A bearing safety assessment method for a hydroelectric generator set based on coupled inversion and safety margin

Abstract

This invention discloses a method for safety assessment of bearing bushes in hydro-generator units based on coupled inversion and safety margin, belonging to the field of bearing bush safety assessment for hydro-generator units. It solves the problem that existing bearing bush monitoring relies on single-parameter threshold alarms, making it difficult to reflect the true load-bearing state of the oil film and unable to quantitatively assess safety margins and early degradation trends. The solution involves: real-time acquisition of multi-source monitoring data, preprocessing and classifying operating conditions; establishing a coupled inversion model of thermal balance and oil film dynamic response to obtain the equivalent thickness change and dynamic stiffness characteristics of the oil film; fusing multi-source data to construct an oil film stability index; calculating the bearing bush safety margin index, and using its rate of change to determine degradation trends and provide graded early warnings; and outputting control suggestions or commands when a preset early warning level is reached. This invention improves the accuracy and foresight of bearing bush anomaly identification and is applicable to the safety assessment of bearing bushes in hydro-generator units.

Description

Technical Field

[0001] This invention relates to the field of safety assessment of bearings in hydro-generator units, and specifically to a method for safety assessment of bearings in hydro-generator units based on coupled inversion and safety margin. Background Technology

[0002] The bearing bushes of a hydro-generator unit are key components that ensure the stable support and safe operation of the unit's rotating parts. Their operating status directly affects the unit's vibration level, rotor stability, and overall safety and reliability.

[0003] In existing technologies, monitoring methods for the operating status of bearings in hydro-generator units mainly rely on online monitoring of parameters such as bearing temperature, lubricating oil temperature, lubricating oil pressure, vibration value, and shaft runout. Abnormal risks are identified through fixed threshold alarms, empirical rule judgments, or single parameter trend analysis. These methods are relatively simple to implement and widely used in engineering, playing a role in detecting significant exceedances and sudden anomalies. However, the operating status of bearings is inherently influenced by multiple factors, including oil film carrying capacity, thermal balance, load distribution, shaft posture, and changes in operating conditions. Relying solely on a single monitoring parameter or simple threshold judgment often only reflects external phenomena and fails to accurately reveal the changes in the internal oil film state of the bearing, making it difficult to effectively identify early risks.

[0004] Furthermore, bearing safety issues are usually not directly caused by a single parameter abrupt change. Instead, they gradually manifest as a complex process under the combined effects of factors such as load fluctuations, speed changes, thrust changes, cooling condition changes, and lubrication condition fluctuations. This process includes reduced oil film thickness, decreased oil film stiffness, increased local temperature rise, and abnormal evolution of the bearing trajectory. Existing methods mostly focus on monitoring and alarming single indicators such as temperature, pressure, or vibration, lacking a comprehensive analysis of the coupling relationship between thermal and dynamic states. An effective technical approach has not yet been developed to indirectly invert the oil film state using multi-source measurable signals and further quantify the bearing safety margin.

[0005] Meanwhile, existing technologies lack quantitative descriptions of the changing trends, risk proximity, and potential critical thinning processes during the degradation process, making it difficult to answer key questions such as how far the current bearing is from the danger boundary and whether the risk is evolving slowly or deteriorating rapidly.

[0006] Therefore, there is an urgent need to propose a new method for assessing the safety of turbine generator bearings. This method should fully utilize multi-source operating data obtained from existing monitoring systems to establish the correlation between the thermal state of the bearing and the dynamic state of the oil film. It should be able to identify key characteristics of the oil film without directly measuring the oil film thickness. Furthermore, it should construct assessment indicators that reflect the safety level and evolution trend of the bearings, so as to achieve quantitative assessment, graded early warning, and control assistance for bearing operation risks, thereby improving the early risk identification capability and safe operation level of turbine generator bearings. Summary of the Invention

[0007] The purpose of this invention is to overcome the shortcomings of the prior art and provide a method for safety assessment of turbine generator bearings based on coupled inversion and safety margin, which improves the early risk identification capability and safe operation level of turbine generator bearings.

[0008] The present invention achieves the above objectives by adopting the following technical solution: The present invention provides a method for safety assessment of turbine generator bearings based on coupled inversion and safety margin, comprising the following steps: S1. Real-time acquisition of multi-source monitoring data during the operation of the hydro-generator unit; S2. Preprocess the collected data and divide it into operating conditions to form an input dataset corresponding to the current unit operating status; S3. Based on multi-source monitoring data, establish a thermo-dynamic coupled inversion model to obtain the equivalent thickness change of the oil film and the dynamic stiffness characteristic of the oil film. S4. The equivalent thickness change of the oil film and the dynamic stiffness characteristic of the oil film are fused with multi-source monitoring data to construct an oil film stability index. S5. Calculate the bearing safety margin index based on the historical healthy operating range or the preset safety boundary; S6. Combining the bearing safety margin index and its rate of change, perform safety margin evolution judgment and degradation trend identification; S7. Graded early warning determination; S8. When the warning level reaches the preset conditions, output the corresponding operation control suggestions or control commands to perform cooling flow regulation, load change rate limitation or load reduction operation.

