A distributed intelligent lubrication monitoring and fault warning system for wind turbine gearboxes

Abstract

The present invention discloses a distributed intelligent lubrication monitoring and fault warning system for wind power gearboxes, which relates to the technical field of wind power gearbox monitoring. The system includes a data acquisition module, a data storage module, a data processing module, a lubrication monitoring module, a lubrication evaluation module, a fault warning module, and a user interface module. The data acquisition module is used to collect state parameters of the lubrication system of the wind power gearbox; the data storage module is used to store the collected state parameters; the data processing module is used to pre-process the state parameters; the lubrication monitoring module is used to give risk prompts based on the predicted state parameter values; the lubrication evaluation module is used to comprehensively evaluate the lubricating oil state; and the fault warning module is used to give early warnings. The present invention can give prompts in advance for possible risks, thereby reducing the number of equipment maintenance and replacements, reducing maintenance costs and equipment failure rates, and facilitating operation and maintenance personnel to arrange maintenance plans in a timely manner based on the comprehensive evaluation status of the lubricating oil.

Description

Technical Field

[0001] The present invention relates to the technical field of wind power gearbox monitoring, and more particularly to a distributed intelligent lubrication monitoring and fault early warning system for wind power gearboxes. Background Art

[0002] With the rapid development of the wind energy industry, the proportion of wind power continues to increase, and the production capacity is increasing day by day, gradually becoming the main power generation. Wind turbines, especially large megawatt-class wind turbines, have higher and higher requirements for equipment reliability, availability and operation and maintenance costs. As the core transmission component of wind turbines, wind turbine gearboxes bear huge torque and dynamic loads. The lubrication of wind turbine gears is the basic condition to ensure the normal operation of the gearbox, and its lubrication condition directly affects the stability and service life of the entire system. Oil monitoring is the most direct and effective means to evaluate the lubrication and wear status of gearboxes. Traditional lubrication management and fault warning methods often rely on regular manual inspections and centralized monitoring systems. They mainly compare with preset thresholds to trigger warnings when the monitored data exceeds the threshold to remind maintenance personnel to pay attention. Although this warning method is simple and fast, it has problems such as delayed response, low accuracy, and single consideration factors. At the same time, the complex working conditions inside the wind turbine gearbox make the changes in the lubricating oil state very sensitive. Any slight lubrication problem may cause serious wear or even sudden failure.

[0003] The existing Chinese patent with publication number CN114607571A discloses a method and system for identifying faults of offshore wind power gearboxes using lubrication system monitoring, which includes using the wind speed power curve relationship and the quartile method to eliminate abnormal data and wind curtailment and power restriction data in the SCADA historical operation data of the wind turbine set; extracting the gear oil pump outlet pressure and the gear box inlet oil pressure in the historical normal data after cleaning, expanding the data according to the slope relationship between the two, substituting the original data and the expanded data into a single classification support vector machine model, training and obtaining the boundary curve of the two parameter distributions of the gear oil pump outlet pressure and the gear box inlet oil pressure under normal conditions; substituting the normal historical data into the trained model, and calculating the upper and lower thresholds of the normal data; obtaining the smoothed output value for the unit data to be tested, and when the output index is within the threshold range, it indicates that the wind turbine gearbox is normal, otherwise the gearbox is abnormal; the present invention can timely detect gearbox faults and reduce the workload and operation and maintenance costs of wind farm operation and maintenance personnel.

[0004] This method uses a support vector machine model to train the inlet and outlet pressure parameter distribution curves of the wind turbine gearbox oil pump, and compares the calculated results with the threshold to determine the gearbox fault anomaly. Although it can detect gearbox faults in time, it is not accurate enough to determine gearbox faults only by the gearbox inlet and outlet pressures, which may increase the false alarm rate and cannot provide a comprehensive assessment of the lubrication status of the lubrication system. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a distributed intelligent lubrication monitoring and fault warning system for wind turbine gearboxes. The invention monitors the lubrication status parameters of the wind turbine gearbox in real time, and predicts a period of time in the future based on historical offline status parameters. The predicted status parameters are used to conduct a comprehensive status evaluation of the lubrication system through a fuzzy comprehensive evaluation method. According to the evaluation results, rapid identification and early warning of lubrication abnormalities can be achieved, thereby shortening the fault diagnosis time window and reducing the risk of equipment downtime.

