux density data and environmental parameter data under different working conditions are collected through a micro thermocouple array pre-arranged in the sealing structure of the oil-free air compressor to obtain a heat flux characteristic data set of the sealing structure;

[0052] In one embodiment of the present invention, the sealing structure of the oil-free air compressor is first analyzed to determine the arrangement position and number of the micro-thermocouple array. As a key component of the oil-free air compressor, the sealing structure is usually composed of a multi-layer maze structure, which is used to isolate external pollutants from entering the air path system. According to the geometric characteristics of the sealing structure, the micro-thermocouple array is arranged at its key position, specifically including: each chamber of the maze seal, the transition area between the bearing seat and the sealing structure, and the area around the outer ring of the bearing. The micro-thermocouple used is a K-type thermocouple with a diameter of 0.1mm and a special high-temperature wear-resistant alloy material to ensure stability and accuracy under high-speed operation and high-temperature environment. When each measuring point is installed, the thermocouple probe is accurately implanted into the predetermined position inside the sealing structure through the microhole, and the probe is flush with the surface of the structure to avoid interference with the airflow. When collecting data, the oil-free air compressor is tested under a variety of typical working conditions, including the startup phase, rated load operation, overload operation, and shutdown phase. For each working condition, the complete temperature change cycle is recorded, and the real-time temperature data of each measuring point of the micro-thermocouple array is collected. The frequency of temperature data acquisition is dynamically adjusted according to the working conditions. It is once per minute during steady-state operation and increased to ten times per second during start-up and stop or load changes to capture transient thermal changes. By calculating the temperature difference between adjacent time points and dividing it by the time interval, the temperature rise rate data of each measuring point is obtained, which reflects the dynamic process of heat accumulation in the sealing structure. At the same time, a heat flux density sensor is installed at the key position of the sealing structure. A thin film heat flux sensor is used with a thickness of 0.2mm, a sensitivity of 2μV / (W / m²), and a measurement range of 0-50kW / m². The heat flux density sensor is arranged at the connection between the bearing support structure and the sealing structure to capture the direction and intensity of heat flow. The heat flux density data is collected synchronously with the temperature data to provide direct evidence of the heat transfer path. In addition, environmental parameter data such as ambient temperature, humidity, and air flow velocity around the air compressor are also recorded. The ambient temperature is measured by a PT100 temperature sensor with an accuracy of ±0.1°C; the relative humidity is measured by a capacitive humidity sensor with an accuracy of ±2%RH; the airflow velocity is measured by a hot wire anemometer with a range of 0.1-25m / s. The environmental parameter data is collected every 10 minutes for environmental impact correction during subsequent temperature field analysis. All collected data are recorded by a high-speed data acquisition system with a sampling rate of 100Hz and 16-bit A / D conversion accuracy, and are preliminarily processed to eliminate obvious noise and outliers. The processed temperature rise rate data, heat flux density data, and environmental parameter data constitute a complete thermal flow characteristic data set of the sealing structure, which contains the thermal behavior characteristics of each part of the sealing structure under different working conditions, and fully reflects the generation, transfer and accumulation of heat in the sealing structure. The obtained thermal flow characteristic data set is stored in the form of a time series matrix.

[0053] 102. Calculate the temperature gradient change between adjacent measuring points in the micro-thermocouple array based on the heat flow characteristic data set, and analyze the accumulation trend of heat in the sealing structure in combination with the geometric characteristics of the sealing structure and the material thermophysical parameters to obtain a three-dimensional hot spot distribution map and heat accumulation rate data;

[0054] In one embodiment of the present invention, the temperature gradient change between adjacent measuring points in the micro-thermocouple array is calculated according to the heat flow characteristic data set, and the accumulation trend of heat in the sealing structure is analyzed in combination with the geometric characteristics of the sealing structure and the material thermophysical parameters to obtain a three-dimensional hot spot distribution map and heat accumulation rate data, including: eliminating noise from the temperature rise rate data by smoothing and filtering the time series data to obtain filtered temperature rise field data; calculating the temperature gradient vector and the change rate of adjacent measuring points in the micro-thermocouple array according to the temperature rise field data and the heat flux density data, and correcting the temperature field in combination with the environmental parameter data to obtain a preliminary temperature distribution model; establishing a heat transfer differential equation in combination with the preliminary temperature distribution model and the geometric characteristics of the sealing structure, solving the heat transfer differential equation by finite element analysis to obtain a continuous temperature field inside the sealing structure; identifying temperature gradient mutation areas and heat flux density abnormal areas for the continuous temperature field and heat flux density data, and calculating the heat accumulation index of each temperature gradient mutation area and heat flux density abnormal area in combination with the material thermophysical parameters to obtain a three-dimensional hot spot distribution map and heat accumulation rate data.

[0055] Specifically, the temperature rise rate data collected from the sealed structure micro-thermocouple array is subjected to time series data smoothing and filtering to eliminate the noise interference generated during the measurement process. The Savitzky-Golay filter is specifically used to process the temperature rise rate data. The filter realizes data smoothing through local polynomial fitting while retaining the peak characteristics of the original data. During the processing, the window width is selected as 11 time points and the polynomial order is 3. Based on experimental verification, this parameter combination effectively filters out high-frequency noise components while retaining the temperature change characteristics. For the start-stop stage data with a high sampling rate, an adaptive window size is used, and the window width is dynamically adjusted with the sampling rate to ensure the consistency of the filtering effect. After filtering, the noise of the temperature rise rate data is significantly reduced, and the signal-to-noise ratio is increased by about 40%, and the filtered temperature rise field data is obtained. The data retains the essential characteristics of temperature change, while eliminating abnormal fluctuations caused by factors such as sensor fluctuations and electromagnetic interference, providing a reliable basis for the subsequent accurate calculation of temperature gradients. The filtered temperature rise field data is stored in the form of a three-dimensional matrix, and each element represents the temperature rise rate value at a specific spatial position and time point.

[0056] Specifically, the temperature gradient vector and change rate of adjacent measuring points in the micro-thermocouple array are calculated based on the filtered temperature rise field data and heat flux density data. First, a three-dimensional spatial coordinate system of the micro-thermocouple array is established, and the physical position of each measuring point is mapped to the coordinate system to form a spatial distribution map of the temperature sampling points. For any two adjacent measuring points, the temperature difference is calculated and divided by the spatial distance to obtain the scalar value of the temperature gradient in this direction. The temperature gradient components in the three orthogonal directions of x, y, and z are calculated by the central difference method, and the temperature gradient vector is formed by combining them. For irregularly arranged measuring points, the irregular grid interpolation algorithm is used to convert the temperature field into a temperature distribution on a regular grid, and then the gradient value is calculated. At the same time, the direction of the temperature gradient is verified in combination with the heat flux density data. According to the basic principle of heat conduction, the direction of the heat flow should be consistent with the direction of the temperature gradient. This principle is used to verify and correct the calculation results. In addition, the temperature field is corrected using environmental parameter data, specifically including: eliminating the influence of background temperature fluctuations according to environmental temperature data; correcting the surface evaporation heat dissipation effect using relative humidity data; and adjusting the convective heat transfer coefficient based on the surrounding air flow velocity. The correction process uses a multiple linear regression model to establish a relationship function between environmental parameters and temperature deviations, and quantitatively eliminate the interference of environmental factors. After the above processing, a preliminary temperature distribution model is obtained.

