Continuous furnace roller fault diagnosis and analysis method based on multi-source sensor monitoring

Abstract

The application belongs to the technical field of continuous furnace equipment fault diagnosis, and specifically discloses a continuous furnace roller fault diagnosis and analysis method based on multi-source sensor monitoring, which comprises the following steps: determining the current stage of the continuous furnace according to the axial elongation rate and calling the corresponding subset for cooperative diagnosis; determining the thermal expansion fault by the axial displacement change rate in the starting stage, and subdividing the fault type by the vibration speed and the cooling water temperature difference; determining the mechanical jam fault by the axial displacement deviation in the stable stage, and performing double-dimension cooperative evaluation by the vibration speed and the cooling water temperature difference, and distinguishing the thermal expansion from the plastic deformation by combining the creep slope; determining the shrinkage jam fault by the axial retraction amount in the shutdown stage, and subdividing the fault type by the vibration speed; and if there is a fault, determining the fault source according to the consistency of the fault types of adjacent rollers and the transmission influence factors based on the connection mode, the axial spacing and the heat conduction path. The application realizes the stage self-adaptive diagnosis and accurate fault source tracing of the roller.

Description

Technical Field

[0001] This invention belongs to the field of continuous furnace equipment fault diagnosis technology, and relates to a method for continuous furnace idler roller fault diagnosis and analysis based on multi-source sensor monitoring. Background Technology

[0002] Continuous furnaces use a conveyor mechanism to sequentially pass workpieces through multiple temperature zones for continuous heat treatment. As a core load-bearing component, the idler rollers endure high temperatures and heavy loads for extended periods, and their operating condition directly impacts product quality. During continuous furnace operation, there is a risk of axial thermal expansion obstruction during startup, creep deformation and cooling failure during stabilization, and retraction and jamming during shutdown. Furthermore, adjacent idler rollers are physically connected through the furnace structure, conveyor belt, or workpieces, allowing faults to propagate between them, further complicating fault location. Failure to detect these faults in a timely manner will lead to accelerated idler roller wear, conveyor misalignment, and even idler roller breakage or shutdown.

[0003] To address the aforementioned issues, various solutions for idler roller fault diagnosis have been proposed. For example, Chinese invention patent CN121573390A discloses a method for diagnosing belt conveyor idler roller faults based on vibration and temperature data fusion. This method uses synchronous pulses to drive multiple wireless sensor nodes deployed along the conveyor line to synchronously collect data, generating a multi-source sensor data stream. Based on the physical location information of the nodes, it reconstructs a spatiotemporal vibration data map and a spatiotemporal temperature distribution map, identifies and filters out common-mode vibration interference, and obtains a differential vibration signal set. By performing pattern recognition on the differential signals, it extracts the spatial propagation parameters of abnormal fluctuations, thereby tracing back and locking the physical source of the fault to achieve idler roller location. Finally, it generates collaborative maintenance task instructions based on the fault nature. This method overcomes the shortcomings of traditional single-point monitoring, which is susceptible to global vibration interference and has difficulty in identifying propagation paths, effectively improving the accuracy of fault identification and location.

[0004] However, the aforementioned existing technologies still have the following shortcomings:

[0005] First, this method uses a fixed dual-parameter fusion strategy of vibration and temperature for fault diagnosis, without considering the differences in fault characterization parameters of the idler roller at different operating stages. It fails to dynamically adjust the monitoring parameters and judgment logic according to the actual operating stage of the continuous furnace, which leads to a mismatch between the diagnostic strategy and the real-time operating conditions of the equipment, making it difficult to achieve accurate and effective fault identification at each stage.

[0006] Second, this method achieves fault location through spatiotemporal data map reconstruction and reverse tracing, but does not consider fault source identification when adjacent idlers fail simultaneously, making it difficult to effectively distinguish whether the fault originates from the target idler or is transmitted from adjacent idlers.

[0007] Therefore, there is an urgent need for a roller fault diagnosis scheme that can adaptively adjust the diagnostic strategy according to the continuous furnace operation stage and has the ability to identify the fault source of adjacent rollers. Summary of the Invention

[0008] In view of this, in order to solve the problems mentioned in the background technology, a fault diagnosis and analysis method for continuous furnace idler rollers based on multi-source sensor monitoring is proposed.

[0009] The objective of this invention can be achieved through the following technical solution: This invention provides a method for fault diagnosis and analysis of continuous furnace idler rollers based on multi-source sensor monitoring, including: determining the current stage based on the axial thermal elongation rate of each idler roller in the continuous furnace, and selecting a subset of parameters corresponding to the current stage from the mapping relationship table for collaborative diagnosis.

[0010] During the startup phase, the axial thermal expansion rate is compared with its threshold to determine whether a thermal expansion fault exists. If it does, the fault type is determined based on the vibration velocity and the cooling water temperature difference.