[0006] To achieve the above object, the present invention provides the following technical solution: including a data acquisition module, a data storage module, a data processing module, a lubrication monitoring module, a lubrication evaluation module, a fault warning module, and a user interface module, wherein:

[0007] The data acquisition module is used to collect the state parameters of the wind power gearbox lubricating oil system in real time through sensors according to the acquisition method;

[0008] The data storage module is used to integrate, encrypt and store the lubricating oil state parameters collected by the data collection module in time series;

[0009] The data processing module is used to pre-process the lubricating oil state parameter value;

[0010] The lubrication monitoring module is used to train the state parameter prediction model based on the pre-processed historical offline lubrication state parameter values, predict the lubrication state parameter values ​​for a period of time in the future by inputting real-time collected data into the trained data, and issue risk warnings based on abnormal data;

[0011] The lubrication assessment module is used to assess the current lubrication status through dynamic weights combined with fuzzy comprehensive evaluation based on the real-time collected lubricating oil status parameter values ​​when there is a risk warning;

[0012] The fault warning module is used to generate warning information according to the lubricating oil status;

[0013] The user interface module is used to display relevant warning information to operation and maintenance personnel in an intuitive manner, and also provides data query and report generation functions.

[0014] As a further improvement of the present invention, the sensor includes a temperature sensor, a pressure sensor, an oil sensor, a moisture sensor, and a particle size sensor;

[0015] The state parameters of the lubrication system include the temperature, pressure, flow, viscosity, acid value, moisture, and wear particle size of the lubricating oil;

[0016] Automatically acquire the latest lubrication system status parameters from each sensor according to the preset acquisition frequency;

[0017] Controlled collection is carried out according to event triggers. Event triggers include manual collection by staff and increasing the collection frequency when the collection status parameters exceed the safety threshold range.

[0018] As a further improvement of the present invention, the method for determining the safety threshold range specifically includes:

[0019] S1: The operating state parameter values ​​of the lubrication system are organized into a time state sequence according to the acquisition time;

[0020] S2: Draw the probability density distribution diagram of each state parameter based on the collected state parameter values, and calculate the standard deviation of each state parameter value:

[0021]

[0022] in, represents the standard deviation of the i-th lubricant state parameter value, represents the jth measured value of the i-th lubricant state parameter, k is the total number of state parameter values ​​collected, is the average value of the measured value of the i-th lubricating oil state parameter;

[0023] S3: Determine a safety threshold according to the standard deviation. The safety threshold includes a first safety threshold and a second safety threshold, which are respectively:

[0024]

[0025]

[0026] in, represents the first safety threshold of the i-th state parameter, represents the second safety threshold of the i-th state parameter, is the threshold coefficient of the i-th state parameter; the safety threshold range is determined as [ X ic1 , X ic2 ] .

[0027] As a further improvement of the present invention, the preprocessing specifically includes dividing the collected data into intervals, and performing outlier removal, missing value filling, and data normalization on the state parameter values ​​of each interval, wherein:

[0028] Interval division is done by setting time intervals;

[0029] The abnormal point elimination includes determining the safety threshold of each lubricating oil state parameter value in each interval, defining the state parameter value as an abnormal data point when the collected state parameter value exceeds the safety threshold of the interval, and replacing the abnormal data point with the mean value of the interval;

[0030] Missing value filling includes replacing the missing value of the state parameter value that is missing during the collection process with the previous state parameter value;

[0031] Data normalization is calculated using the normalization formula.

[0032] As a further improvement of the present invention, the lubrication monitoring module specifically further includes: a model training unit, a real-time prediction unit, and a risk warning unit, wherein:

[0033] The model training unit is used to train the state parameter prediction model through historical offline state parameter values;

[0034] The real-time prediction unit is used to input the lubricating oil state parameter values ​​collected in real time according to the trained state parameter prediction model, and output the predicted value of each state parameter of the lubricating oil in the future period;

[0035] The risk warning unit is used to provide early risk warnings based on the predicted values ​​of the state parameters output by the real-time prediction unit.

[0036] As a further improvement of the present invention, the state parameter prediction model adopts an ARIMA model, and the training method specifically includes the following steps:

[0037] S10: construct the time series of each state parameter value of the wind turbine gearbox lubricating oil;

[0038] Obtain the preprocessed lubrication state parameter values ​​from the data processing module, and establish the time series of each lubricant state parameter value in the following form:

[0039] ;

[0040] Wherein, G represents the time series of any lubricating oil state parameter value among the lubricating oil temperature, pressure, flow rate, viscosity, acid value and particle size; , , , , Respectively represent the lubricating oil state parameter values ​​at the 1st, 2nd, ..., nth moments; n is the maximum number of lubricating oil state parameter values ​​in the time series G;

[0041] S20: Perform a stationarity test on each lubricating oil state parameter value to determine the differential order d;

[0042] S30: Determine the lag order of the ARIMA model;

[0043] S40: Estimate the parameters of the ARIMA model using the maximum likelihood estimation method. The parameters to be estimated include the constant parameter c, the autoregressive parameter , sliding average parameter , error term ; Thus completing the training of the state parameter prediction model.