[0057] Specifically, based on the preliminary temperature distribution model, the heat transfer differential equation is established in combination with the geometric characteristics of the sealing structure, and the finite element analysis method is used to solve it. First, according to the actual geometric shape of the sealing structure of the oil-free air compressor, a detailed three-dimensional CAD model is established, including key parts such as the labyrinth seal chamber, the transition area between the bearing seat and the sealing structure. The CAD model is obtained by carefully measuring the dimensions of the actual sealing structure, with an accuracy of 0.1mm. The CAD model is imported into the finite element analysis software for meshing. The meshing is performed using tetrahedral units with a maximum unit size of 1mm. The mesh is encrypted at the corners of the sealing structure, the cross-sectional change area and other geometric shape mutation positions, and the minimum unit size reaches 0.2mm. The non-steady-state heat conduction equation is established, considering three heat transfer modes: conduction, convection and radiation, and the boundary conditions are set according to the actual operating conditions. The material thermophysical parameters are determined according to the actual materials of the components of the sealing structure, including thermal conductivity, density and specific heat capacity. These parameters change with temperature and are described by piecewise linear functions. The preliminary temperature distribution model is used as the initial condition and calibration basis for the solution. The implicit time integration format is used to solve the unsteady heat conduction equation, and the time step is set to 1 second to ensure the balance between calculation accuracy and efficiency. Through iterative calculation, when the maximum temperature difference between two adjacent iteration results is less than 0.1℃, the iteration is considered to have converged. The solution result is the temperature field at any position inside the sealed structure, forming a continuous temperature distribution function.

[0058] Specifically, the continuous temperature field and heat flux data obtained by the solution are analyzed to identify the temperature gradient mutation area and the heat flux abnormal area. The spatial second-order derivative of the temperature field data, that is, the rate of change of the temperature gradient, is calculated. When the absolute value of the second-order derivative exceeds the preset threshold (the threshold is determined by statistical analysis, generally the 95% quantile of the temperature gradient change rate distribution), it is marked as a temperature gradient mutation area. These areas represent locations where the temperature changes dramatically, and are often key areas for heat accumulation or dissipation. At the same time, the heat flux data is analyzed. When the heat flux value deviates from the normal value range (the normal range is obtained by historical data statistics) by more than 30%, it is marked as a heat flux abnormal area. These abnormal areas are usually related to obstructed heat transfer paths or abnormal heat sources. For the identified temperature gradient mutation areas and heat flux abnormal areas, the heat accumulation index is calculated in combination with the material thermophysical parameters. The calculation of this index takes into account the heat capacity of the material (the product of density and specific heat capacity), the regional volume, and the temperature rise rate, reflecting the degree of heat accumulation in the area. According to the size of the heat accumulation index, each area is graded to generate a three-dimensional hot spot distribution map, which intuitively shows the spatial location and severity of the hot spot in the sealed structure. At the same time, the heat accumulation rate data of each hot spot area is calculated, that is, the rate of change of the heat accumulation index per unit time, reflecting the dynamic trend of the development of the hot spot.

[0059] Furthermore, the temperature gradient vectors and the rate of change of adjacent measuring points in the micro-thermocouple array are calculated according to the temperature rise field data and the heat flux density data, and the temperature field is corrected in combination with the environmental parameter data to obtain a preliminary temperature distribution model, including: spatially mapping the temperature value of each measuring point in the three-dimensional space coordinate system of the micro-thermocouple array according to the temperature rise field data to obtain an initial temperature spatial distribution diagram; calculating the temperature gradient vector for the temperature values ​​of adjacent measuring points in the initial temperature spatial distribution diagram, and using the central difference method to calculate the temperature change rate of each measuring point along the three directions of x, y, and z to obtain a temperature gradient vector field; according to the correspondence between the heat flux density data and the temperature gradient vector field, using the least squares method to calculate the equivalent thermal conductivity of each area of ​​the sealing structure, and establishing a heat flux-temperature relationship model; using the environmental parameter data to correct the heat flux-temperature relationship model to obtain a preliminary temperature distribution model.

[0060] Specifically, the temperature value of each measuring point is spatially mapped in the three-dimensional spatial coordinate system of the micro-thermocouple array according to the filtered temperature rise field data to obtain the initial temperature spatial distribution diagram. In specific implementation, a rectangular coordinate system with the geometric center of the sealing structure as the origin is established, with the x-axis along the axial direction, the y-axis along the radial direction, and the z-axis along the circumferential direction. The physical position of each micro-thermocouple is obtained by precise measurement. The measurement uses a three-coordinate measuring machine with an accuracy of 0.01mm to ensure the accuracy of the position data. The actual physical coordinates of each measuring point are converted into the coordinate system to form a spatial distribution diagram of the measuring point. For each measuring point, the latest collected temperature rise rate data is integrated with time, and the initial temperature value is added to calculate the current temperature value. The temperature value is associated with the spatial coordinates of the measuring point to form a temperature-position mapping relationship. Since the distribution of the micro-thermocouples in the sealing structure is not a regular grid, but is optimized according to the structural characteristics and hot spot prediction, it is necessary to interpolate the temperature data of these discrete points. The three-dimensional spatial interpolation algorithm based on radial basis function is used to interpolate the temperature values ​​of discrete measuring points to the entire sealed structure space to form a continuous initial temperature spatial distribution map. The distribution map is stored in the form of voxels with a voxel size of 0.5mm×0.5mm×0.5mm to ensure sufficient spatial resolution. The selection of the weight function in the interpolation process is based on the spatial distance between the measuring points. The closer the distance, the greater the contribution of the measuring point to the temperature value of the target point.

[0061] Specifically, the temperature gradient vector is calculated for the temperature values ​​of adjacent measuring points in the initial temperature spatial distribution diagram, and the temperature change rate of each measuring point along the three directions of x, y, and z is calculated using the central difference method to obtain the temperature gradient vector field. During the calculation process, for the regularly arranged measuring point area, the central difference formula is directly used to calculate the temperature gradient components in the three orthogonal directions. For example, for the area located at (x 0 ,y 0 ,z 0), the temperature gradient in the x direction is calculated by dividing the temperature difference of the adjacent measuring points on the left and right sides of the point by the spatial distance. For boundary measuring points or irregularly arranged areas, a one-sided difference or a modified difference format is used. For measuring points with completely irregular arrangement, the moving least squares method is first used to fit the temperature distribution function near the measuring point, and then the fitting function is differentiated to obtain the gradient value. This method avoids the error caused by direct difference of irregular grids. The temperature gradient components in three directions of each measuring point are combined to form a temperature gradient vector. At the same time, the rate of change of the temperature gradient over time is calculated, that is, the temperature gradient vectors of two adjacent time points are divided by the time interval. The calculated temperature gradient vector field is stored in the form of a vector field, and each spatial position corresponds to a three-dimensional vector, which indicates the direction and rate of temperature change at the point. The temperature gradient vector field intuitively reflects the flow direction and intensity of heat in the sealing structure. Heat always flows from the high temperature area to the low temperature area, and the flow intensity is proportional to the temperature gradient. The gradient change rate reflects the stability of the heat flow state. Areas with large change rates are often areas where the heat flow state is unstable and need to be paid special attention.

[0062] Specifically, according to the correspondence between the heat flux density data and the temperature gradient vector field, the least squares method is used to calculate the equivalent thermal conductivity of each region of the sealing structure, and a heat flux-temperature relationship model is established. According to Fourier's law of heat conduction, the heat flux density is proportional to the temperature gradient, and the proportional coefficient is the thermal conductivity of the material. The actual sealing structure is composed of a variety of materials, and there may be anisotropy inside the material, so it is necessary to estimate the equivalent thermal conductivity of each region. For the location where the heat flux density sensor is installed, the heat flux density value of the point is directly obtained and paired with the temperature gradient vector calculated at the same location. Collect heat flux density-temperature gradient data pairs under multiple working conditions to form a sample set. For each region, assuming that the thermal conductivity remains constant within a certain range, the least squares method is used to fit the linear relationship between the heat flux density and the temperature gradient, and the slope is the equivalent thermal conductivity of the region. The influence of measurement error is considered in the fitting process, and abnormal data points are eliminated or downgraded. For areas where the heat flux density is not directly measured, interpolation or extrapolation is performed based on the estimated values ​​of the material type and similar areas. In this way, the equivalent thermal conductivity distribution map of each region of the sealing structure is obtained. Based on the equivalent thermal conductivity and temperature gradient vector field, the heat flux distribution of the entire sealing structure is calculated to form a complete heat flux-temperature relationship model.