[0009] Furthermore, the multi-source monitoring data includes at least one or more of the following: unit operating parameters, load-related parameters, bearing thermal state parameters, lubrication system parameters, and rotor shaft attitude and dynamic response parameters; Load-related parameters, including axial load, thrust load, equivalent thrust, or equivalent load parameters that reflect changes in the rotor's stress state; The thermal parameters of the bearing include one or more of the following: multi-point temperature distribution of the bearing, oil inlet temperature, oil return temperature, and cooling medium temperature; Lubrication system parameters include one or more of the following: lubricating oil supply pressure, return oil pressure, oil pressure fluctuation value, oil temperature, oil level, and cooling flow rate; Rotor shaft attitude and dynamic response parameters, including one or more of the following: shaft trajectory, shaft swing, eccentricity, vibration displacement, vibration velocity, and vibration acceleration.

[0010] Furthermore, the preprocessing procedure for the collected data is as follows: Time alignment aligns data from different sources and with different sampling frequencies according to a unified time reference; Data cleaning removes out-of-range values, communication interruption values, obviously invalid values, and abnormal jump values; Missing data repair uses interpolation, neighborhood completion, or sliding estimation to repair short-term missing data. Denoising and smoothing: Smooth high-frequency disturbances in temperature, pressure, and vibration-related signals, while retaining effective trend information reflecting state changes; Feature window construction: Construct sliding windows, fixed windows, or event-triggered windows according to preset analysis cycles.

[0011] Furthermore, the parameters used for classifying operating conditions include the rate of change of rotational speed, the rate of change of active power, the rate of change of axial load, and the rate of change of guide vane opening. The classification of operating conditions includes at least one or more of the following: steady-state operation, load increase, load decrease, start-up, shutdown, and transition. For different operating conditions, separate input datasets and corresponding inversion parameter benchmarks are established to reduce the interference of operating condition changes on the oil film state inversion results.

[0012] Furthermore, in step S3, establishing a thermo-dynamic coupled inversion model based on multi-source monitoring data specifically includes: The average temperature and temperature difference of the bearing bush are calculated based on the multi-point temperature distribution of the bearing bush. The temperature rise and average temperature rise rate of the lubricating oil are calculated based on the inlet and outlet temperatures of the lubricating oil. At the same time, the temperature dispersion index is introduced. A thermal state characterization quantity for bearings is constructed based on bearing average temperature, bearing temperature difference, temperature dispersion index, lubricating oil temperature rise and average temperature rise rate. Dynamic state characterization parameters are constructed based on eccentricity, trajectory dispersion, oil pressure fluctuation coefficient, and vibration characteristic values. The specific process for determining the equivalent thickness change of the oil film and the dynamic stiffness characteristic of the oil film is as follows: The current equivalent thickness change of the oil film is inverted by coupling the thermal state characterization quantity and the dynamic state characterization quantity of the bearing bush. The characteristic quantity of oil film dynamic stiffness is constructed based on eccentricity, oil pressure fluctuation coefficient and vibration characteristic value.

[0013] Furthermore, based on the current equivalent oil film thickness and the benchmark value of equivalent oil film thickness under healthy reference conditions, a sub-index of oil film thickness is calculated. Based on the current dynamic stiffness characteristics of the oil film and the benchmark value of the dynamic stiffness of the oil film under the healthy reference working condition, the sub-index of dynamic stiffness of the oil film is calculated. Based on the bearing temperature difference, temperature dispersion index and average temperature rise rate, the temperature field stability sub-index is calculated. The oil pressure stability sub-index is calculated based on the oil pressure fluctuation coefficient; Based on the eccentricity, trajectory dispersion, and vibration characteristic values, the trajectory vibration stability sub-index is calculated. The oil film stability index is obtained by weighting and integrating the various sub-indicators. The weight can be one of the following: fixed weight, working condition-specific weight, or weight that is dynamically adjusted according to the operating status.

[0014] Furthermore, in step S5, the bearing safety margin index is obtained by comparing the current oil film stability index with the historical healthy operating range, the standard reference range under the same working conditions, or the preset critical boundary. The bearing safety margin index is used to represent the distance between the current bearing operating state and the safety boundary.

[0015] Furthermore, in step S6, the specific process of determining the safety margin evolution and identifying the degradation trend by combining the bearing safety margin index and its rate of change is as follows: The safety margin index of the bearing bush is calculated by sliding calculation over a continuous period to obtain the rate of change of the safety margin, and then a judgment is made: When the safety margin index continues to decline and its rate of change exceeds the preset threshold, the bearing oil film is determined to be in a degraded state. When the safety margin index is lower than the preset safety threshold and the rate of change continues to worsen, it is determined that there is a potential critical thinning risk. Furthermore, in step S7, the tiered early warning determination specifically includes: Risk grading and early warning systems include at least four levels: normal, attention, warning, and alarm. The normal level indicates that the safety margin is within a healthy range; The level of concern indicates a shift in the safety margin compared to the historical healthy range. A warning level indicates a significant decrease in safety margin and a continued trend of degradation; An alarm level indicates that the safety margin is below the critical boundary or that there is a critical risk of thinning.

[0016] Furthermore, When abnormal bearing temperature rise becomes the dominant factor, a cooling flow adjustment suggestion or a cooling system adjustment command is triggered. When a decrease in oil film stability is related to excessively rapid load changes, a load change rate limiting suggestion or limiting command is triggered. When the safety margin is lower than the preset threshold, a load reduction operation suggestion or load reduction control command is triggered.