[0044] As a further improvement of the present invention, the risk warning unit provides a warning in the following manner:

[0045] The total number of state parameters collected during the prediction period is N;

[0046] When one or more lubricating oil state parameter prediction values ​​exceed the safety threshold range, the number of abnormal data points is ;

[0047] Calculate the abnormal rate of state parameters , configure the status parameter abnormality rate threshold;

[0048] When the abnormality rate is greater than or equal to the abnormality rate threshold, a risk warning is issued in advance; when the abnormality rate is less than the abnormality rate threshold, the model continues to predict the predicted value of each state parameter of the lubricant.

[0049] As a further improvement of the present invention, the lubrication state evaluation method includes:

[0050] S100: The state parameter set of indicators for constructing the lubricant comprehensive evaluation system is denoted as , of the following form:

[0051]

[0052] in, is the i-th state parameter of the lubrication state of the wind turbine gearbox, i=1, 2, ..., l, l is the total number of lubricating oil state parameters;

[0053] S200: Construct a wind turbine gearbox lubrication status level set, denoted as Q, in the following form:

[0054]

[0055] in, Indicates normal status. Indicates a state of attention. Indicates the warning state. Indicates alert status. Indicates a fault state;

[0056] S300: Calculate the dynamic weight of each lubrication state parameter, and construct a weight set of the state parameters through the calculated dynamic weight, in the following form:

[0057]

[0058] in, represents the dynamic weight of the i-th state parameter; i=1, 2, ..., l;

[0059] S400: Select the normal membership function and calculate the membership of each state parameter to each state level, recorded as , , , , Respectively represent the state parameters Status level ~ The membership degree is composed into a relational fuzzy matrix R;

[0060] S500: Perform matrix synthesis calculation according to the weight set of each lubrication state parameter and the relationship fuzzy matrix to obtain the comprehensive evaluation level of the lubricating oil state. The matrix synthesis calculation formula is as follows:

[0061]

[0062] in, , Corresponding to the lubrication status level of wind turbine gearbox The membership degree of the lubricating oil state level with the largest membership degree is selected as the comprehensive evaluation result of the lubrication state of the wind turbine gearbox to realize the classification of the current lubrication state.

[0063] As a further improvement of the present invention, the dynamic weight is calculated as follows:

[0064] Calculate the information entropy of each state parameter, the calculation formula is as follows:

[0065]

[0066] in, represents the information entropy of the k state parameter values ​​collected by the i-th state parameter, is the normalization coefficient, is the state parameter value The normalized value of

[0067] Determine the information entropy weight of each state parameter, and the calculation formula is as follows:

[0068]

[0069] in, is the information entropy weight of the i-th state parameter;

[0070] The dynamic weight is obtained through the information entropy weight of each state parameter;

[0071]

[0072] in, is the dynamic weight of the i-th state parameter, , They represent the maximum and minimum values ​​of the i-th state parameter that can allow the wind turbine gearbox lubrication system to operate normally, represents the average value of the measured value of the i-th state parameter, Represents the penalty coefficient.

[0073] As a further improvement of the present invention, the fault warning module is When the alarm is triggered, the operator generates early warning information to the staff, and the early warning information includes the collected lubricating oil status parameters and early warning level.

[0074] As a further improvement of the present invention, the user interface module specifically includes:

[0075] Provides a real-time updated lubricant status parameter dashboard, showing the trend changes of key parameters in chart form;

[0076] Fault case database, which records historical fault information, handling process, and effect evaluation, and provides historical fault data to operation and maintenance personnel;

[0077] Provides custom alarm categories and priorities, and color codes and labels for warning information of different levels.