[0011] During the stabilization phase, the mechanical jamming state is determined based on the comparison between the axial displacement and its threshold, combined with the vibration velocity and creep slope. At the same time, the cooling failure state is determined based on the comparison between the cooling water temperature difference and the furnace pressure fluctuation and their thresholds. The fault type is determined by the combined determination of the mechanical jamming state and the cooling failure state.

[0012] During the shutdown phase, the presence of shrinkage jamming fault is determined based on the axial retraction amount. If it exists, the fault type is determined based on the vibration velocity.

[0013] If the current stage diagnosis result is a fault, then obtain the diagnosis results and fault types of adjacent idlers, determine the consistency of fault types, and combine the transmission influence factors based on connection method, axial spacing and heat conduction path to determine the fault source attribute.

[0014] Accordingly, the present invention determines the current stage of the continuous furnace based on the axial thermal elongation rate of each idler roller in the continuous furnace, and constructs a parameter subset corresponding to each operating stage based on historical fault records, so that the diagnostic strategy is dynamically adjusted with the actual working conditions, thereby overcoming the defect of the existing technology that uses fixed monitoring parameters, which leads to the disconnect between the diagnostic strategy and the equipment working conditions.

[0015] Compared with the prior art, the beneficial effects of the present invention are as follows: (1) The present invention determines the current stage by statistical representative values ​​of the axial thermal elongation rate of each idler roller in the continuous furnace, avoiding the interference of abnormal rate caused by uneven local heating or slight jamming of a single idler roller on the determination of the whole machine stage, and constructs a parameter subset corresponding to each operating stage based on historical fault records, so that the diagnosis is dynamically adjusted with the actual working conditions, overcoming the defect of the current fixed monitoring parameters causing the diagnosis strategy to be out of sync with the equipment working conditions.

[0016] (2) This invention establishes a phased diagnostic architecture by determining whether there is a thermal expansion fault by the axial thermal expansion rate during the start-up phase, and further subdividing the thermal expansion fault type according to the vibration speed and cooling water temperature difference. During the stabilization phase, the mechanical jamming state is determined by the axial displacement, and the cooling failure state is determined by the cooling water temperature difference combined with the furnace pressure fluctuation. The fault type is determined according to the combination of the two states. During the shutdown phase, the shrinkage amount is used to determine whether there is a shrinkage jamming fault, and the type is further subdivided according to the vibration speed. This solves the problem of unclear fault type positioning caused by the current multi-parameter parallel fusion.

[0017] (3) This invention combines axial displacement and vibration velocity in the stable phase, and further combines creep slope to determine mechanical jamming when axial displacement is abnormal but vibration velocity is normal. This accurately distinguishes between recoverable thermal expansion deviation and irreversible plastic deformation, avoids misjudging normal thermal expansion as mechanical jamming fault, and improves the accuracy of fault determination.

[0018] (4) This invention obtains the diagnostic results and fault types of each adjacent idler roller, and determines the fault source attribute based on the consistency of fault type and the transmission influence factor when there is a fault in the adjacent idler roller. This achieves accurate differentiation between local faults and transmitted faults. At the same time, the transmission influence factor comprehensively considers three physical dimensions: the connection method between adjacent idler rollers, the axial spacing and the heat conduction path, so that the assessment of fault transmission has a clear physical basis and solves the current problem of difficulty in effectively tracing the source when adjacent idler rollers have faults at the same time. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram showing the connections between the steps of the method of the present invention.

[0021] Figure 2 This is a schematic diagram of the mechanical jamming state determination steps of the present invention.

[0022] Figure 3 This is a schematic diagram showing the connection steps for calculating the impact factor in this invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] Please see Figure 1 As shown, this invention provides a method for fault diagnosis and analysis of continuous furnace idler rollers based on multi-source sensor monitoring. The method includes:

[0025] Unless otherwise specified, all preset thresholds involved in this invention are determined through statistical analysis based on historical normal operating data of the continuous furnace at the corresponding operating stage. Thresholds for the same parameter may differ at different operating stages and should be calculated separately based on historical data for each stage.

[0026] For example, the vibration velocity threshold is obtained as follows:

[0027] The continuous vibration velocity data of the target idler roller is collected within a preset statistical time period under the corresponding operating stage of the continuous furnace, when all process parameters are within the preset normal range. The root mean square value of the collected vibration velocity data is calculated, and the statistical mean of the root mean square value is used as the benchmark value. Three times the standard deviation of the benchmark value is used as the vibration velocity threshold value for that stage. The vibration velocity threshold values ​​for the start-up stage, the stabilization stage, and the shutdown stage are obtained in this way based on the historical normal operation data of their respective stages.

[0028] The cooling water temperature difference threshold, furnace pressure fluctuation threshold, axial displacement threshold, axial thermal elongation rate range, preset deviation threshold, preset frequency threshold, and preset transmission threshold are all obtained in the same way as described above, and will not be repeated here.