[0063] Specifically, the heat flux-temperature relationship model is corrected using environmental parameter data to obtain a preliminary temperature distribution model. Environmental parameter data include ambient temperature, relative humidity, and ambient air flow velocity, etc. These parameters affect the heat exchange process between the sealing structure and the outside world. The correction process first constructs environmental influence factors, such as the ambient temperature influence factor is defined as the difference between the current ambient temperature and the reference ambient temperature. For each environmental parameter, its influence function on the heat flux is established, such as the function of ambient temperature affecting the heat flux using a linear model, and the coefficient is obtained by historical data regression. Based on these influence functions, the comprehensive correction value of the environmental parameters on the heat flux is calculated. For surface heat transfer, the influence of ambient humidity on evaporative heat dissipation is considered. The higher the humidity, the weaker the evaporative heat dissipation; the influence of air flow velocity on the convective heat transfer coefficient is considered. The greater the air flow velocity, the stronger the convective heat transfer. For internal heat conduction, the influence of ambient temperature on the thermophysical parameters of the material is mainly considered, such as the relationship between thermal conductivity and temperature. These corrections are applied to the heat flux-temperature relationship model to obtain the heat flux distribution after environmental correction. Based on the corrected heat flux distribution and the current temperature field, the temperature evolution trend in a short time is predicted to obtain a preliminary temperature distribution model.

[0064] 103. Construct a thermal path network model of the sealing structure and the bearing system based on the three-dimensional hot spot distribution map and the heat accumulation rate data, determine the locations of the fixed monitoring points and dynamic monitoring points in the potential thermal anomaly area of ​​the bearing through the thermal path network model, and obtain the monitoring point layout plan;

[0065] In one embodiment of the present invention, the thermal path network model of the sealing structure and the bearing system is constructed according to the three-dimensional hot spot distribution map and the heat accumulation rate data, and the positions of the fixed monitoring points and the dynamic monitoring points of the potential thermal anomaly zone of the bearing are determined by the thermal path network model to obtain the monitoring point arrangement plan, including: clustering analysis of the hot spots in the sealing structure according to the three-dimensional hot spot distribution map, merging similar hot spots into hot zones based on the results of the cluster analysis, and calculating the heat flow propagation direction and rate of each hot zone according to the heat accumulation rate data to obtain a hot zone influence map; constructing a thermal path network model in the form of a directed graph according to the hot zone influence map and the geometric relationship between the sealing structure and the bearing system, wherein the nodes represent the hot zones, the edges represent the heat transfer paths, and the weights represent the heat conduction coefficients; performing a critical path analysis on the thermal path network model, calculating the heat propagation time and the heat attenuation coefficient from each hot zone to the key parts of the bearing, and ranking each hot zone according to the degree of influence on the bearing temperature based on the heat propagation time and the heat attenuation coefficient to obtain a hot zone importance ranking table; determining the positions of the fixed monitoring points and the dynamic monitoring points in the potential thermal anomaly zone of the bearing according to the hot zone importance ranking table to obtain a monitoring point arrangement plan.

[0066] Specifically, cluster analysis is performed on the hot spots in the sealed structure according to the three-dimensional hot spot distribution map. Based on the results of cluster analysis, similar hot spots are merged into hot zones, and the heat flow propagation direction and rate of each hot zone are calculated according to the heat accumulation rate data to obtain the hot zone impact map. In the specific implementation, the density-based clustering algorithm DBSCAN is used to cluster the hot spots in the three-dimensional hot spot distribution map. The algorithm is suitable for clustering of irregular shapes and has strong robustness to noise points. In the algorithm parameter setting, the neighborhood radius ε is set to 5mm, that is, when the spatial distance between two hot spots is less than 5mm, they are considered to be adjacent; the minimum number of points is set to 3, that is, there are at least 3 neighboring hot spots around a hot spot to form a cluster core. In the clustering process, the spatial position and heat accumulation index of the hot spot are considered at the same time, and the two features are given equivalent weights through normalization. For noise points or isolated hot spots, it is determined whether to be a separate hot zone or merged into the nearest hot zone according to the size of their heat accumulation index. After clustering, each cluster is regarded as a hot zone, and the center position, range and average heat accumulation index of the hot zone are calculated. Next, based on the heat accumulation rate data, the changing trend of heat in each hot zone is analyzed, the heat accumulation curve is fitted, and the heat accumulation rate and acceleration are extracted. Combined with the temperature gradient vector of each point in the hot zone, the dominant heat flow direction of the hot zone is calculated, that is, the weighted average of the temperature gradient vectors of all points in the hot zone, and the weight is the heat accumulation index of each point. The heat flow propagation rate is determined by the temperature change rate on the hot zone boundary. The specific method is to select multiple points along the dominant heat flow direction on the hot zone boundary, calculate the time difference required for the temperature of these points to rise to a specific threshold, and divide it by the distance between the points to obtain the heat propagation speed. Finally, a hot zone impact map is generated, which displays the location, range, heat flow direction and propagation rate of each hot zone in a three-dimensional visual form, intuitively reflecting the distribution and propagation characteristics of heat in the sealed structure.

[0067] Specifically, a thermal path network model in the form of a directed graph is constructed based on the thermal zone influence map and the geometric relationship between the sealing structure and the bearing system, in which the node represents the thermal zone, the edge represents the heat transfer path, and the weight represents the thermal conductivity coefficient. The construction process first regards each thermal zone as a node in the network, and the node attributes include position coordinates, thermal zone range, heat accumulation index and heat flow direction. In addition, the key parts of the bearing are also added to the network as special nodes, including the outer ring, inner ring and rolling element contact area of ​​the bearing. Then, based on the heat flow direction and the geometric connectivity of the sealing structure, the connection relationship between the nodes is determined. When there is direct heat transfer between two thermal zones, a directed edge is established between the corresponding nodes, and the direction of the edge follows the direction of heat transfer, that is, from the high temperature zone to the low temperature zone. For thermal zones that are not directly adjacent but connected by structural parts, jump connections are established considering the thermal conductivity of the structure. The weight of the edge represents the thermal conductivity coefficient, and the calculation method is based on the physical distance between the two thermal zones, the cross-sectional area of ​​the connection area and the thermal conductivity of the material. For structural connections with complex shapes, the equivalent thermal conductivity is used, which is calibrated by finite element analysis or experimental data. The larger the weight value, the easier the heat transfer. In addition, the heat transfer path from the hot zone to the key parts of the bearing is considered, and the connection relationship and weight from the hot zone to the bearing node are determined according to the structural characteristics around the bearing. After completing the connection relationship and weight calculation, a complete directed graph structure is formed, which fully describes the heat transfer network in the sealing structure, including the heat source location, transfer path and transfer efficiency.

[0068] Specifically, the critical path analysis is performed on the thermal path network model to calculate the heat propagation time and thermal attenuation coefficient from each hot zone to the key parts of the bearing. Based on the heat propagation time and thermal attenuation coefficient, each hot zone is ranked according to the degree of influence on the bearing temperature, and the importance ranking table of the hot zones is obtained. The critical path analysis adopts an improved shortest path algorithm. Different from the traditional Dijkstra algorithm, this method considers both the path length and the heat conduction efficiency. For each hot zone node in the network, all possible paths to the nodes of the key parts of each bearing are calculated. For each path, the heat propagation time is calculated, that is, the time required for the heat to propagate from the hot zone to the bearing part. The calculation method is to divide the length of each edge on the path by the heat propagation speed of the segment, and then accumulate to obtain the total time. The heat propagation speed is proportional to the weight of the edge (heat conduction coefficient) and inversely proportional to the heat capacity of each node on the path. At the same time, the thermal attenuation coefficient is calculated to indicate the degree of heat loss during the transfer process. The thermal attenuation coefficient takes into account the path length, the thermal conductivity of each segment material and the boundary heat dissipation conditions. The calculation method is to connect the thermal resistance of each segment on the path in series and combine it with the lateral heat dissipation effect. The greater the thermal resistance, the greater the attenuation coefficient, which means that the proportion of heat transferred to the bearing is smaller. Based on the heat propagation time and the thermal attenuation coefficient, the influence index of each hot zone on the bearing temperature is calculated. The calculation formula is influence index = heat accumulation index of hot zone × (1 / heat propagation time) × (1-thermal attenuation coefficient). This index reflects how much heat generated by the hot zone can affect the bearing temperature in a short time. All hot zones are sorted according to the influence index to form a hot zone importance ranking table, which lists the priority of each hot zone's influence on the bearing temperature.