[0078] Beneficial effects of the present invention:

[0079] Through real-time collection of multiple state parameters of wind turbine gearbox lubricating oil, online monitoring of wind turbine gearbox lubricating oil is realized, and the collected historical state parameters are stored in time series. The state parameter prediction model is constructed through historical offline state data training, which can predict the state parameters of lubricating oil in the future, so as to give early warning of possible risks, and evolve the monitoring of lubricating oil state into active monitoring and risk warning, giving operation and maintenance personnel sufficient time windows so that they can take preventive measures in advance for possible problems to ensure the smooth and safe operation of the wind turbine gearbox lubrication system and reduce unexpected downtime and maintenance costs;

[0080] According to the risk warning for the future time period, the real-time status parameter values ​​of the wind turbine gearbox are collected, and the lubricating oil of the wind turbine gearbox is further evaluated through fuzzy comprehensive evaluation combined with dynamic weights. It can more accurately quantify the impact of changes in lubricating oil status parameters on the overall lubrication status, and issue fault warnings in time according to the lubricating oil status. By reasonably and timely monitoring the status parameters of the lubricating oil in the wind turbine gearbox, the service life of the lubricating oil can be extended, the number of equipment maintenance and replacement can be reduced, and the maintenance cost and equipment failure rate can be reduced. At the same time, it is convenient for operation and maintenance personnel to arrange maintenance plans in time according to the comprehensive evaluation status of the lubricating oil, thereby significantly improving the refinement of equipment lubrication management and fault prevention capabilities. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Figure 1 A schematic diagram of the structure of a distributed intelligent lubrication monitoring and fault warning system for wind power gearboxes provided by the present invention;

[0082] Figure 2 An interactive process diagram of a distributed intelligent lubrication monitoring and fault warning system for a wind turbine gearbox provided by the present invention. DETAILED DESCRIPTION

[0083] In order to make the purpose, technical solution and advantages of the embodiment of the present invention clearer, the technical solution of the embodiment of the present invention will be clearly and completely described below in conjunction with the drawings of the embodiment of the present invention. Obviously, the described embodiment is a part of the embodiment of the present invention, not all of the embodiments. Based on the described embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0084] Unless otherwise defined, the technical terms or scientific terms used in the present invention should be understood by people with ordinary skills in the field to which the present invention belongs. The words "first", "second" and similar words used in the present invention do not indicate any order, quantity or importance, but are only used to distinguish different components. "Include" or "comprise" and similar words mean that the elements or objects appearing before the word include the elements or objects listed after the word and their equivalents, without excluding other elements or objects. "Connect" or "connected" and similar words are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0085] In order to keep the following description of the embodiments of the present invention clear and concise, detailed descriptions of well-known functions and well-known components are omitted.

[0086] refer to Figure 1 to Figure 2 As shown, a specific implementation of a distributed intelligent lubrication monitoring and fault warning system for wind power gearboxes of the present invention is shown, the system includes a data acquisition module, a data storage module, a data processing module, a lubrication monitoring module, a lubrication evaluation module, a fault warning module, and a user management module, wherein:

[0087] The data acquisition module is used to collect the state parameters of the wind turbine gearbox lubrication system in real time through sensors, including:

[0088] Sensors include oil sensor, particle size sensor, moisture sensor, temperature sensor, and pressure sensor;

[0089] The state parameters of the lubrication system include lubricating oil temperature, pressure, flow, viscosity, acid value, moisture, and wear particle size;

[0090] The collection methods include automatic collection and controlled collection, among which:

[0091] Automatically acquire the latest lubrication system status parameters from each sensor according to the preset acquisition frequency;

[0092] Controlled collection is carried out according to event triggers, which include manual collection by staff and increasing the collection frequency when the collection status parameters exceed the safety threshold range;

[0093] The method for determining the safety threshold specifically includes the following steps:

[0094] S1: The operating state parameter values ​​of the lubrication system are organized into a time state sequence according to the acquisition time;

[0095] S2: Draw the probability density distribution diagram of each state parameter based on the collected state parameter values, and calculate the standard deviation of each state parameter value:

[0096]

[0097] in, represents the standard deviation of the i-th lubricant state parameter value, represents the jth measured value of the i-th lubricant state parameter, k is the total amount of state parameter values ​​collected, is the average value of the measured value of the i-th lubricant state parameter, = ;

[0098] S3: Determine a safety threshold according to the standard deviation. The safety threshold includes a first safety threshold and a second safety threshold, which are respectively:

[0099]

[0100]

[0101] in, represents the first safety threshold of the i-th state parameter, represents the second safety threshold of the i-th state parameter, is the threshold coefficient of the i-th state parameter, which can be set according to the characteristics of the state parameter to determine the safety threshold range. [ X ic1 , X ic2 ] .