[0017] The beneficial effects of this invention are as follows: Compared with existing technologies, this invention fully utilizes multi-source data such as rotational speed, load, bearing temperature, lubricating oil parameters, shaft trajectory, and vibration that can be obtained from existing monitoring systems of hydro-generator units. It constructs a coupled inversion model combining thermal and dynamic states, achieving indirect identification of the equivalent thickness change and dynamic stiffness characteristics of the oil film without the need for directly deploying dedicated oil film thickness sensors. Furthermore, it quantitatively assesses the bearing operating status, risk proximity, and degradation evolution trend through oil film stability index, safety margin index, and their rate of change. This method overcomes the problems of traditional bearing safety monitoring, which mainly relies on single-parameter threshold alarms, struggles to reflect the true load-bearing state of the oil film, and fails to identify early risks. It not only improves the accuracy, foresight, and interpretability of bearing anomaly identification but also provides direct evidence for control measures such as cooling regulation, load change rate limiting, and load reduction operation. Attached Figure Description

[0018] Figure 1 This is a flowchart of a method for assessing the bearing safety of a hydro-generator unit based on coupled inversion and safety margin, provided by an embodiment of the present invention. Figure 2 This is a schematic diagram of the characterization quantities reflecting the state of the bearing oil film provided in the embodiments of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings.

[0020] This invention provides a method for assessing the bearing safety of hydro-generator units based on coupled inversion and safety margin, such as... Figure 1 As shown, it specifically includes: Step S1: Multi-source monitoring data acquisition; Real-time acquisition of multi-source monitoring data during the operation of the hydro-generator unit, wherein the multi-source monitoring data includes at least one or more of the following: Unit operating parameters include one or more of the following: speed, active power, reactive power, guide vane opening, head, and flow rate; Load-related parameters, including axial load, thrust load, equivalent thrust, or equivalent load parameters that can reflect changes in the rotor's stress state; The thermal parameters of the bearing include one or more of the following: multi-point temperature distribution of the thrust bearing or guide bearing, oil inlet temperature, oil return temperature, and cooling medium temperature; Lubrication system parameters include one or more of the following: lubricating oil supply pressure, return oil pressure, oil pressure fluctuation value, oil temperature, oil level, and cooling flow rate; Rotor shaft attitude and dynamic response parameters, including one or more of the following: shaft trajectory, shaft swing, eccentricity, vibration displacement, vibration velocity, and vibration acceleration.

[0021] In specific implementation, the multi-point temperature of the bearing bush is preferably arranged at different positions of the same bearing bush or at corresponding positions of different bearing bushes to reflect the temperature field distribution characteristics of the bearing bush; the shaft center trajectory is preferably obtained through an orthogonal displacement sensor to characterize the motion trajectory of the rotor in the bearing; the lubricating oil pressure can be characterized by its dynamic change characteristics in the form of average pressure value, fluctuation amplitude, fluctuation standard deviation, short-time change rate, etc.

[0022] Step S2: Data preprocessing and working condition division; The multi-source monitoring data collected in step S1 is preprocessed to form a standard input dataset that meets the requirements of subsequent inversion and evaluation. Preprocessing includes at least the following: Time alignment: Aligning data from different sources and with different sampling frequencies according to a unified time reference; Data cleaning: Remove out-of-range values, communication interruption values, obviously invalid values, and abnormal jump values; Missing data repair: short-term missing data is repaired using interpolation, neighborhood completion, or sliding estimation methods; Denoising and smoothing: Smoothing signals with large high-frequency disturbances such as temperature, pressure, and vibration, while retaining effective trend information that reflects changes in state; Feature window construction: Construct sliding windows, fixed windows, or event-triggered windows according to preset analysis cycles.

[0023] After data preprocessing, the unit's operating status is divided into operating conditions. These operating condition divisions include at least one or more of the following: steady-state operation, load increase, load decrease, start-up, shutdown, and transition. The criteria for operating condition division may include parameters such as the rate of change of rotational speed, the rate of change of active power, the rate of change of axial load, and the rate of change of guide vane opening.

[0024] The purpose of dividing the work conditions is that the thermal state of the bearing bush, the oil film bearing state, and the characteristics of the shaft center trajectory differ significantly under different work conditions. If a unified model is directly used for inversion and evaluation, it is easily affected by the migration of work conditions, leading to deviations in the evaluation results. Therefore, this invention establishes corresponding characteristic benchmarks, health reference intervals, and inversion parameter constraints for different work conditions to improve the accuracy and stability of the inversion results and safety assessment results.

[0025] Step S3: Establish a thermo-dynamic coupled inversion model and invert key features of the oil film; In this invention, the bearing oil film state is not directly measured by an oil film thickness sensor. Instead, it utilizes multi-source monitoring parameters such as bearing temperature distribution, lubricating oil parameters, shaft trajectory, and vibration response to construct thermal and dynamic state sub-models. These sub-models are then coupled and solved to obtain the equivalent oil film thickness change and the oil film dynamic stiffness characteristic quantity, such as... Figure 2 As shown.