[0029] The preset frequency threshold ranges from 60% to 80%, and the specific value is adjusted based on the total number of historical fault records. The base value of the preset transmission threshold is 0.5, which can be dynamically adjusted according to the actual operating conditions of the continuous furnace, with an adjustment range of ±0.2.

[0030] S1, Stage Identification and Dynamic Subset Calling

[0031] The current stage is determined by the axial thermal elongation rate of each idler roller in the continuous furnace, and a subset of parameters corresponding to the current stage is selected from the mapping table for collaborative diagnosis. The continuous furnace exhibits different thermodynamic characteristics at different operating stages. During startup, the furnace temperature continuously rises from room temperature to the process setpoint, causing significant positive axial elongation of the idler rollers. During the stable stage, the furnace temperature field becomes more uniform, and the axial elongation of the idler rollers remains relatively constant. During shutdown, the furnace temperature gradually decreases, and the idler rollers retract axially. Therefore, the axial thermal elongation rate of the idler rollers can reflect the operating stage of the continuous furnace.

[0032] For example, determining the current stage of the continuous furnace includes:

[0033] A time window width is set to obtain the axial elongation values ​​of multiple displacement monitoring points on each idler roller in the continuous furnace at the current moment and at historical moments corresponding to the time window width, tracing back from the current moment. The value of the time window width ranges from 5 to 10 minutes, and its specific value is determined according to the heating rate during the start-up phase of the continuous furnace. The higher the heating rate, the smaller the time window width, and the lower the heating rate, the larger the time window width.

[0034] The median of the axial elongation values ​​of each idler roller at each displacement monitoring point at the current time and at historical time is taken to obtain the representative axial elongation values ​​of each idler roller at the current time and at historical time.

[0035] It should be noted that multiple displacement monitoring points are used for measurement because the idler roller may experience slight skewness during axial thermal elongation due to uneven heating or mechanical installation deviations. Single-point measurement cannot accurately reflect the overall axial thermal elongation state of the idler roller. By coordinating measurements from multiple monitoring points and taking the median, the interference of local outliers on the measurement results is effectively eliminated.

[0036] The difference between the representative axial elongation values ​​of each idler roller at the current time and at historical time is calculated as the axial elongation change, and then divided by the time window width to obtain the axial thermal elongation rate of each idler roller.

[0037] The current stage of the continuous furnace is determined based on the axial thermal elongation rate of each idler roller.

[0038] Optionally, determining the current stage of the continuous furnace based on the axial thermal elongation rate of each idler roller includes:

[0039] Calculate the statistical representative value of the axial thermal elongation rate of each idler roller, compare the statistical representative value with the preset axial thermal elongation rate range, and determine the current stage of the continuous furnace based on the comparison result; the statistical representative value is the average, median or maximum value of the axial thermal elongation rate of each idler roller.

[0040] Preferably, the average axial thermal elongation rate of each idler roller is used as the statistical representative value. Since the idler rollers in the continuous furnace are arranged sequentially along the conveying direction, the temperature fields of each idler roller differ, and the axial thermal elongation rate of a single idler roller may have individual deviations. By taking the average value, abnormal rates caused by uneven heating or slight jamming of local idler rollers can be effectively suppressed, avoiding misjudgments of the overall machine operating condition due to abnormal data from a single idler roller.

[0041] The statistical representative value is compared with the axial thermal elongation rate range: if the statistical representative value is greater than or equal to the upper limit of the range, the continuous furnace is determined to be in the start-up stage; if the statistical representative value is within the range, the continuous furnace is determined to be in the stable stage; if the statistical representative value is less than or equal to the lower limit of the range, the continuous furnace is determined to be in the shutdown stage.

[0042] For example, the construction of the mapping table includes:

[0043] Based on historical fault records of the continuous furnace at various operating stages, the fault period is defined by extending a preset time length both before and after the fault occurrence time as a baseline. The preset time length can be determined according to the fault type and the response characteristics of the monitoring parameters; for example, it can be selected as 5 minutes before the fault occurs to 10 minutes after the fault occurs. The selection of this time span is based on the following: data within the 5 minutes before the fault occurs reflects the normal operating status of the equipment before the fault occurs, providing a data basis for calculating the normal operating baseline value; data within the 10 minutes after the fault occurs can completely cover the entire process from fault occurrence to development, ensuring that the maximum deviation of each monitoring parameter during the fault period is effectively captured. For different types of faults, the preset time length can be appropriately adjusted according to the fault evolution speed; for example, the extended time length can be shortened for sudden faults and extended for gradual faults.

[0044] The sampling data of each monitoring parameter are obtained before the fault occurs and during a period of the same duration as the fault. The average value of each monitoring parameter during that period is calculated and used as the operating benchmark value of each monitoring parameter.