[0102] The data storage module is used to integrate, encrypt and store the lubricating oil state parameters collected by the data collection module in time series;

[0103] The data processing module is used to pre-process the lubricating oil state parameter values, including interval division, and outlier removal, missing value filling, and data normalization for the state parameter values ​​in each interval, where:

[0104] Interval division is done by setting time intervals;

[0105] Determine a safety threshold value for each lubricating oil condition parameter value in each interval;

[0106] When the collected state parameter value exceeds the safety threshold of the interval, the state parameter value is defined as an abnormal data point, and the abnormal data point is replaced by the mean value of the interval;

[0107] Missing value filling includes replacing the missing value of the state parameter value that is missing during the collection process with the previous state parameter value;

[0108] Data normalization is calculated using the following formula:

[0109]

[0110] in, represents the normalized value of the i-th state parameter, represents the initial acquisition value of the i-th state parameter, Indicates the standard value of each state parameter, and Respectively represent the maximum and minimum values ​​of each state parameter;

[0111] The lubrication monitoring module is used to train the state parameter prediction model according to the preprocessed historical offline lubrication state parameter values, and predict the lubrication state parameter values ​​for a period of time in the future by inputting real-time collected data into the trained data. The lubrication monitoring module specifically includes a model training unit, a real-time prediction unit, and a risk warning unit, wherein:

[0112] The model training unit is used to train the state parameter prediction model through the historical offline state parameter values. The state parameter prediction model is an ARIMA model. The training method specifically includes the following steps:

[0113] S10: construct the time series of each state parameter value of the wind turbine gearbox lubricating oil;

[0114] Obtain the preprocessed lubrication state parameter values ​​from the data processing module, and establish the time series of each lubricant state parameter value in the following form:

[0115] ;

[0116] Wherein, G represents the time series of any lubricating oil state parameter value among the lubricating oil temperature, pressure, flow rate, viscosity, acid value and particle size; , , , , Respectively represent the lubricating oil state parameter values ​​at the 1st, 2nd, ..., nth moments; n is the maximum number of lubricating oil state parameter values ​​in the time series G;

[0117] S20: Perform a stationarity test on each lubricating oil state parameter value to determine the differential order d;

[0118] ADF unit root test is used. If the lubricating oil state parameter value time series G is non-stationary, a first-order difference is performed on G, recorded as ΔG. ADF unit root test is performed on ΔG. If ΔG is non-stationary, the difference order is increased successively and a stationarity test is performed. The minimum difference order d for a stationary time series after difference is found, where d is a non-negative integer. The lubricating oil state parameter value time series that is stable after d differences is used as the data set for training the ARIMA model, recorded as G, in the following form:

[0119] ;

[0120] in, , ,……, , respectively , , , The value after d differences;

[0121] S30: Determine the lag order of the ARIMA model;

[0122] Determine the autoregressive lag order p and the moving average lag order q of the ARIMA model based on the autocorrelation diagram and partial autocorrelation diagram; draw the time series The autocorrelation graph and partial autocorrelation graph of G; the autocorrelation graph drops sharply after the lag order p and is truncated at p; the partial autocorrelation graph PACF image drops sharply after the lag order q and is truncated at q;

[0123] S40: Parameter estimation of ARIMA model;

[0124] The equation for the ARIMA model is as follows:

[0125] ;

[0126] in, For time series The tth element of g; c is a constant parameter; is the autoregressive parameter, the value range of i is 1, 2, ..., p; is the sliding average parameter, the value range of j is 1, 2, ..., q; ,……, , is the error term, where ,……, Expressed as residual, is the parameter to be estimated;

[0127] The maximum likelihood estimation method is used to estimate the parameters of the ARIMA model. The parameters to be estimated include the constant parameter c, the autoregressive parameter , sliding average parameter , error term ; Thus completing the training of the state parameter prediction model.

[0128] The real-time prediction unit is used to input the lubricating oil state parameter values ​​collected in real time according to the trained state parameter prediction model, and output the predicted value of each state parameter of the lubricating oil in the future period;

[0129] The risk warning unit is used to provide early risk warnings based on the predicted values ​​of the state parameters output by the real-time prediction unit. The specific risk warning methods include:

[0130] The total number of state parameters collected during the prediction period is N;

[0131] When one or more lubricating oil state parameter prediction values ​​exceed the safety threshold range, the number of abnormal data points is ;

[0132] Calculate the abnormal rate of state parameters , configure the status parameter abnormality rate threshold;

[0133] When the abnormality rate is greater than or equal to the abnormality rate threshold, a risk warning is issued in advance; when the abnormality rate is less than the abnormality rate threshold, the model continues to predict the predicted value of each state parameter of the lubricant.

[0134] By constructing a state parameter prediction model, the various state parameters of the lubricant can be predicted in advance for a period of time in the future, and the monitoring of the lubricant state can be transformed into active monitoring and risk warning, giving operation and maintenance personnel a sufficient time window so that they can take preventive measures in advance for possible problems to ensure the smooth and safe operation of the wind turbine gearbox lubrication system and reduce unexpected downtime and maintenance costs.