[0069] Specifically, the locations of fixed monitoring points and dynamic monitoring points in the potential thermal anomaly areas of the bearing are determined according to the ranking table of the importance of the thermal areas, and the monitoring point layout plan is obtained. The monitoring point layout process first classifies the thermal areas in the ranking table of the importance of the thermal areas, and divides the thermal areas into three levels: high, medium, and low according to the impact index. Usually, the first 20% of the thermal areas in the ranking table are selected as high-level thermal areas, the middle 50% are medium-level thermal areas, and the remaining 30% are low-level thermal areas. Focus on high-level thermal areas, where thermal anomalies are most likely to cause abnormal bearing temperature. Next, analyze the performance of the position stability of the high-level thermal areas under different working conditions, calculate the standard deviation of the center position of the thermal area between various working conditions, and determine the thermal areas with a standard deviation less than the preset threshold (for example, 2mm) as stable thermal areas, which are suitable for placing fixed monitoring points; thermal areas with a large standard deviation are determined as hot areas with changing positions, and dynamic monitoring points need to be configured. For the layout of fixed monitoring points, a greedy algorithm is used to determine the minimum number of monitoring points so that each high-level stable thermal area is covered by at least one monitoring point. The algorithm first calculates the set of hot zones that can be covered by each candidate monitoring point location, and then iteratively selects the location that can cover the most uncovered hot zones to install monitoring points until all high-level stable hot zones are covered. For high-level hot zones with changing positions, dynamic monitoring points need to be arranged. The dynamic monitoring points are designed with adjustable position sensors, which can adjust the monitoring position according to the changes in the hot zone position. The activity range of the dynamic monitoring points is determined by the position distribution of the hot zone under various working conditions, while the design of the adjustment mechanism takes into account the space limitations and response speed requirements. The layout scheme of fixed monitoring points and dynamic monitoring points is combined to form a complete monitoring point layout scheme, which clearly specifies the type of each monitoring point, the installation location (for fixed points) or the activity range (for dynamic points), and the target hot zone to be monitored. The monitoring point layout scheme is presented in the form of engineering drawings and three-dimensional models, which intuitively shows the spatial layout and coverage of the monitoring system, and provides detailed guidance for actual installation and commissioning.

[0070] Furthermore, the positions of fixed monitoring points and dynamic monitoring points of potential thermal anomaly zones of the bearing are determined according to the hot zone importance ranking table to obtain a monitoring point layout plan, including: grading the hot zones according to the hot zone importance ranking table, and calculating the coverage radius of each level of hot zones to obtain a hot zone classification table; performing position stability analysis on high-level hot zones in the hot zone classification table, calculating the standard deviation of the position offset of each high-level hot zone under different working conditions, and dividing the high-level hot zones into stable hot zones and variable hot zones according to the standard deviation of the position offset; setting fixed monitoring points according to the positions of the stable hot zones, optimizing the number of fixed monitoring points using a set covering algorithm so that each stable hot zone is covered by at least one fixed monitoring point, and minimizing the total number of fixed monitoring points to obtain a fixed monitoring point layout plan; setting dynamic monitoring points for the variable hot zones according to the moving range and change frequency of the variable hot zones to obtain a dynamic monitoring point layout plan, and integrating the fixed monitoring point layout plan with the dynamic monitoring point layout plan to form a monitoring point layout plan.

[0071] Specifically, the hot zones are graded according to the hot zone importance ranking table, and the coverage radius of each hot zone is calculated to obtain the hot zone classification table. The hot zone importance ranking table contains the priority information of each hot zone's impact on the bearing temperature. The hot zones are divided into three levels: high, medium, and low by setting thresholds. In specific implementation, the percentage split method is used to classify the hot zones with the top 20% of the impact index as high-level hot zones. These hot zones have the most significant impact on the bearing temperature and need to be monitored; the hot zones ranked between 20% and 70% are classified as medium-level hot zones. These hot zones have a certain impact on the bearing temperature and need to be properly monitored; the hot zones ranked after 70% are classified as low-level hot zones. These hot zones have little impact on the bearing temperature and do not need to be specifically monitored. For hot zones of different levels, the corresponding coverage radius, that is, the radius of the hot zone's influence range, is calculated. The calculation of the coverage radius is based on the heat flow propagation model, taking into account the heat accumulation index, dominant heat flow direction, and heat conduction characteristics of the hot zone. The specific calculation method is: determine a temperature threshold, usually 10% of the difference between the center temperature of the hot zone and the surrounding reference temperature, and then calculate the radial distance required for the temperature to decay from the center of the hot zone to the threshold. The higher the core temperature of the hot zone, the larger the coverage radius; the better the thermal conductivity of the medium, the larger the coverage radius. For high-level hot zones, due to their high importance, the coverage radius is relatively conservative to ensure that all potential impact areas can be covered; for medium and low-level hot zones, the coverage radius is more accurate to avoid wasting monitoring resources. The level and coverage radius information of each hot zone is recorded in the hot zone classification table, which provides basic data for the layout of monitoring points. The hot zone classification table is stored in a table, containing information such as hot zone ID, level, coverage radius, center coordinates, etc.

[0072] Specifically, the position stability analysis is performed on the advanced hot zones in the hot zone classification table, the standard deviation of the position offset of each advanced hot zone under different working conditions is calculated, and the advanced hot zones are divided into stable hot zones and variable hot zones according to the standard deviation of the position offset. The position stability analysis aims to identify those hot zones with relatively fixed positions and those with variable positions, and provide a basis for the subsequent determination of fixed monitoring points and dynamic monitoring points. In specific implementation, the hot spot distribution data of the oil-free air compressor under typical working conditions are collected, including rated load operation, partial load operation, overload operation, cold start, hot start and other working conditions. For each advanced hot zone, its center position coordinates under various working conditions are tracked, and the average and standard deviation of these positions are calculated. The standard deviation reflects the degree of fluctuation of the hot zone position. The smaller the standard deviation, the more stable the hot zone position. The position standard deviation calculation adopts the three-dimensional Euclidean distance, and the offsets in the three directions of x, y and z are considered at the same time. The position offset standard deviation threshold is set to 3mm. When the hot zone position standard deviation is less than the threshold, it is classified as a stable hot zone; when the standard deviation is greater than or equal to the threshold, it is classified as a variable hot zone. The selection of the threshold takes into account the sensitivity range and installation accuracy of the temperature sensor. In addition to position stability, the changes in the shape and size of the hot spot are also analyzed, and these changes are characterized by the coefficient of variation of the hot spot coverage volume. For hot spots with significant shape changes, even if their center positions are relatively stable, they tend to be classified as variable hot spots, because it is difficult for fixed-position sensors to effectively cover hot spots with variable shapes. Through position stability analysis, high-level hot spots are clearly divided into stable hot spots and variable hot spots.