[0045] Calculate the absolute value of the difference between the sampled value of each monitoring parameter at each sampling time and the operating reference value during the fault period, and select the maximum value as the maximum deviation value of each monitoring parameter.

[0046] The monitoring parameters whose maximum deviation value is greater than the preset deviation threshold are marked as the associated parameters of the corresponding operation stage.

[0047] The frequency of each monitoring parameter being marked as an associated parameter in each operation phase is counted separately. Monitoring parameters whose frequency exceeds a preset frequency threshold are included in the parameter subset of that operation phase, and a mapping relationship table is generated.

[0048] S2. Determination and Classification of Thermal Expansion Faults During Startup Phase

[0049] During the startup phase, the axial thermal expansion rate is compared with its threshold to determine whether a thermal expansion fault exists. If it does, the fault type is determined based on the vibration velocity and the cooling water temperature difference.

[0050] This step aims to identify and classify thermal expansion faults during the startup phase. The principle is as follows: During the startup phase of a continuous furnace, the furnace temperature rises continuously, causing the idler rollers to elongate axially. If there is mechanical jamming or limiting interference in the axial elongation direction, the axial thermal expansion rate will be lower than expected. Therefore, by comparing the axial thermal expansion rate with its threshold, it can be determined whether a thermal expansion fault exists. Simultaneously, vibration velocity and cooling water temperature difference are used to further classify the identified thermal expansion faults: when the idler roller's expansion is obstructed, if the obstruction has caused abnormal bearing vibration, the vibration velocity will increase accordingly; if the furnace temperature field deviates from the normal state due to cooling system abnormalities, the cooling water temperature difference will change accordingly. By comprehensively considering the abnormal states of vibration velocity and cooling water temperature difference, the thermal expansion faults can be further subdivided into different types.

[0051] For example, determining the thermal expansion fault and its type includes:

[0052] Calculate the rate of change of axial displacement per unit time as the axial thermal expansion rate, and compare it with the axial thermal expansion rate threshold.

[0053] When the axial thermal expansion rate is less than the threshold, a thermal expansion fault is determined, and the fault type is determined based on the abnormal judgment results of vibration velocity and cooling water temperature difference; the fault types include: expansion obstruction type, vibration-accompanied type, temperature rise-accompanied type, and combined type.

[0054] When the axial thermal expansion rate is greater than or equal to the threshold, it is determined that there is no thermal expansion fault.

[0055] The axial thermal expansion rate threshold is obtained as follows: the theoretical axial thermal expansion rate of the idler roller under unobstructed conditions is calculated by multiplying the linear expansion coefficient of the idler roller material, the axial length of the idler roller, and the heating rate during the start-up phase of the continuous furnace. The axial thermal expansion rate threshold is obtained by subtracting a margin value from the theoretical axial thermal expansion rate. The linear expansion coefficient is an inherent material parameter of the idler roller material, the heating rate is taken from the statistical value of the historical operating data of the continuous furnace, and the margin value is taken as 10% to 20% of the theoretical axial thermal expansion rate, and is not less than 3 times the resolution of the displacement sensor.

[0056] Optionally, the fault type for determining thermal expansion faults includes: when neither the vibration velocity nor the cooling water temperature difference exceeds the corresponding threshold, the fault type is expansion-impeded; when the vibration velocity exceeds the corresponding threshold but the cooling water temperature difference does not exceed the corresponding threshold, the fault type is vibration-accompanied; when the vibration velocity does not exceed the corresponding threshold but the cooling water temperature difference exceeds the corresponding threshold, the fault type is temperature rise-accompanied; and when both the vibration velocity and the cooling water temperature difference exceed the corresponding threshold, the fault type is a combination of both.

[0057] S3. Dual-dimensional collaborative assessment of mechanical jamming and cooling failure in the stable phase.

[0058] During the stabilization phase, the mechanical jamming state is determined based on the comparison between the axial displacement and its threshold, combined with the vibration velocity and creep slope. At the same time, the cooling failure state is determined based on the comparison between the cooling water temperature difference and the furnace pressure fluctuation and their thresholds. The fault type is determined by the combined determination of the mechanical jamming state and the cooling failure state.

[0059] The principle of the collaborative evaluation is as follows: the physical manifestation of mechanical jamming is abnormal axial displacement, accompanied by changes in vibration characteristics. Its determination is based on the comprehensive comparison results of axial displacement, vibration velocity and creep slope; the physical manifestation of cooling failure is an increase in cooling water temperature difference. Its determination needs to exclude abnormal temperature difference caused by furnace pressure fluctuation. That is, when the cooling water temperature difference increases but the furnace pressure fluctuation is small, it can be attributed to the decrease in the heat exchange capacity of the cooling system itself.

[0060] For example, the determination of mechanical jamming state and cooling failure state includes:

[0061] Please see Figure 2 As shown, S3-1, mechanical jamming state determination.