[0135] The status assessment module is used to assess the current lubrication status based on the real-time collected lubricating oil status parameter values ​​by combining dynamic weights with fuzzy comprehensive evaluation when there is an advance risk warning. The lubrication status assessment method includes the following steps:

[0136] S100: The state parameter set of indicators for constructing the lubricant comprehensive evaluation system is denoted as , of the following form:

[0137]

[0138] in, is the i-th state parameter of the lubrication state of the wind turbine gearbox, i=1, 2, ..., l, l is the total number of lubricating oil state parameters;

[0139] S200: Construct a wind turbine gearbox lubrication status level set, denoted as Q, in the following form:

[0140]

[0141] in, Indicates normal status. Indicates a state of attention. Indicates the warning state. Indicates alert status. Indicates a fault state;

[0142] S300: Calculate the dynamic weight of each lubrication state parameter, and construct a weight set of the state parameters through the calculated dynamic weight, in the following form:

[0143]

[0144] in, represents the dynamic weight of the i-th state parameter; i=1, 2, ..., l;

[0145] The dynamic weight is calculated by the entropy weight method and variable weight theory to calculate the dynamic weight of each state parameter. By assigning different weights to different state parameters, it indicates the difference in the degree to which different state parameters affect the lubrication state. It can highlight the deviation of each lubricating oil state parameter and increase the ability to evaluate the lubrication state. The specific calculation method is as follows:

[0146] Calculate the information entropy of each state parameter, the calculation formula is as follows;

[0147]

[0148] in, represents the information entropy of the k state parameter values ​​collected by the i-th state parameter, is the normalization coefficient, is the state parameter value The normalized value of ;

[0149] Determine the information entropy weight of each state parameter, and the calculation formula is as follows:

[0150]

[0151] in, is the information entropy weight of the i-th state parameter;

[0152] The dynamic weight is obtained through the information entropy weight of each state parameter;

[0153]

[0154] in, is the dynamic weight of the i-th state parameter, It indicates the maximum value of the i-th state parameter that allows the wind turbine gearbox lubrication system to operate normally. It represents the minimum value of the i-th state parameter that allows the wind turbine gearbox lubrication system to operate normally. represents the average value of the measured value of the i-th state parameter, Expressed as a penalty coefficient, it is determined according to the characteristics and attributes of each state parameter. The larger it is, the more obvious the penalty effect is, and even a small change will cause the state of the lubricant to change.

[0155] S400: Select the normal membership function and calculate the membership of each state parameter to each state level, recorded as , , , , Respectively represent the state parameters Status level ~ The degree of membership;

[0156] The membership degrees are combined into a relational fuzzy matrix R in the following form:

[0157] R=

[0158] S500: Perform matrix synthesis calculation according to the weight set of each lubrication state parameter and the relationship fuzzy matrix to obtain the comprehensive evaluation level of the lubricating oil state. The matrix synthesis calculation formula is as follows:

[0159]

[0160] in, , Corresponding to the lubrication status level The membership degree of the lubricating oil state level with the largest membership degree is selected as the comprehensive evaluation result of the lubrication state of the wind turbine gearbox to realize the classification of the current lubrication state.

[0161] If the status level is , it means that the lubricating oil status parameters of the wind turbine gearbox are all in the normal operating range, and the wind turbine gearbox lubrication system can operate normally for a long time, and no maintenance is temporarily arranged.

[0162] If the status level is , it means that there are obvious fluctuations in the lubricating oil status parameters of the wind turbine gearbox, but it does not affect the continued operation of the wind turbine gearbox lubrication system, and maintenance is arranged according to the normal cycle.

[0163] If the status level is , it means that some of the lubricating oil status parameters of the wind turbine gearbox may have exceeded the normal operating range. It can still continue to operate, but the maintenance cycle should be shortened.

[0164] If the status level is , it means that there are multiple state parameters of the lubricating oil state parameters of the wind turbine gearbox that have exceeded the normal operating range, which has affected the functioning of the wind turbine gearbox lubrication system, and maintenance should be arranged as soon as possible.

[0165] If the status level is , it means that the lubricating oil status parameters of the wind turbine gearbox have obviously exceeded the normal operating range, affecting the normal operation of the wind turbine gearbox, and it is necessary to arrange a power outage for maintenance immediately.