[0073] Specifically, fixed monitoring points are set according to the location of the stable hot zone, and the set covering algorithm is used to optimize the number of fixed monitoring points so that each stable hot zone is covered by at least one fixed monitoring point, while minimizing the total number of fixed monitoring points to obtain a fixed monitoring point layout plan. The core issue of fixed monitoring point layout is to achieve effective coverage of all stable hot zones under limited constraints. First, determine the set of candidate monitoring point locations, which need to meet installation conditions, such as sufficient installation space, easy maintenance, and no interference with the normal operation of the equipment. Generally, a series of points that meet the above conditions are selected on the surface of the sealing structure or the bearing seat as candidate points. For each candidate point, calculate its monitoring coverage range, that is, the spatial range that the point can effectively monitor. The monitoring coverage range is related to factors such as sensor type, structural characteristics of the installation location, and thermal conductivity of the surrounding materials. Then, a coverage relationship matrix between the candidate point and the stable hot zone is established. The matrix elements indicate whether a candidate point can cover a stable hot zone. The coverage judgment criterion is that the distance between the candidate point and the center of the hot zone is less than the coverage radius of the hot zone. Based on the coverage relationship matrix, the set covering algorithm is applied to solve the optimal monitoring point combination. Specifically, a greedy algorithm is used. Each time, candidate points that can cover the most uncovered hot spots are selected and added to the monitoring point set until all stable hot spots are covered. In order to improve the algorithm effect, a weight factor is introduced. The weight factor considers the importance of the hot spot and the difficulty of installing the candidate point. Points with high importance and easy installation are given priority. The optimization process also considers the redundancy between monitoring points, appropriately increases the coverage redundancy of key hot spots, and improves the reliability of the monitoring system.

[0074] Specifically, dynamic monitoring points are set for the variable hot zone according to the moving range and changing frequency of the variable hot zone, and the dynamic monitoring point layout scheme is obtained, and the fixed monitoring point layout scheme is integrated with the dynamic monitoring point layout scheme to form a monitoring point layout scheme. The key to the design of dynamic monitoring points is to meet the real-time tracking requirements of the variable hot zone. First, the activity characteristics of each variable hot zone are analyzed in detail, and its position distribution under various working conditions is recorded. The moving range of the hot zone, that is, the maximum offset distance of the center position of the hot zone, and the characteristics of the moving path, such as linear movement, random movement within the area, are calculated. At the same time, the frequency and trigger conditions of the hot zone position change, such as load change, start-stop process, etc., are analyzed. Based on these activity characteristics, customized dynamic monitoring points are designed for each variable hot zone. The implementation forms of dynamic monitoring points are diverse, including: multi-point array sensors, array sensors are arranged within the range where the hot zone may appear, and virtual movement is achieved through selective reading; mechanical adjustment sensors, sensor probes are adjusted by mechanical devices such as small stepping motors; magnetic suspension sensors, non-contact adjustment of sensor positions is achieved using magnetic control technology. The design of dynamic monitoring points also needs to consider factors such as space constraints, response speed, and control methods. For example, for small areas, miniaturized array sensors are used; for hot spots with rapid position changes, magnetic levitation sensors with fast response speeds are used. The control system of dynamic monitoring points uses closed-loop control to automatically adjust the sensor position based on real-time temperature field analysis to track the movement of hot spots. After completing the dynamic monitoring point layout plan, integrate it with the fixed monitoring point layout plan to form a complete monitoring point layout plan. During the integration process, pay attention to the collaborative work between monitoring points, such as information complementarity between fixed points and dynamic points, and seamless connection of monitoring areas.

[0075] 104. According to the monitoring point layout plan, a temperature monitoring device is installed in the potential thermal anomaly area of ​​the bearing to collect the bearing temperature data, and the thermal distribution change characteristics are extracted from the bearing temperature data. Based on the thermal distribution change characteristics, the abnormal heat flow pattern in the sealing structure is identified to obtain the bearing thermal anomaly diagnosis result, and a temperature alarm is issued based on the bearing thermal anomaly diagnosis result.

[0076] In one embodiment of the present invention, the temperature monitoring device is installed in the potential thermal anomaly area of ​​the bearing according to the monitoring point arrangement scheme to collect bearing temperature data, the heat distribution change characteristics are extracted from the bearing temperature data, and the abnormal heat flow pattern in the sealing structure is identified based on the heat distribution change characteristics to obtain the bearing thermal anomaly diagnosis result, including: installing the temperature monitoring device in the potential thermal anomaly area of ​​the bearing according to the monitoring point arrangement scheme to collect bearing temperature data; performing multi-scale decomposition of the bearing temperature data by wavelet transform, extracting the frequency domain characteristics and time domain characteristics of the bearing temperature data, and obtaining a bearing temperature change feature vector; constructing a heat flow state matrix of the sealing structure according to the bearing temperature change feature vector, and performing similarity calculation with a pre-set normal heat flow pattern template to obtain a heat flow anomaly index; performing multivariate analysis on the heat flow anomaly index in combination with equipment operating condition parameters, classifying the abnormal patterns using a fault pattern recognition algorithm based on the characteristics of the sealing structure, and calculating the bearing thermal damage risk assessment index to obtain a bearing thermal anomaly diagnosis result, and performing a temperature alarm based on the bearing thermal anomaly diagnosis result.

[0077] Specifically, according to the monitoring point layout plan, a temperature monitoring device is installed in the potential thermal anomaly area of ​​the bearing to collect bearing temperature data. In the specific implementation, according to the fixed monitoring point layout plan, a mounting hole is drilled at the specified position. The hole diameter is determined according to the selected sensor model, usually 3-5mm. The depth of the mounting hole needs to reach a position of 0.5-1mm from the surface of the target monitoring area to ensure that the sensor can accurately capture the hot spot temperature. The fixed monitoring point uses a high-precision thermal resistor temperature sensor PT1000, with a temperature measurement range of -50℃ to 200℃ and an accuracy of ±0.1℃. The sensor is fixed by a special metal sleeve, and thermal grease is filled between the sleeve and the hole wall. For dynamic monitoring points, the corresponding system is installed according to the active characteristics of the variable hot zone. The multi-point array design area uses a micro-thermocouple array, each array contains 4×4 measuring points; the mechanical adjustment design area is installed with a micro-drive mechanism, and the temperature sensor probe is driven by a stepper motor. All sensors are connected to a 16-bit analog-to-digital converter via a shielded cable with a sampling frequency of 10Hz. Temperature data collection is carried out under various typical operating conditions of the oil-free air compressor, including start-stop process, different load levels, etc.

[0078] Specifically, the bearing temperature data is decomposed at multiple scales by wavelet transform, and the frequency domain and time domain features of the bearing temperature data are extracted to obtain the bearing temperature change feature vector. First, the original temperature data is preprocessed, including outlier detection, missing value interpolation and data normalization. The improved Z-score algorithm is used for outlier detection, and the data points that deviate from the mean by more than 3 times the standard deviation are marked as abnormal; the missing values ​​are filled by spline interpolation; and the temperature data is mapped to the [0,1] interval by normalization. Then, the db4 wavelet basis function is selected to perform 5-layer wavelet decomposition on the preprocessed temperature data, and 5 detail coefficient sequences and 1 approximate coefficient sequence are obtained. Statistical features are extracted from these coefficient sequences, including: energy density, which characterizes the intensity of temperature fluctuations in each frequency band; entropy value, which reflects the complexity of temperature signals; kurtosis and skewness, which describe the temperature distribution morphology; maximum and minimum values, which mark extreme temperatures; autocorrelation coefficient, which reflects the periodicity of temperature time series. At the same time, time domain features such as temperature rise rate and temperature stability index are extracted from the original data. These features are combined into feature vectors, and the dimension is reduced by principal component analysis, retaining the principal components that explain 90% of the variance.