[0026] The process of constructing thermal state characterization parameters is as follows: Suppose that at a certain analysis time or within an analysis window, there are n temperature measurement points arranged on the bearing, and the temperatures at each measurement point are respectively... Then the average temperature of the bearing bush is: ; The maximum temperature of the bearing bush is: ; The bearing temperature difference is: ; In the formula, min() represents finding the minimum value; To characterize the degree of non-uniformity in the temperature field distribution, a temperature dispersion index is introduced: ; Assume the lubricating oil inlet temperature is The outlet temperature is The lubricating oil temperature rises to To reflect the rate of temperature change, let the average bearing temperature at the current moment and the previous moment be respectively... and Where Δt represents the time interval between the current moment and the previous moment, the average temperature rise rate is: ; The thermal state characterization quantity H of the bearing T for: ; Where H T This is a characteristic quantity of the thermal state of the bearing; Let be the weighting coefficients of each characteristic term of the thermal state, and satisfy . ; These represent the normalized average bearing temperature, bearing temperature difference, temperature dispersion index, lubricating oil temperature rise, and average temperature rise rate, respectively.

[0027] Normalization can be performed using a relatively healthy baseline, for example, for any feature. Its normalized expression is: ; in For the current eigenvalue, These are the normalized current eigenvalues; For health reference values; This is the risk boundary value corresponding to this feature. The larger the value, the further the characteristic deviates from a healthy state.

[0028] The process of constructing dynamic state characterization quantities is as follows: Let the displacements of the axis trajectory in the two orthogonal directions be respectively and Then, within an analysis window, the coordinates of the trajectory center are: ; Where m is the number of sampling points within the analysis window; Let be the trajectory coordinates of the j-th sampling point.

[0029] The axial eccentricity is defined as follows: ; If the bearing is designed with a radial clearance of c, then the eccentricity is: ; To reflect changes in trajectory shape, trajectory dispersion is introduced: ; Assume the average value of the lubricating oil supply pressure is The standard deviation of pressure is Then the oil pressure fluctuation coefficient is: ; Let the vibration characteristic quantity be The effective value of vibration displacement, vibration velocity, or vibration acceleration can be taken as the dynamic state characterization quantity. Defined as: ; in As a dynamic state characterization quantity, This is the normalized value of the eccentricity. This is the normalized value of the oil pressure fluctuation coefficient; The normalized value of the vibration characteristic quantity, This is the normalized value of the trajectory dispersion; Let be the weight coefficients of each characteristic term of the dynamic state, and satisfy . .

[0030] The process of inverting the equivalent thickness change of the oil film is as follows: Let the equivalent thickness of the oil film under the healthy reference operating condition be... The equivalent thickness of the oil film under the current operating conditions is Since oil film thickness is difficult to measure directly online, this invention uses the coupled offset of thermal state characterization and dynamic state characterization to inversely determine the current equivalent thickness change of the oil film.

[0031] Define the change in equivalent thickness of the oil film. for: ; Preferably, the following inversion relationship can be used: ; Alternatively, it can be denoted as the equivalent thickness of the oil film under the current operating conditions: ; in This represents the change in the equivalent thickness of the oil film. The normalized load characteristic quantity can be obtained by normalizing the axial load, equivalent thrust, or related force parameters; This is a normalized rotational speed characteristic. These are the coefficients corresponding to each characteristic quantity in the formula for calculating the equivalent thickness of the oil film under the current operating conditions.

[0032] The physical meaning expressed by the above formula is as follows: deterioration of thermal state and abnormal dynamic state usually lead to a worsening of oil film bearing conditions, thus reducing the equivalent thickness of the oil film; while changes in load and speed have a coupled effect on oil film formation, so load and speed terms are introduced for correction.

[0033] In a preferred embodiment, the and They are defined as follows: ; Where L is the current load characteristic value; For health reference load values; The upper boundary value of the load; N is the current rotational speed; For health reference speed; This serves as the upper boundary for the rotational speed reference.

[0034] The inversion process of the dynamic stiffness characteristic of the oil film is as follows: To reflect the variation in the oil film's support capability for the rotor system, this invention further constructs a dynamic stiffness characteristic of the oil film. Preferably, it is defined as a comprehensive characteristic related to trajectory eccentricity, oil pressure fluctuation, and vibration response: ; in This represents the current characteristic quantity of the dynamic stiffness of the oil film; This is the reference value for the dynamic stiffness of the oil film under healthy reference operating conditions; This is the influence coefficient; when , , An increase in this value indicates increased shaft offset, decreased oil pressure stability, and enhanced dynamic response, typically corresponding to a decrease in oil film support rigidity. It shows a decreasing trend.

[0035] Define the oil film dynamic stiffness offset: ; in This indicates that the oil film support stiffness has decreased compared to a healthy state.

[0036] Step S4: Construct the oil film stability index; After obtaining the equivalent thickness change and dynamic stiffness characteristics of the oil film in step S3, these are combined with characteristics such as temperature, oil pressure, trajectory, and vibration to construct the oil film stability index.

[0037] The process of constructing each sub-indicator is as follows: Oil film thickness sub-index for: ;

[0038] in The closer the value is to 1, the closer the oil film thickness is to a healthy state; The smaller the value, the more significant the thinning of the oil film.

[0039] Oil film dynamic stiffness sub-index for: ; in The closer it is to 1, the closer the oil film support stiffness is to a healthy state; The smaller the value, the more significant the decrease in support capacity.