[0166] The lubrication assessment module conducts fuzzy comprehensive evaluation on the real-time status parameters collected during the time period when risk warnings are generated. By assigning different dynamic weights to each status parameter, the impact of changes in various status parameters of the lubricating oil on its overall performance can be accurately quantified. Timely issuance of risk warnings greatly improves the refinement of wind turbine gearbox lubrication management and fault prevention capabilities, and reduces the error recognition rate of faults.

[0167] The fault warning module is used to generate corresponding warning information according to the lubricating oil status. When the alarm is triggered, early warning information will be generated for the operation and maintenance personnel. The early warning information includes: the collected lubricating oil status parameters, early warning level, and possible fault analysis.

[0168] The user interface module is used to display relevant warning information to operation and maintenance personnel in an intuitive way, and also provides data query and report generation functions, including:

[0169] Provides a real-time updated lubricant status parameter dashboard, showing the trend changes of key parameters in the form of charts, so that operation and maintenance personnel can monitor the overall status of the wind turbine gearbox lubrication system in real time;

[0170] Provides custom alarm categories and priorities, and color codes and labels different levels of warning information to facilitate rapid response to emergencies;

[0171] Establish a fault case library to record historical fault information, handling process, and effect evaluation, provide historical fault data to operation and maintenance personnel, and improve fault handling efficiency.

[0172] Working principle and its effect:

[0173] The present invention realizes online monitoring of the lubricating oil in the wind turbine gearbox by real-time collection of multiple state parameters of the lubricating oil in the wind turbine gearbox, stores the collected historical state parameters in time series, and builds a state parameter prediction model through historical offline state data training, which can predict the state parameters of the lubricating oil in the future, thereby giving advance warnings of possible risks, and evolving the monitoring of the lubricating oil state into active monitoring and risk warning, giving operation and maintenance personnel a sufficient time window so that they can take preventive measures in advance for possible problems to ensure the smooth and safe operation of the lubrication system of the wind turbine gearbox, and reduce unexpected downtime and maintenance costs;

[0174] According to the risk warning for the future time period, the real-time status parameter values ​​of the wind turbine gearbox are collected, and the lubricating oil of the wind turbine gearbox is further evaluated through fuzzy comprehensive evaluation combined with dynamic weights. It can more accurately quantify the impact of changes in lubricating oil status parameters on the overall lubrication status, and issue fault warnings in time according to the lubricating oil status. By reasonably and timely monitoring the status parameters of the lubricating oil in the wind turbine gearbox, the service life of the lubricating oil can be extended, the number of equipment maintenance and replacement can be reduced, and the maintenance cost and equipment failure rate can be reduced. At the same time, it is convenient for operation and maintenance personnel to arrange maintenance plans in time according to the comprehensive evaluation status of the lubricating oil, thereby significantly improving the refinement of equipment lubrication management and fault prevention capabilities.

[0175] In addition, although exemplary embodiments have been described in the present invention, the scope includes any and all embodiments based on the present invention with equivalent elements, modifications, omissions, combinations (e.g., various embodiments intersecting schemes), adaptations or changes. The elements in the claims will be interpreted broadly based on the language adopted in the claims, and are not limited to the examples described in this specification or during the execution of this application, and the examples will be interpreted as non-exclusive. Therefore, this specification and examples are intended to be considered as examples only, and the true scope and spirit are indicated by the following claims and the full scope of their equivalents.

[0176] The above description is intended to be illustrative rather than restrictive. For example, the above examples (or one or more schemes thereof) may be used in combination with each other. For example, a person of ordinary skill in the art may use other embodiments when reading the above description. In addition, in the above-mentioned specific embodiments, various features may be grouped together to simplify the present invention. This should not be interpreted as an intention that a disclosed feature that is not required to be protected is necessary for any claim. On the contrary, the subject matter of the present invention may be less than all the features of a particular disclosed embodiment. Thus, the following claims are incorporated herein as examples or embodiments in the specific embodiments, wherein each claim is independently a separate embodiment, and it is considered that these embodiments can be combined with each other in various combinations or arrangements. The scope of the present invention should be determined with reference to the appended claims and the full scope of equivalent forms granted by these claims.

[0177] The above embodiments are only exemplary embodiments of the present invention and are not intended to limit the present invention. The protection scope of the present invention is defined by the claims. Those skilled in the art may make various modifications or equivalent substitutions to the present invention within the essence and protection scope of the present invention, and such modifications or equivalent substitutions shall also be deemed to fall within the protection scope of the present invention.