[0079] Specifically, the heat flow state matrix of the sealing structure is constructed according to the characteristic vector of the bearing temperature change, and the similarity is calculated with the pre-set normal heat flow mode template to obtain the heat flow abnormality index. The construction process of the heat flow state matrix is ​​as follows: the monitoring points are arranged according to the spatial position to form a topological structure; the characteristic vector of each point is divided into time windows, with a window length of 10 minutes and a sliding step of 1 minute; the temperature gradient and heat flow direction between adjacent points are calculated; and the heat flow change rate of adjacent time windows is compared. The dimension of the three-dimensional heat flow state matrix formed is the number of monitoring points × characteristic dimension × number of time windows. The normal heat flow mode template is a standard model obtained by analyzing the normal operation data of the equipment. The specific method is to collect heat flow state matrix samples during normal operation, identify typical modes under different working conditions through cluster analysis, and calculate the statistical boundary to form an allowable range. The similarity calculation uses the improved Mahalanobis distance metric to calculate the normalized distance between the current state and the best matching template. The heat flow abnormality index is defined as the logarithmic transformation value of the distance, ranging from [0,∞). The larger the value, the higher the degree of abnormality.

[0080] Specifically, a multivariate analysis is performed on the heat flow abnormality index combined with the equipment operating condition parameters, and the abnormal mode is classified by using a fault pattern recognition algorithm based on the characteristics of the sealing structure, and the bearing thermal damage risk assessment index is calculated to obtain the bearing thermal abnormality diagnosis result, and a temperature alarm is performed based on the bearing thermal abnormality diagnosis result. The multivariate analysis first obtains the equipment operating parameters, including speed, load rate, etc., establishes a correlation model between the heat flow abnormality and the operating condition parameters, calculates the expected impact of the operating condition on the heat flow abnormality, and obtains the corrected abnormality index. Based on the index, the fault recognition algorithm is applied for classification. The algorithm includes three modules: feature mapping, pattern matching, and judgment, and identifies fault types such as local overheating of the sealing structure and heat flow path obstruction. The bearing thermal damage risk index is calculated for each fault type. The index considers factors such as abnormal severity and duration, and is divided into five levels: safety, attention, warning, danger, and emergency. Finally, a diagnostic report is generated, which includes abnormality type, location, severity, handling suggestions, etc., to provide a basis for maintenance decisions.

[0081] Furthermore, the heat flow abnormality index is combined with the equipment operating condition parameters for multivariate analysis, the abnormal mode is classified by a fault pattern recognition algorithm based on the sealing structure characteristics, and the bearing thermal damage risk assessment index is calculated to obtain the bearing thermal abnormality diagnosis result, including: using the three-dimensional hot spot distribution map and the heat accumulation rate data in combination with the operating parameters under different working conditions to construct a high-dimensional feature space, and pre-classifying the data points in the feature space according to the structural characteristics of the sealing structure to form a benchmark heat flow abnormality pattern library related to the working condition; after combining the heat flow abnormality index with the current working condition parameters, the index is positioned in the constructed high-dimensional feature space, and the Mahalanobis distance is calculated with each pattern in the benchmark heat flow abnormality pattern library, and the abnormal mode with the closest distance is selected as the abnormal mode classification result; based on the abnormal mode classification result and the heat flow path network model, the process of heat transfer to the bearing along the heat flow path is simulated, the temperature change trend of each part of the bearing is predicted, and the thermal damage risk assessment index is calculated according to the thermal tolerance characteristics of the bearing to obtain the bearing thermal abnormality diagnosis result, and a temperature alarm is issued based on the bearing thermal abnormality diagnosis result.

[0082] Specifically, a high-dimensional feature space is constructed by combining the three-dimensional hot spot distribution map and heat accumulation rate data with the operating parameters under different working conditions, and the data points in the feature space are pre-classified according to the structural characteristics of the sealing structure to form a benchmark thermal flow abnormal pattern library related to the working conditions. During the construction process, the spatial distribution characteristics of the hot spots are extracted from the three-dimensional hot spot distribution map, including the hot spot center position coordinates, hot spot range, heat intensity and other attributes, and the dynamic characteristics of the hot spots, such as heat accumulation rate, heat diffusion direction, etc., are extracted from the heat accumulation rate data. The operating parameters include air compressor speed, load rate, operating time, ambient temperature, etc. These parameters are recorded by the equipment control system and synchronized with the hot spot data time. In the feature selection stage, the principal component analysis method and random forest feature importance evaluation are used to screen out the feature set with high contribution to abnormal pattern recognition, and generally 15-20 feature dimensions are retained. A high-dimensional feature space is constructed through these features, and each data point represents the heat flow state under a specific working condition. The data points in the feature space are pre-classified, and the density clustering algorithm DBSCAN is used, which can effectively process irregular point clusters. During the clustering process, the density parameter ε is dynamically adjusted according to the data distribution to ensure the clustering effect. After the clustering is completed, the formation reasons of each cluster are analyzed, and the abnormality type is determined by combining the geometric characteristics and material characteristics of the sealing structure. Common abnormalities include overheating of the sealing chamber, abnormal heat conduction of the bearing seat, and poor heat dissipation of the bearing. For each abnormality type, a statistical model is established to describe its characteristic distribution to form a benchmark thermal flow abnormality pattern library. The pattern library is stored in the form of structured data, and each pattern contains information such as feature description, working condition association, and severity classification.

[0083] Specifically, the heat flow anomaly index is combined with the current operating parameters to locate in the constructed high-dimensional feature space, and the Mahalanobis distance is calculated with each mode in the benchmark heat flow anomaly pattern library, and the abnormal mode with the closest distance is selected as the abnormal mode classification result. The positioning process first standardizes the current heat flow anomaly index to eliminate the dimension difference, and then combines it with the current operating parameters (speed, load rate, etc.) to form a feature vector. Each component of the feature vector is weighted, and the weight is determined according to the importance of each feature in the abnormal pattern recognition, and the importance is calculated by the information gain method. The combined feature vector is mapped to the pre-constructed high-dimensional feature space to determine its position coordinates in the space. After the positioning is completed, the distance between the point and each mode in the benchmark heat flow anomaly pattern library is calculated, and the Mahalanobis distance measurement method is used. This method takes into account the correlation between features and is more accurate than the Euclidean distance. The covariance matrix in the Mahalanobis distance calculation formula is estimated by normal samples in the feature space to ensure the accuracy of the distance calculation. In order to improve the reliability of classification, multiple distance metrics from the point to each mode are calculated at the same time, such as Euclidean distance, cosine similarity, etc., and the final classification result is selected by weighted voting. For multiple modes that are close to each other, the confidence scores are calculated. When the highest confidence exceeds the threshold (usually set to 80%), it is determined to be the abnormal mode; when the confidence of all modes is lower than the threshold, it is marked as "unknown abnormality" and the manual intervention process is triggered. The classification results contain information such as abnormality type, confidence, similar patterns, etc., which provide a basis for subsequent risk assessment.

[0084] Specifically, based on the abnormal pattern classification results and the heat flow path network model, the process of heat transfer to the bearing along the heat flow path is simulated, the temperature change trend of each part of the bearing is predicted, and the thermal damage risk assessment index is calculated according to the thermal tolerance characteristics of the bearing to obtain the bearing thermal anomaly diagnosis result, and a temperature alarm is performed based on the bearing thermal anomaly diagnosis result. The simulation process adopts the finite element thermal analysis method to establish a three-dimensional heat conduction model including the sealing structure and the bearing system. The location and intensity of the heat source are determined according to the abnormal pattern classification results, and the heat source is set to the actual temperature and heat flow value of the abnormal hot spot. The boundary conditions are set according to the current operating parameters, such as the convection heat transfer coefficient is adjusted according to the air compressor speed, and the ambient temperature is taken from the real-time monitoring value. The material thermophysical parameters such as thermal conductivity, specific heat capacity, density, etc. are set according to the actual material properties, and temperature dependence is considered. The implicit time integration algorithm is used to solve the transient heat conduction equation, with a time step of 1 minute and a simulation time of 2 hours to predict the development trend of thermal anomalies. The simulation results obtain the temperature time series of each key part of the bearing (such as the outer ring, inner ring, and rolling element). Based on the temperature prediction results, the thermal damage risk is evaluated in combination with the thermal tolerance characteristics of the bearing material. The evaluation method considers factors such as absolute temperature, temperature gradient, and number of thermal cycles to construct a comprehensive risk index, which reflects the potential degree of damage to the bearing caused by thermal factors. The risk index is divided into five levels: normal (0-20), slight (20-40), moderate (40-60), severe (60-80), and dangerous (80-100). According to the risk level, corresponding processing suggestions are generated, such as normal continuous monitoring, shortening the inspection cycle, reducing load operation, planned shutdown and maintenance, emergency shutdown processing, etc. Finally, a bearing thermal anomaly diagnosis report is formed. The report includes the type of anomaly, the location of occurrence, the predicted temperature curve, the risk assessment results and the processing suggestions. It is presented intuitively in the form of pictures and texts, providing a scientific basis for equipment maintenance decisions, and alarming based on the content of the bearing thermal anomaly diagnosis report.