[0062] The axial displacement is compared with an axial displacement threshold. If the axial displacement does not exceed the threshold, and the creep slope is less than or equal to a preset early warning slope threshold, it is determined that there is no mechanical jamming. Otherwise, it is determined that there is an early risk of mechanical jamming and a mechanical jamming condition exists. An early risk of mechanical jamming indicates that although the current axial displacement is not exceeded, the idler roller material has shown signs of accelerated creep, posing a potential risk of developing into mechanical jamming. No fault alarm is triggered temporarily, but a creep anomaly warning signal is output to remind operators to pay attention to the axial displacement and creep trend and arrange timely maintenance.

[0063] When the axial displacement exceeds the threshold, if the vibration velocity exceeds the preset vibration velocity threshold, it is determined that there is a mechanical jamming state.

[0064] If the vibration velocity does not exceed the preset vibration velocity threshold, the creep slope is further compared with the preset fault determination slope threshold. If the creep slope is less than or equal to the fault determination slope threshold, it is determined that there is no mechanical jamming state; otherwise, it is determined that there is a mechanical jamming state.

[0065] The creep slope threshold includes the early warning slope threshold and the fault determination slope threshold, which can be determined in the following ways:

[0066] For a given idler roller material, its creep limit corresponds to a critical creep rate. When the creep slope is less than or equal to this critical value, the axial displacement deviation is mainly caused by recoverable thermal expansion. When the creep slope is greater than this critical value, it indicates that the idler roller material has undergone irreversible plastic deformation, and its axial displacement deviation includes a creep deformation component. This critical value can be obtained through high-temperature tensile creep tests on the idler roller material.

[0067] The critical creep rate is used as the fault determination slope threshold to determine whether there is an abnormality in the mechanical jamming state; the critical creep rate multiplied by a preset proportional coefficient (preferably, the proportional coefficient ranges from 0.5 to 0.8) is used as the early warning slope threshold to determine whether there is an early hidden danger of mechanical jamming. The proportional coefficient is selected based on the following: when the creep slope exceeds 50% to 80% of the critical value, it indicates that the idler roller has entered the early stage of creep acceleration. Although the fault determination standard has not been reached, it is necessary to issue an early warning so that the operator can intervene in advance. For different materials and operating temperatures, this proportional coefficient can be adjusted according to the actual working conditions and safety margin requirements.

[0068] For example, taking a roller made of nickel-based superalloy GH4202 as an example, its... , and The steady-state creep rates under the operating conditions are approximately , and Therefore, the steady-state creep rate value at the corresponding temperature can be selected as a reference benchmark for the creep slope threshold of the roller made of that material. For other materials, those skilled in the art can follow... or The standard involves conducting a high-temperature tensile creep test, calculating the deformation rate during the steady-state creep stage of the specimen, and using the obtained steady-state creep rate as the creep slope threshold for the corresponding material and operating temperature. When the actual operating temperature differs from the example temperature, it can be obtained through interpolation or supplementary experiments. The example values ​​mentioned above are for illustrative purposes only and do not constitute a limitation on the invention.

[0069] S3-2, Cooling Failure Status Determination

[0070] Cooling water temperature difference is a direct parameter reflecting the heat exchange capacity of the cooling system. An increased temperature difference indicates that the cooling system is removing less heat, resulting in insufficient cooling of the idler rollers. However, furnace pressure fluctuations can also affect the cooling system's heat dissipation efficiency. Increased furnace pressure alters the gas flow state within the furnace, reducing the convective heat transfer coefficient between the cooling medium and the idler roller surface, thus causing an increase in the cooling water temperature difference. Therefore, it is necessary to rule out temperature difference anomalies caused by furnace pressure fluctuations to avoid misinterpreting temperature difference anomalies caused by operating condition fluctuations as a failure of the cooling system itself.

[0071] The temperature difference between the cooling water inlet and outlet is compared with a preset cooling water temperature difference threshold. If the cooling water temperature difference does not exceed the threshold, it is determined that there is no cooling failure state.

[0072] When the cooling water temperature difference exceeds the threshold, the furnace pressure fluctuation is further compared with the preset furnace pressure fluctuation threshold. If the furnace pressure fluctuation is less than the preset furnace pressure fluctuation threshold, it is determined that there is a cooling failure state; otherwise, it is determined that there is no cooling failure state.

[0073] For example, determining the fault type includes:

[0074] When there is no mechanical jamming or cooling failure, the fault is determined to be non-faulty. When only mechanical jamming exists, the fault type is determined to be mechanical jamming.

[0075] When only the cooling failure state exists, the fault type is determined to be cooling failure type; when both exist, the fault type is determined to be combined type.

[0076] The core of the dual-dimensional collaborative judgment lies in judging mechanical jamming and cooling failure as two independent dimensions, and then combining the judgment results of the two to accurately distinguish between single faults and compound faults, and avoid the ambiguity of fault type positioning due to misjudgment of a single dimension.