[0040] Temperature field stability sub-index for: ; in These are the weighting coefficients; The larger the value, the more stable the temperature distribution; The smaller the value, the more obvious the phenomenon of local hotspots, widening temperature differences, or accelerated warming.

[0041] Define oil pressure stability sub-indices for: ; in The larger the value, the more stable the oil pressure; The smaller the value, the more pronounced the oil pressure fluctuation.

[0042] Trajectory vibration stability sub-index for: ; in These are the weighting coefficients in the calculation of the trajectory vibration stability sub-index. The larger the value, the more stable the trajectory and vibration state; The smaller the value, the more obvious the axis offset, trajectory disorder, or vibration abnormality.

[0043] By weighting and combining the above sub-indices, we obtain the oil film stability index: ; in The oil film stability index; The weight coefficients of each sub-indicator in the oil film stability index calculation term are, and satisfy the following conditions: .

[0044] Preferably, and The sum of the values ​​is greater than , , The weight of any single item is determined to highlight the core role of the inverted oil film thickness and dynamic stiffness characteristics in this invention.

[0045] The larger the value, the better the overall stability of the oil film; The smaller the value, the worse the oil film stability and the higher the risk of bearing operation.

[0046] Step S5: Calculate the bearing safety margin index; Step S5, based on step S4, further compares the current oil film stability index with the healthy reference range or critical boundary to form a bearing safety margin index that can directly characterize the safety level.

[0047] Assuming that, under current operating conditions, the lower limit of the healthy reference value for the oil film stability index obtained from historical healthy operation samples is: The risk critical boundary is And satisfy .

[0048] The bearing safety margin index is: ; in The bearing safety margin index has the following physical meaning: when A value ≥1 indicates that the current state is within the healthy reference range or better than the lower limit of health. when When the value is less than 1, it indicates that although the current state has not crossed the critical boundary, it has entered the safety margin contraction zone. when When the value is ≤0, it indicates that the current state has reached or exceeded the critical boundary, and there is a high risk.

[0049] To avoid bias caused by direct comparison under different operating conditions, separate systems can be established for each operating condition. and This enables the assessment of safety margins for different operating conditions.

[0050] Step S6: Perform safety margin evolution judgment and degradation trend identification; After obtaining the bearing safety margin index, it is necessary not only to judge the safety status at the current moment, but also to analyze the change process of the safety margin over a continuous period of time in order to identify the degradation trend and potential critical thinning risk.

[0051] To further determine the evolution trend of the bearing condition, the rate of change of safety margin R is calculated at consecutive time points. m : ; in This represents the current bearing safety margin indicator. This refers to the bearing safety margin indicator at the previous moment; The rate of change of the safety margin.

[0052] Its discriminative significance is as follows: when When the value is approximately 0, it indicates that the safety margin is basically stable. when When the value is less than 0, it indicates that the safety margin is decreasing and the condition of the bearing is deteriorating. when Continuously less than the preset threshold If the time frame is too high, it indicates that the safety margin is decreasing rapidly, and it can be determined that there is a risk of rapid deterioration.

[0053] Step S7: Determine the level of early warning; Based on the safety margin index obtained in step S5 and the safety margin change rate obtained in step S6, the bearing operating status is classified and warned. Preferably, the warning levels include at least four levels: normal, attention, warning, and alarm.

[0054] Specifically, based on and Preferably, the state can be determined according to the following rules: when ≥1 and ≥ When this occurs, it is considered a normal state; When 0.6≤ <1 and When the value is less than 0, it is considered to be in a state of concern. When 0 < <0.6 and < When this occurs, it is determined to be a warning state; when ≤0 or Continued decline and Significantly smaller than When this occurs, it is determined to be an alarm state.

[0055] in The threshold for the rate of change of safety margin can be set based on historical sample statistics, unit test results, or engineering experience.

[0056] Step S8: Control suggestion output; When the warning level reaches the preset conditions, the present invention outputs corresponding control suggestions or control instructions based on the risk type, dominant abnormal characteristics and current operational constraints.

[0057] The control recommendations or control instructions include at least one or more of the following: Cooling flow rate adjustment suggestions or adjustment instructions; Recommendations for lubrication system inspection; Recommendations or directives to limit the rate of load change; Recommendations for reduced-load operation or instructions for reduced-load control; Recommendation to switch to a more conservative operating mode; Inspection and repair or manual key inspection prompts.

[0058] When temperature-related anomalies dominate, cooling flow adjustment recommendations are triggered first; when oil pressure fluctuations and trajectory eccentricity anomalies dominate and are related to rapid load changes, load change rate limiting recommendations are triggered first; when the safety margin is lower than the preset threshold and continues to deteriorate, load reduction operation recommendations or control commands are triggered.

[0059] Control suggestions can be sent to the monitoring and alarm system, the operation monitoring interface, or the intelligent operation and maintenance platform; control commands can be sent to the unit control system to execute corresponding control actions under the conditions of meeting safety interlocking conditions and manual confirmation.

[0060] The invention will be further illustrated below using the online safety assessment of the thrust bearing bush of a certain hydroelectric generator set as an example.