[0085] In this embodiment, the temperature rise rate, heat flux density and environmental parameter data under different working conditions are collected through the micro-thermocouple array in the sealing structure to obtain a heat flow characteristic data set; based on the data set, the temperature gradient change between adjacent measuring points is calculated, and the heat accumulation trend is analyzed in combination with the geometric characteristics of the sealing structure and the thermal physical parameters of the material, and a three-dimensional hot spot distribution map and heat accumulation rate data are generated; according to the hot spot distribution map and the heat accumulation rate data, a thermal path network model of the sealing structure and the bearing system is constructed to determine the location of the monitoring point in the potential thermal anomaly area of ​​the bearing; according to the monitoring point layout plan, a temperature monitoring device is installed in the potential thermal anomaly area of ​​the bearing, temperature data is collected and the characteristics of thermal distribution changes are extracted to identify abnormal heat flow patterns in the sealing structure. The present invention can accurately capture the local thermal anomaly area of ​​the bearing caused by the sealing structure, and prevent the hot spot blind area from causing early failure of the bearing.

[0086] The above describes the oil-free air compressor bearing temperature alarm method in the embodiment of the present invention. The following describes the oil-free air compressor bearing temperature alarm device in the embodiment of the present invention. Figure 2 In one embodiment of the present invention, an oil-free air compressor bearing temperature alarm device comprises:

[0087] The data acquisition module 201 is used to collect temperature rise rate data, heat flux density data and environmental parameter data under different working conditions through a micro thermocouple array pre-arranged in the sealing structure of the oil-free air compressor to obtain a heat flux characteristic data set of the sealing structure;

[0088] The hot spot mapping module 202 is used to calculate the temperature gradient change between adjacent measuring points in the micro-thermocouple array according to the heat flow characteristic data set, and analyze the accumulation trend of heat in the sealing structure in combination with the geometric characteristics of the sealing structure and the material thermophysical parameters to obtain a three-dimensional hot spot distribution map and heat accumulation rate data;

[0089] A monitoring arrangement module 203 is used to construct a thermal path network model of the sealing structure and the bearing system according to the three-dimensional hot spot distribution map and the heat accumulation rate data, determine the positions of the fixed monitoring points and the dynamic monitoring points in the potential thermal anomaly area of ​​the bearing through the thermal path network model, and obtain a monitoring point arrangement plan;

[0090] The thermal anomaly diagnosis module 204 is used to install a temperature monitoring device in the potential thermal anomaly area of ​​the bearing according to the monitoring point arrangement plan to collect bearing temperature data, extract the thermal distribution change characteristics from the bearing temperature data, and identify the abnormal heat flow pattern in the sealing structure based on the thermal distribution change characteristics, obtain the bearing thermal anomaly diagnosis result, and issue a temperature alarm based on the bearing thermal anomaly diagnosis result.

[0091] In an embodiment of the present invention, the oil-free air compressor bearing temperature alarm device runs the above oil-free air compressor bearing temperature alarm method, and the oil-free air compressor bearing temperature alarm device collects temperature rise rate, heat flux density and environmental parameter data under different working conditions through the micro-thermocouple array in the sealing structure to obtain a heat flow characteristic data set; based on the data set, the temperature gradient change between adjacent measuring points is calculated, and the heat accumulation trend is analyzed in combination with the geometric characteristics of the sealing structure and the thermal physical parameters of the material, and a three-dimensional hot spot distribution map and heat accumulation rate data are generated; according to the hot spot distribution map and the heat accumulation rate data, a thermal path network model of the sealing structure and the bearing system is constructed to determine the location of the monitoring point of the potential thermal anomaly zone of the bearing; according to the monitoring point layout plan, a temperature monitoring device is installed in the potential thermal anomaly zone of the bearing, temperature data is collected and the characteristics of the thermal distribution change are extracted to identify the abnormal heat flow pattern in the sealing structure. The present invention can accurately capture the local thermal anomaly zone of the bearing caused by the sealing structure, and prevent the hot spot blind area from causing early failure of the bearing.

[0092] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described system, device, or unit can refer to the corresponding process in the aforementioned method embodiment and will not be repeated here.

[0093] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art or the whole or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM), random access memory (RAM), disk or optical disk and other media that can store program code.

[0094] As described above, the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A bearing temperature alarm method for an oil-free air compressor, characterized in that: The oil-free air compressor bearing temperature alarm method comprises: The temperature rise rate data, heat flux density data and environmental parameter data under different working conditions are collected through the micro-thermocouple array pre-arranged in the sealing structure of the oil-free air compressor to obtain the heat flux characteristic data set of the sealing structure; Calculate the temperature gradient change between adjacent measuring points in the micro-thermocouple array according to the heat flow characteristic data set, and analyze the accumulation trend of heat in the sealing structure in combination with the geometric characteristics of the sealing structure and the material thermophysical parameters to obtain a three-dimensional hot spot distribution map and heat accumulation rate data; A thermal path network model of the sealing structure and the bearing system is constructed according to the three-dimensional hot spot distribution map and the heat accumulation rate data, and the positions of the fixed monitoring points and the dynamic monitoring points in the potential thermal anomaly area of ​​the bearing are determined by the thermal path network model to obtain a monitoring point arrangement plan; According to the monitoring point arrangement plan, a temperature monitoring device is installed in the potential thermal anomaly area of ​​the bearing to collect bearing temperature data, and the thermal distribution change characteristics are extracted from the bearing temperature data. Based on the thermal distribution change characteristics, the abnormal heat flow pattern in the sealing structure is identified to obtain the bearing thermal anomaly diagnosis result, and a temperature alarm is issued based on the bearing thermal anomaly diagnosis result.

2. The oil-free air compressor bearing temperature alarm method according to claim 1 is characterized in that: The temperature gradient change between adjacent measuring points in the micro-thermocouple array is calculated according to the heat flow characteristic data set, and the accumulation trend of heat in the sealing structure is analyzed in combination with the geometric characteristics of the sealing structure and the material thermophysical parameters to obtain a three-dimensional hot spot distribution map and heat accumulation rate data including: Eliminate noise from the temperature rise rate data by smoothing and filtering the time series data to obtain filtered temperature rise field data; Calculating the temperature gradient vectors and the change rates of adjacent measuring points in the micro-thermocouple array according to the temperature rise field data and the heat flux density data, and performing temperature field correction in combination with the environmental parameter data to obtain a preliminary temperature distribution model; A heat transfer differential equation is established in combination with the preliminary temperature distribution model and the geometric characteristics of the sealing structure, and the heat transfer differential equation is solved by finite element analysis to obtain a continuous temperature field inside the sealing structure; The temperature gradient mutation area and the heat flux density abnormal area are identified from the continuous temperature field and heat flux density data, and the heat accumulation index of each temperature gradient mutation area and heat flux density abnormal area is calculated in combination with the material thermophysical parameters to obtain a three-dimensional hot spot distribution map and heat accumulation rate data.