[0077] S4. Fault diagnosis and classification of shrinkage jamming during shutdown phase

[0078] During the shutdown phase, the presence of shrinkage jamming fault is determined based on the axial retraction amount. If it exists, the fault type is determined based on the vibration velocity.

[0079] For example, the determination of the contraction jamming fault and its fault type includes:

[0080] Using the axial displacement value of the target idler roller at the time of entering the shutdown stage as the displacement reference value, the axial displacement value at the current moment is obtained, and the axial retraction amount is calculated. The displacement reference value is established as follows: when the system determines that the continuous furnace has entered the shutdown stage based on the axial thermal elongation rate, the axial displacement value of the target idler roller at that moment is recorded, and it is used as the starting reference for subsequent retraction amount calculation.

[0081] Obtain the theoretical retraction threshold during the shutdown phase. If the axial retraction is less than the theoretical retraction threshold, it is determined that there is a shrinkage jamming fault; otherwise, it is determined that there is no shrinkage jamming fault.

[0082] When a shrinkage jamming fault is detected, the vibration velocity is compared with a vibration velocity threshold, which is determined based on the statistical average vibration velocity of the idler roller under normal operating conditions during the shutdown phase. If the vibration velocity is less than the vibration velocity threshold, the fault type is determined to be shrinkage obstruction type; otherwise, the fault type is determined to be vibration-related type.

[0083] S5. Fault Source Attribute Determination

[0084] If the current stage diagnosis result is a fault, then obtain the diagnosis results and fault types of adjacent idlers, determine the consistency of fault types, and combine the transmission influence factors based on connection method, axial spacing and heat conduction path to determine the fault source attribute.

[0085] In a continuous furnace, the idlers are arranged linearly along the conveying direction, with adjacent idlers physically connected and connected by the furnace structure, conveyor belt, or workpieces, forming a heat transfer path. When one idler malfunctions, its abnormal vibration, thermal expansion, or changes in motion may be transmitted to adjacent idlers through these physical connections, causing similar fault characteristics in the adjacent idlers. Therefore, the source of the fault cannot be determined solely based on the diagnostic results of the target idler; a comprehensive assessment combining diagnostic information from adjacent idlers is necessary.

[0086] For example, the fault type consistency determination includes:

[0087] If the diagnostic results for all adjacent idlers are all fault-free, then the fault types are determined to be inconsistent.

[0088] When adjacent idlers are diagnosed as faulty, if there is an adjacent idler with the same fault type as the target idler, the fault type is determined to be the same.

[0089] If the fault type of all adjacent idlers is different from the fault type of the target idler, then the fault types are determined to be inconsistent.

[0090] Please see Figure 3 As shown, exemplarily, the calculation of the transmission influence factor includes:

[0091] Obtain the connection method, axial spacing, and number of heat conduction paths between the target idler and each adjacent idler. The number of heat conduction paths refers to the total number of solid-contact heat transfer paths and fluid-medium heat transfer paths between the target idler and adjacent idlers.

[0092] The base transfer coefficient is determined based on the connection method between the target idler and its adjacent idler. When the connection is rigid, the base transfer coefficient is assigned a first preset value. When the connection is flexible, the base transfer coefficient is assigned a second preset value, which is less than the first preset value. Rigid connections effectively transfer vibration and thermal stress, resulting in a higher base transfer coefficient; flexible connections provide vibration isolation and buffering, but have weaker fault transmission capabilities, resulting in a lower base transfer coefficient. The first preset value is determined through vibration transmissibility experiments under rigid connection conditions, with a preferred range of 0.6–0.9; the second preset value is determined through vibration transmissibility experiments under flexible connection conditions, with a preferred range of 0.2–0.5. Preferably, the first preset value is 0.8 and the second preset value is 0.3.

[0093] The maximum axial distance between the target idler and each adjacent idler is used as the reference axial distance. The distance between the target idler and the adjacent idler is subtracted from the reference axial distance, and then divided by the reference axial distance to obtain the distance correction factor. The distance correction factor is calculated as follows: In the formula This is the spacing correction factor. The axial distance between the target idler and the adjacent idler. The axial distance is the maximum axial distance between the target idler and its adjacent idlers. The smaller the axial distance, the closer the distance correction coefficient is to 1, and the larger the transmission influence factor. The larger the axial distance, the closer the distance correction coefficient is to 0, and the smaller the transmission influence factor.

[0094] The maximum number of heat transfer paths between the target idler and each adjacent idler is used as the reference number of heat transfer paths. The number of heat transfer paths between the target idler and that adjacent idler is divided by the reference number of heat transfer paths to obtain the path correction factor. The path correction factor is calculated as follows: In the formula This is the path correction factor. This represents the number of heat conduction paths between the target idler and its adjacent idler. This represents the maximum number of heat transfer paths between the target idler and its adjacent idlers. If the number of heat transfer paths is 0, the path correction factor is 0. When the number of heat transfer paths reaches the maximum number of heat transfer paths, the path correction factor is 1.