[0061] The unit is connected to a computer monitoring system, a condition monitoring system, and a lubricating oil monitoring system, enabling it to collect real-time data on unit operating parameters, bearing temperature, lubricating oil parameters, shaft trajectory, and vibration parameters. The system is set to perform online calculations every 60 seconds, using the most recent 10 minutes of data as the analysis window to provide a real-time assessment of the current bearing operating status.

[0062] Real-time data acquisition: At a certain evaluation time t, the system acquires real-time data from the most recent 10 minutes, mainly including: Current rotational speed N = 150 r / min; Active power P = 280MW; The current load characteristic value L = 0.82; The temperatures at the six temperature measuring points of the thrust bearing are as follows: ; Lubricating oil inlet temperature is =38.4℃, outlet temperature =43.6℃; Average lubricating oil supply pressure Pressure standard deviation ; The coordinates of the trajectory center are calculated within the current analysis window using the displacement sampling sequence of the two orthogonal directions of the axis trajectory. ; The bearing is designed with a radial clearance of c = 0.20 mm; Effective value of vibration characteristic quantity =1.92mm / s.

[0063] The system also saves health reference data under the same operating conditions, specifically including: Equivalent thickness of oil film under healthy reference operating conditions =0.120mm; Reference value of oil film dynamic stiffness under healthy reference conditions =1.00; Lower limit of health reference =0.85; Risk critical boundary =0.40.

[0064] Calculation of thermal state characterization parameters: The average temperature of the bearing bush is calculated based on the real-time collected bearing temperature data: ; Maximum temperature is The minimum temperature is The bearing temperature difference is .

[0065] The temperature dispersion index is: ; The lubricating oil temperature rises to: ; If the average temperature of the bearing bush at the previous evaluation time was 68.95℃, and the interval between the two evaluations was 60s, then the average temperature rise rate was: ; Assume the health reference values ​​and risk boundary values ​​under this operating condition are as follows: The healthy reference value for the average temperature of the bearing bush is 64.0℃, and the corresponding risk boundary value is 75.0℃. The healthy reference value for the temperature difference of the bearing bush is 1.5℃, and the corresponding risk boundary value is 5℃. The healthy reference value for the bearing temperature dispersion index is 0.5℃, and the corresponding risk boundary value is 2.0℃. The healthy reference value for lubricating oil temperature rise is 3.0℃, and the corresponding risk boundary value is 7.0℃; The healthy reference value for the average temperature rise rate of the bearing bush is 0, and the corresponding risk boundary value is 0.02℃ / s.

[0066] The normalization results for each feature are as follows:

[0067] Let the thermal state weights be: ; The characteristic quantity of the thermal state of the bearing is:

[0068] Calculation of dynamic state characterization parameters: Based on the coordinates of the trajectory center, the axial eccentricity is: ; The eccentricity rate is: ; Suppose the calculated trajectory dispersion within the analysis window is as follows: ; The oil pressure fluctuation coefficient is: ; Vibration characteristics are measured by the effective value of vibration velocity: ; Let the health reference values ​​and risk boundary values ​​corresponding to each feature be as follows: The healthy reference value for eccentricity is 0.20, and the corresponding risk boundary value is 0.60. The health reference value for trajectory dispersion is 0.010 mm, and the corresponding risk boundary value is 0.040 mm. The healthy reference value for the oil pressure fluctuation coefficient is 0.020, and the corresponding risk boundary value is 0.100. The healthy reference value for the vibration characteristic quantity is 1.00 mm / s, and the corresponding risk boundary value is 4.00 mm / s.

[0069] The normalization result is:

[0070] Let the dynamic state weights be: ; The dynamic state characterization quantity is:

[0071] The process of inverting the equivalent thickness change of the oil film is as follows: Let the normalized load characteristic quantity under the current operating condition be... The normalized rotational speed characteristic is .

[0072] Let the inversion model coefficients be: ; The current equivalent thickness of the oil film is:

[0073] The change in the equivalent thickness of the oil film is: ; The results show that the current equivalent thickness of the oil film has been reduced by approximately 0.0204 mm compared to the healthy reference state.

[0074] The inversion process of the dynamic stiffness characteristic of the oil film is as follows: Let the influence coefficient be The characteristic quantity of the dynamic stiffness of the oil film is: ; The oil film dynamic stiffness offset is: ; The results show that the current oil film support stiffness has decreased significantly compared to the healthy state.

[0075] The calculation process for the oil film stability index is as follows: First, calculate each sub-indicator.

[0076] Oil film thickness sub-indicator: ; Oil film dynamic stiffness sub-index: ; Let the temperature field stability weight be... The temperature field stability sub-index is: ;

[0077] The sub-indices for oil pressure stability are: ; Let the trajectory vibration stability weight be... The trajectory vibration stability sub-index is:

[0078] Let the weight of the oil film stability index be... The oil film stability index is:

[0079] The calculation process for the bearing safety margin and rate of change is as follows: Based on the health reference lower limit and risk threshold set in this embodiment: ; The bearing safety margin is: ; Similarly, the safety margin at the previous assessment time was The rate of change of the safety margin is (unit: s). -1 That is, per second): ; The results show that the current bearing bush is still above the critical boundary, but the safety margin has shrunk significantly and is showing a continuous downward trend.