3. The oil-free air compressor bearing temperature alarm method according to claim 2 is characterized in that: The step of calculating the temperature gradient vectors and the change rates of adjacent measuring points in the micro-thermocouple array according to the temperature rise field data and the heat flux density data, and correcting the temperature field in combination with the environmental parameter data to obtain a preliminary temperature distribution model includes: According to the temperature rise field data, the temperature value of each measuring point is spatially mapped in the three-dimensional space coordinate system of the micro-thermocouple array to obtain an initial temperature spatial distribution diagram; Calculate the temperature gradient vector for the temperature values ​​of adjacent measuring points in the initial temperature spatial distribution diagram, and use the central difference method to calculate the temperature change rate of each measuring point along the three directions of x, y, and z to obtain the temperature gradient vector field; According to the corresponding relationship between the heat flux density data and the temperature gradient vector field, the equivalent thermal conductivity of each area of ​​the sealing structure is calculated by using the least square method to establish a heat flux-temperature relationship model; The heat flow-temperature relationship model is modified using the environmental parameter data to obtain a preliminary temperature distribution model.

4. The oil-free air compressor bearing temperature alarm method according to claim 1, characterized in that: The heat path network model of the sealing structure and the bearing system is constructed according to the three-dimensional hot spot distribution map and the heat accumulation rate data, and the positions of the fixed monitoring points and the dynamic monitoring points in the potential thermal anomaly area of ​​the bearing are determined by the heat path network model to obtain the monitoring point arrangement scheme, which includes: Performing cluster analysis on the hot spots in the sealing structure according to the three-dimensional hot spot distribution map, merging similar hot spots into hot zones based on the results of the cluster analysis, and calculating the heat flow propagation direction and rate of each hot zone according to the heat accumulation rate data to obtain a hot zone influence map; Constructing a thermal path network model in the form of a directed graph according to the thermal zone influence map and the geometric relationship between the sealing structure and the bearing system, wherein the nodes represent the thermal zones, the edges represent the heat transfer paths, and the weights represent the heat transfer coefficients; Performing a critical path analysis on the thermal path network model, calculating the heat propagation time and thermal attenuation coefficient from each thermal zone to the key parts of the bearing, and ranking each thermal zone according to the degree of influence on the bearing temperature based on the heat propagation time and thermal attenuation coefficient, to obtain a ranking table of thermal zone importance; The positions of fixed monitoring points and dynamic monitoring points in the potential thermal anomaly zone of the bearing are determined according to the hot zone importance ranking table, and a monitoring point arrangement plan is obtained.

5. The oil-free air compressor bearing temperature alarm method according to claim 4, characterized in that: Determining the positions of fixed monitoring points and dynamic monitoring points in the potential thermal anomaly zone of the bearing according to the hot zone importance ranking table to obtain the monitoring point arrangement plan includes: The hot zones are graded according to the hot zone importance ranking table, and the coverage radius of each level of hot zone is calculated to obtain a hot zone classification table; Performing position stability analysis on the advanced hot zones in the hot zone classification table, calculating the standard deviation of the position offset of each advanced hot zone under different working conditions, and dividing the advanced hot zones into stable hot zones and variable hot zones according to the standard deviation of the position offset; According to the location of the stable hot zone, fixed monitoring points are set, and the number of fixed monitoring points is optimized by using a set covering algorithm so that each stable hot zone is covered by at least one fixed monitoring point, while minimizing the total number of fixed monitoring points, thereby obtaining a fixed monitoring point arrangement plan; Dynamic monitoring points are set for the variable hot zone according to the moving range and the changing frequency of the variable hot zone to obtain a dynamic monitoring point arrangement plan, and the fixed monitoring point arrangement plan and the dynamic monitoring point arrangement plan are integrated to form a monitoring point arrangement plan.

6. The oil-free air compressor bearing temperature alarm method according to claim 1, characterized in that: The step of installing a temperature monitoring device in a potential thermal anomaly area of ​​the bearing according to the monitoring point arrangement scheme to collect bearing temperature data, extracting heat distribution change characteristics from the bearing temperature data, and identifying abnormal heat flow patterns in the sealing structure based on the heat distribution change characteristics, and obtaining a bearing thermal anomaly diagnosis result includes: According to the monitoring point arrangement plan, a temperature monitoring device is installed in the potential thermal anomaly area of ​​the bearing to collect bearing temperature data; Perform multi-scale decomposition on the bearing temperature data by wavelet transform, extract the frequency domain characteristics and time domain characteristics of the bearing temperature data, and obtain the bearing temperature change feature vector; A heat flow state matrix of the sealing structure is constructed according to the bearing temperature change characteristic vector, and a similarity calculation is performed with a preset normal heat flow pattern template to obtain a heat flow abnormality index; A multivariate analysis is performed on the heat flow anomaly index in combination with the equipment operating condition parameters, the abnormal mode is classified using a fault pattern recognition algorithm based on the sealing structure characteristics, and the bearing thermal damage risk assessment index is calculated to obtain the bearing thermal anomaly diagnosis result, and a temperature alarm is issued based on the bearing thermal anomaly diagnosis result.

7. The oil-free air compressor bearing temperature alarm method according to claim 6, characterized in that: The heat flow abnormality index is combined with the equipment operating condition parameters for multivariate analysis, the abnormal mode is classified by a fault mode recognition algorithm based on the sealing structure characteristics, and the bearing thermal damage risk assessment index is calculated to obtain the bearing thermal abnormality diagnosis results including: The three-dimensional hot spot distribution map and the heat accumulation rate data are combined with the operating parameters under different operating conditions to construct a high-dimensional feature space, and the data points in the feature space are pre-classified according to the structural characteristics of the sealing structure to form a benchmark thermal flow anomaly pattern library related to the operating conditions; After combining the heat flow anomaly index with the current operating condition parameters, the index is located in the constructed high-dimensional feature space, and the Mahalanobis distance is calculated with each mode in the reference heat flow anomaly mode library, and the anomaly mode with the closest distance is selected as the anomaly mode classification result; Based on the abnormal pattern classification results and the heat flow path network model, the process of heat transfer to the bearing along the heat flow path is simulated, the temperature change trend of each part of the bearing is predicted, and the thermal damage risk assessment index is calculated according to the thermal tolerance characteristics of the bearing to obtain the bearing thermal anomaly diagnosis result, and a temperature alarm is issued based on the bearing thermal anomaly diagnosis result.

8. An oil-free air compressor bearing temperature alarm device, characterized in that: The oil-free air compressor bearing temperature alarm device comprises: A data acquisition module is used to collect temperature rise rate data, heat flux density data and environmental parameter data under different working conditions through a micro-thermocouple array pre-arranged in the sealing structure of the oil-free air compressor, so as to obtain a heat flux characteristic data set of the sealing structure; A hotspot mapping module is used to calculate the temperature gradient change between adjacent measuring points in the micro-thermocouple array according to the heat flow characteristic data set, and analyze the accumulation trend of heat in the sealing structure in combination with the geometric characteristics of the sealing structure and the material thermophysical parameters to obtain a three-dimensional hotspot distribution map and heat accumulation rate data; A monitoring arrangement module is used to construct a thermal path network model of the sealing structure and the bearing system according to the three-dimensional hot spot distribution map and the heat accumulation rate data, determine the positions of the fixed monitoring points and the dynamic monitoring points in the potential thermal anomaly area of ​​the bearing through the thermal path network model, and obtain a monitoring point arrangement plan; The thermal anomaly diagnosis module is used to install a temperature monitoring device in a potential thermal anomaly area of ​​the bearing according to the monitoring point arrangement plan to collect bearing temperature data, extract thermal distribution change characteristics from the bearing temperature data, and identify abnormal heat flow patterns in the sealing structure based on the thermal distribution change characteristics, obtain bearing thermal anomaly diagnosis results, and issue a temperature alarm based on the bearing thermal anomaly diagnosis results.

Images