[0095] Multiplying the base transfer coefficient, the spacing correction coefficient, and the path correction coefficient yields the transfer influence factor between the target idler and its adjacent idler. The value of the transfer influence factor ranges from 0 to 1.

[0096] When the target idler and an adjacent idler are rigidly connected, the axial distance is zero, and the number of heat conduction paths is at its maximum value, the transmission influence factor approaches the first preset value. When the connection method is flexible, the transmission influence factor does not exceed the second preset value.

[0097] For example, the attributes for determining the fault source include:

[0098] When the consistency determination result of the fault type is inconsistent, the fault source attribute is determined to be a local fault.

[0099] When the fault type consistency determination result is consistent, the transmission influence factor of each adjacent idler with the same fault type as the target idler is obtained. If the transmission influence factor is less than the preset transmission threshold, the fault source attribute is determined to be a local fault; otherwise, the fault source attribute is determined to be a transmission fault.

[0100] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, in the form of a computer program product.

[0101] Those skilled in the art will recognize that the modules and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0102] In addition, the functional modules in the various embodiments of this application can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module.

[0103] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

[0104] Finally, the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for fault diagnosis and analysis of continuous furnace idler rollers based on multi-source sensor monitoring, characterized in that: The method includes: The current stage is determined based on the axial thermal elongation rate of each idler roller in the continuous furnace, and a subset of parameters corresponding to the current stage is selected from the mapping table for collaborative diagnosis. During the startup phase, the axial thermal expansion rate is compared with its threshold to determine whether there is a thermal expansion fault. If so, the fault type is determined based on the vibration velocity and the cooling water temperature difference. During the stable phase, the mechanical jamming state is determined based on the comparison between the axial displacement and its threshold, combined with the vibration velocity and creep slope. At the same time, the cooling failure state is determined based on the comparison between the cooling water temperature difference and the furnace pressure fluctuation and their thresholds. The fault type is determined by the combined determination of the mechanical jamming state and the cooling failure state. During the shutdown phase, the presence of shrinkage jamming fault is determined based on the axial retraction amount. If it exists, the fault type is determined based on the vibration velocity. If the current stage diagnosis result is a fault, then obtain the diagnosis results and fault types of adjacent idlers, determine the consistency of fault types, and combine the transmission influence factors based on connection method, axial spacing and heat conduction path to determine the fault source attribute.

2. The method for fault diagnosis and analysis of continuous furnace idler rollers based on multi-source sensor monitoring according to claim 1, characterized in that: Determining the current stage of the continuous furnace includes: Set the time window width and obtain the axial elongation values ​​of multiple displacement monitoring points on each idler roller in the continuous furnace at the current time and at historical times corresponding to the time window width, tracing back from the current time. The median of the axial elongation values ​​of each idler roller at each displacement monitoring point at the current time and at a historical time is taken to obtain the representative axial elongation values ​​of each idler roller at the current time and at a historical time. The difference between the representative axial elongation values ​​of each idler roller at the current time and at the historical time is calculated as the axial elongation change, and divided by the time window width to obtain the axial thermal elongation rate of each idler roller. The current stage of the continuous furnace is determined based on the axial thermal elongation rate of each idler roller.

3. The method for fault diagnosis and analysis of continuous furnace idler rollers based on multi-source sensor monitoring according to claim 1, characterized in that: The construction of the mapping table includes: Based on the historical fault records of the continuous furnace at each operating stage, the fault period is defined by extending the preset time lengths before and after the fault occurrence time as the benchmark. The sampling data of each monitoring parameter are obtained before the fault occurs and during a period of the same duration as the fault. The average value of each monitoring parameter during that period is calculated and used as the operating benchmark value of each monitoring parameter. Calculate the absolute value of the difference between the sampled value of each monitoring parameter at each sampling time during the fault period and the operating reference value, and select the maximum value as the maximum deviation value of each monitoring parameter. The monitoring parameters whose maximum deviation value is greater than the preset deviation threshold are marked as the associated parameters of the corresponding operation stage; The frequency of each monitoring parameter being marked as an associated parameter in each operation phase is counted separately. Monitoring parameters whose frequency exceeds a preset frequency threshold are included in the parameter subset of that operation phase, and a mapping relationship table is generated.

4. The method for fault diagnosis and analysis of continuous furnace idler rollers based on multi-source sensor monitoring according to claim 1, characterized in that: The determination of thermal expansion faults and their types includes: Calculate the rate of change of axial displacement per unit time as the axial thermal expansion rate, and compare it with the axial thermal expansion rate threshold. When the axial thermal expansion rate is less than the threshold, a thermal expansion fault is determined, and the fault type is determined based on the abnormal judgment results of vibration velocity and cooling water temperature difference; the fault types include: expansion obstruction type, vibration-accompanied type, temperature rise-accompanied type, and combined type. When the axial thermal expansion rate is greater than or equal to the threshold, it is determined that there is no thermal expansion fault.