[0080] The status determination and early warning output process is as follows: Let the system state discrimination rule be: when When the value is ≥1, it is considered a normal state; When 0.6≤ <1 and When the value is less than 0, it is considered to be in a state of concern. When 0 < <0.6 and < -0.00050s -1 When this occurs, it is determined to be a warning state; when When the value is ≤0, it is determined to be an alarm state.

[0081] Due to the following in this embodiment: ; Therefore, the system determines that the current bearing status is a state of interest.

[0082] Furthermore, combining the results of each sub-indicator, we can see that: =0.676 is low, indicating that the oil film support stiffness has decreased significantly; The value of 0.525 is relatively low, indicating that the oil pressure fluctuates significantly. =0.611 and =0.606, both of which are lower than the center value of the healthy range, indicating that the stability of the temperature field and trajectory has deteriorated.

[0083] Therefore, the system outputs the following operating suggestions: It is recommended to focus on checking the stability of the lubricating oil supply; It is recommended to appropriately increase the cooling flow rate; It is recommended to limit short-term, rapid load increases. If the safety margin continues to decline for three consecutive assessment cycles, the alert status will be upgraded and it will be recommended to reduce the unit load.

[0084] In this embodiment, the system continuously receives real-time data every 60 seconds and repeatedly executes the above calculation process, thereby dynamically tracking the bearing safety margin. 𝑠 and the rate of change of safety margin R 𝑚 The process of change. Compared with the alarm method that only uses a fixed temperature threshold, the present invention can not only identify whether there is an anomaly, but also identify early risks such as oil film thinning, decrease in support stiffness and shrinkage of safety margin in advance, and issue attention or warning information before the risk develops into a serious over-limit, so as to give operators time to deal with it.

[0085] In summary, compared with existing technologies, this invention fully utilizes multi-source data such as rotational speed, load, bearing temperature, lubricating oil parameters, shaft trajectory, and vibration obtainable from existing monitoring systems of hydro-generator units. It constructs a coupled inversion model combining thermal and dynamic states, achieving indirect identification of the equivalent thickness change and dynamic stiffness characteristics of the oil film without the need for directly deploying dedicated oil film thickness sensors. Furthermore, it quantitatively assesses the bearing operating status, risk proximity, and degradation evolution trend through oil film stability index, safety margin index, and their rate of change. This method overcomes the problems of traditional bearing safety monitoring relying mainly on single-parameter threshold alarms, difficulty in reflecting the true load-bearing state of the oil film, and difficulty in identifying early risks. It not only improves the accuracy, foresight, and interpretability of bearing anomaly identification but also provides direct evidence for control measures such as cooling regulation, load change rate limiting, and load reduction operation, demonstrating strong engineering feasibility and application value.

[0086] The above description is merely a preferred embodiment of the present invention. It should be understood that the present invention is not limited to the forms disclosed herein and should not be construed as excluding other embodiments. It can be used in various other combinations, modifications, and environments, and can be altered within the scope of the concept described herein through the above teachings or related technologies or knowledge. Modifications and variations made by those skilled in the art that do not depart from the spirit and scope of the present invention should be within the protection scope of the appended claims.

Claims

1. A method for safety assessment of turbine generator bearings based on coupled inversion and safety margin, characterized in that, Includes the following steps: S1. Real-time acquisition of multi-source monitoring data during the operation of the hydro-generator unit; S2. Preprocess the collected data and divide it into operating conditions to form an input dataset corresponding to the current unit operating status; S3. Based on multi-source monitoring data, establish a thermo-dynamic coupled inversion model to obtain the equivalent thickness change of the oil film and the dynamic stiffness characteristic of the oil film. S4. The equivalent thickness change of the oil film and the dynamic stiffness characteristic of the oil film are fused with multi-source monitoring data to construct an oil film stability index. S5. Calculate the bearing safety margin index based on the historical healthy operating range or the preset safety boundary; S6. Combining the bearing safety margin index and its rate of change, perform safety margin evolution judgment and degradation trend identification; S7. Graded early warning determination; S8. When the warning level reaches the preset conditions, output the corresponding operation control suggestions or control commands to perform cooling flow regulation, load change rate limitation or load reduction operation.

2. The method for assessing the bearing safety of a hydro-generator unit based on coupled inversion and safety margin as described in claim 1, characterized in that, The multi-source monitoring data includes at least one or more of the following: unit operating parameters, load-related parameters, bearing thermal state parameters, lubrication system parameters, and rotor shaft attitude and dynamic response parameters; Load-related parameters, including axial load, thrust load, equivalent thrust, or equivalent load parameters that reflect changes in the rotor's stress state; The thermal parameters of the bearing include one or more of the following: multi-point temperature distribution of the bearing, oil inlet temperature, oil return temperature, and cooling medium temperature; Lubrication system parameters include one or more of the following: lubricating oil supply pressure, return oil pressure, oil pressure fluctuation value, oil temperature, oil level, and cooling flow rate; Rotor shaft attitude and dynamic response parameters, including one or more of the following: shaft trajectory, shaft swing, eccentricity, vibration displacement, vibration velocity, and vibration acceleration.