5. The method for fault diagnosis and analysis of continuous furnace idler rollers based on multi-source sensor monitoring according to claim 1, characterized in that: The determination of mechanical jamming state and cooling failure state includes: The axial displacement is compared with the axial displacement threshold. If the creep slope is less than or equal to the preset early warning slope threshold when the axial displacement does not exceed the threshold, it is determined that there is no mechanical jamming state. Otherwise, it is determined that there is an early mechanical jamming risk and that there is a mechanical jamming state. When the axial displacement exceeds the threshold, if the vibration velocity exceeds the preset vibration velocity threshold, it is determined that there is a mechanical jamming state. If the vibration velocity does not exceed the preset vibration velocity threshold, the creep slope is further compared with the preset fault determination slope threshold. If the creep slope is less than or equal to the fault determination slope threshold, it is determined that there is no mechanical jamming state; otherwise, it is determined that there is a mechanical jamming state. The temperature difference between the inlet and outlet of the cooling water is compared with a preset cooling water temperature difference threshold. If the cooling water temperature difference does not exceed the threshold, it is determined that there is no cooling failure state. When the cooling water temperature difference exceeds the threshold, the furnace pressure fluctuation is further compared with the preset furnace pressure fluctuation threshold. If the furnace pressure fluctuation is less than the preset furnace pressure fluctuation threshold, it is determined that there is a cooling failure state; otherwise, it is determined that there is no cooling failure state.

6. The method for fault diagnosis and analysis of continuous furnace idler rollers based on multi-source sensor monitoring according to claim 5, characterized in that: The determination of the fault type includes: When there is no mechanical jamming or cooling failure, the fault is determined to be no fault. When only mechanical jamming exists, the fault type is determined to be mechanical jamming. When only the cooling failure state exists, the fault type is determined to be cooling failure type; when both exist, the fault type is determined to be combined type.

7. The method for fault diagnosis and analysis of continuous furnace idler rollers based on multi-source sensor monitoring according to claim 1, characterized in that: The determination of contraction and jamming faults and their types includes: Using the axial displacement value of the target idler roller when entering the shutdown stage as the displacement reference value, obtain the axial displacement value at the current moment and calculate the axial retraction amount. Obtain the theoretical retraction threshold during the shutdown phase. If the axial retraction is less than the theoretical retraction threshold, it is determined that there is a shrinkage jamming fault; otherwise, it is determined that there is no shrinkage jamming fault. When a contraction jamming fault is detected, the vibration velocity is compared with the vibration velocity threshold. If the vibration velocity is less than the vibration velocity threshold, the fault type is determined to be contraction-impeded; otherwise, the fault type is determined to be vibration-accompanied.

8. The method for fault diagnosis and analysis of continuous furnace idler rollers based on multi-source sensor monitoring according to claim 1, characterized in that: The fault type consistency determination includes: If the diagnostic results for all adjacent idlers are all fault-free, then the fault types are determined to be inconsistent. When there are adjacent idler rollers with fault diagnosis results, if there are adjacent idler rollers with the same fault type as the target idler roller, the fault type is determined to be the same. If the fault type of all adjacent idlers is different from the fault type of the target idler, then the fault types are determined to be inconsistent.

9. The method for fault diagnosis and analysis of continuous furnace idler rollers based on multi-source sensor monitoring according to claim 1, characterized in that: The calculation of the transmission influence factor includes: Obtain the connection method, axial spacing, and number of heat conduction paths between the target idler and each adjacent idler; The basic transfer coefficient is determined based on the connection method between the target idler and the adjacent idler; The maximum axial distance between the target idler and each adjacent idler is used as the reference axial distance. The axial distance between the target idler and the adjacent idler is subtracted from the reference axial distance and then divided by the reference axial distance to obtain the distance correction coefficient. The maximum number of heat conduction paths between the target idler and each of its adjacent idlers is used as the reference number of heat conduction paths. The number of heat conduction paths between the target idler and the adjacent idler is divided by the reference number of heat conduction paths to obtain the path correction coefficient. Multiply the base transfer coefficient, the spacing correction coefficient, and the path correction coefficient to obtain the transfer influence factor between the target idler and the adjacent idler.

10. The method for fault diagnosis and analysis of continuous furnace idler rollers based on multi-source sensor monitoring according to claim 1, characterized in that: The attributes for determining the fault source include: When the consistency determination result of the fault type is inconsistent, the fault source attribute is determined to be a local fault. When the fault type consistency determination result is consistent, the transmission influence factor of each adjacent idler with the same fault type as the target idler is obtained. If the transmission influence factor is less than the preset transmission threshold, the fault source attribute is determined to be a local fault; otherwise, the fault source attribute is determined to be a transmission fault.

Images