3. The method for assessing the bearing safety of a hydro-generator unit based on coupled inversion and safety margin as described in claim 1, characterized in that, The preprocessing process for the collected data is as follows: Time alignment aligns data from different sources and with different sampling frequencies according to a unified time reference; Data cleaning removes out-of-range values, communication interruption values, obviously invalid values, and abnormal jump values; Missing data repair uses interpolation, neighborhood completion, or sliding estimation to repair short-term missing data. Denoising and smoothing: Smooth high-frequency disturbances in temperature, pressure, and vibration-related signals, while retaining effective trend information reflecting state changes; Feature window construction: Construct sliding windows, fixed windows, or event-triggered windows according to preset analysis cycles.

4. The method for safety assessment of turbine generator bearings based on coupled inversion and safety margin as described in claim 1, characterized in that, The parameters used to classify operating conditions include the rate of change of rotational speed, the rate of change of active power, the rate of change of axial load, and the rate of change of guide vane opening. The operating condition classification includes at least one or more of the following: steady-state operation, load increase, load decrease, start-up, shutdown, and transition. For different operating conditions, separate input datasets and corresponding inversion parameter benchmarks are established to reduce the interference of operating condition changes on the oil film state inversion results.

5. The method for safety assessment of turbine generator bearings based on coupled inversion and safety margin as described in claim 2, characterized in that, In step S3, establishing a thermo-dynamic coupled inversion model based on multi-source monitoring data specifically includes: The average temperature and temperature difference of the bearing bush are calculated based on the multi-point temperature distribution of the bearing bush. The temperature rise and average temperature rise rate of the lubricating oil are calculated based on the inlet and outlet temperatures of the lubricating oil. At the same time, the temperature dispersion index is introduced. A thermal state characterization quantity for bearings is constructed based on bearing average temperature, bearing temperature difference, temperature dispersion index, lubricating oil temperature rise and average temperature rise rate. Dynamic state characterization parameters are constructed based on eccentricity, trajectory dispersion, oil pressure fluctuation coefficient, and vibration characteristic values. The specific process for determining the equivalent thickness change of the oil film and the dynamic stiffness characteristic of the oil film is as follows: The current equivalent thickness change of the oil film is inverted by coupling the thermal state characterization quantity and the dynamic state characterization quantity of the bearing bush. The characteristic quantity of oil film dynamic stiffness is constructed based on eccentricity, oil pressure fluctuation coefficient and vibration characteristic value.

6. The method for safety assessment of turbine generator bearings based on coupled inversion and safety margin as described in claim 1, characterized in that, Step S4 specifically includes: Based on the current equivalent oil film thickness and the benchmark value of equivalent oil film thickness under healthy reference conditions, the sub-index of oil film thickness is calculated. Based on the current dynamic stiffness characteristics of the oil film and the benchmark value of the dynamic stiffness of the oil film under the healthy reference working condition, the sub-index of dynamic stiffness of the oil film is calculated. Based on the bearing temperature difference, temperature dispersion index and average temperature rise rate, the temperature field stability sub-index is calculated. The oil pressure stability sub-index is calculated based on the oil pressure fluctuation coefficient; Based on the eccentricity, trajectory dispersion, and vibration characteristic values, the trajectory vibration stability sub-index is calculated. The oil film stability index is obtained by weighting and integrating the various sub-indicators. The weight can be one of the following: fixed weight, working condition-specific weight, or weight that is dynamically adjusted according to the operating status.

7. The method for safety assessment of turbine generator bearings based on coupled inversion and safety margin as described in claim 1, characterized in that, In step S5, the bearing safety margin index is obtained by comparing the current oil film stability index with the historical healthy operating range, the standard reference range under the same working conditions, or the preset critical boundary. The bearing safety margin index is used to represent the distance between the current bearing operating state and the safety boundary.

8. The method for safety assessment of turbine generator bearings based on coupled inversion and safety margin as described in claim 1, characterized in that, In step S6, the specific process of determining the safety margin evolution and identifying the degradation trend by combining the bearing safety margin index and its rate of change is as follows: The safety margin index of the bearing bush is calculated by sliding calculation over a continuous period to obtain the rate of change of the safety margin, and then a judgment is made: When the bearing safety margin index continues to decline and its rate of change exceeds a preset threshold, the bearing oil film is determined to be in a degraded state. When the bearing safety margin index is lower than the preset safety threshold and the rate of change continues to worsen, it is determined that there is a potential critical thinning risk.

9. The method for safety assessment of turbine generator bearings based on coupled inversion and safety margin as described in claim 1, characterized in that, In step S7, the risk classification and early warning determination specifically includes: the risk classification and early warning includes at least four levels: normal, attention, warning, and alarm. The normal level indicates that the safety margin is within a healthy range; The level of concern indicates a shift in the safety margin compared to the historical healthy range. A warning level indicates a significant decrease in safety margin and a continued trend of degradation; An alarm level indicates that the safety margin is below the critical boundary or that there is a critical risk of thinning.

10. The method for assessing the bearing safety of a hydro-generator unit based on coupled inversion and safety margin as described in claim 8, characterized in that, When abnormal bearing temperature rise becomes the dominant factor, a cooling flow adjustment suggestion or a cooling system adjustment command is triggered. When a decrease in oil film stability is associated with excessively rapid load changes, a load change rate limiting suggestion or limiting command is triggered. When the bearing safety margin index is lower than the preset safety threshold, a load reduction operation suggestion or load reduction control command is triggered.

Images