Claims

1. A distributed intelligent lubrication monitoring and fault warning system for wind turbine gearboxes, characterized in that: It includes data acquisition module, data storage module, data processing module, lubrication monitoring module, lubrication evaluation module, fault warning module and user interface module, among which: The data acquisition module is used to collect the state parameters of the wind power gearbox lubricating oil system in real time through sensors according to the acquisition method; The data storage module is used to integrate, encrypt and store the lubricating oil state parameters collected by the data collection module in time series; The data processing module is used to pre-process the lubricating oil state parameter value; The lubrication monitoring module is used to train the state parameter prediction model based on the pre-processed historical offline lubrication state parameter values, predict the lubrication state parameter values ​​for a period of time in the future by inputting real-time collected data into the trained data, and issue risk warnings based on abnormal data; The state parameter prediction model adopts the ARIMA model, and the training method specifically includes the following steps: S10: construct the time series of each state parameter value of the wind turbine gearbox lubricating oil; Obtain the preprocessed lubrication state parameter values ​​from the data processing module, and establish the time series of each lubricant state parameter value in the following form: ； Wherein, G represents the time series of any lubricating oil state parameter value among the lubricating oil temperature, pressure, flow rate, viscosity, acid value and particle size; , , , , Respectively represent the lubricating oil state parameter values ​​at the 1st, 2nd, ..., nth moments; n is the maximum number of lubricating oil state parameter values ​​in the time series G; S20: Perform a stationarity test on each lubricating oil state parameter value to determine the differential order d; S30: Determine the lag order of the ARIMA model; S40: Estimate the parameters of the ARIMA model using the maximum likelihood estimation method. The parameters to be estimated include the constant parameter c, the autoregressive parameter , sliding average parameter , error term ;Thereby completing the training of the state parameter prediction model; The lubrication assessment module is used to assess the current lubrication status through dynamic weights combined with fuzzy comprehensive evaluation based on the real-time collected lubricating oil status parameter values ​​when there is a risk warning; Lubrication condition assessment methods include: S100: The state parameter set of indicators for constructing the lubricant comprehensive evaluation system is denoted as , of the following form: ； in, is the i-th state parameter of the wind turbine gearbox lubrication state, i=1, 2, ..., l , l is the total number of lubricating oil status parameters; S200: Construct a wind turbine gearbox lubrication status level set, denoted as Q, in the following form: ； in, Indicates normal status. Indicates a state of attention. Indicates the warning state. Indicates alert status. Indicates a fault state; S300: Calculate the dynamic weight of each lubrication state parameter, and construct a weight set of the state parameters through the calculated dynamic weight, in the following form: ； in, represents the dynamic weight of the i-th state parameter; i=1, 2, ..., l ; S400: Select the normal membership function and calculate the membership of each state parameter to each state level, recorded as , , , , Respectively represent the state parameters Status level ~ The membership degree is composed into a relational fuzzy matrix R, which is in the following form: R= ； S500: Perform matrix synthesis calculation according to the weight set of each lubrication state parameter and the relationship fuzzy matrix to obtain the comprehensive evaluation level of the lubricating oil state. The matrix synthesis calculation formula is as follows: ； in, , Corresponding to the lubrication status level of wind turbine gearbox The membership degree of the lubricating oil state level with the largest membership degree is selected as the comprehensive evaluation result of the lubrication state of the wind turbine gearbox, so as to realize the classification of the current lubrication state; Formulate maintenance strategies for wind turbine gearbox lubrication systems based on lubrication status levels: If the status level is , it means that the lubricating oil status parameters of the wind turbine gearbox are in the normal operating range, the wind turbine gearbox lubrication system is operating normally, and no maintenance is required; If the status level is , it means that the lubricating oil state parameters of the wind turbine gearbox fluctuate, but it does not affect the continued operation of the wind turbine gearbox lubrication system, and maintenance is scheduled according to the normal cycle; If the status level is , it means that there are no more than c parameters of the lubricating oil state parameters of the wind turbine gearbox that exceed the normal operating range, and the wind turbine gearbox lubrication system can continue to operate, shortening the maintenance cycle and arranging maintenance; If the status level is , it means that there are more than c parameters of the lubricating oil state of the wind turbine gearbox that are beyond the normal operating range, which affects the functioning of the lubrication system of the wind turbine gearbox, and maintenance should be arranged as soon as possible; If the status level is , it means that the lubricating oil state parameters of the wind turbine gearbox are beyond the normal operating range, affecting the normal operation of the wind turbine gearbox, and a power outage for maintenance should be arranged immediately; The dynamic weight is calculated as follows: Calculate the information entropy of each state parameter, the calculation formula is as follows: ； in, represents the information entropy of the k state parameter values ​​collected by the i-th state parameter, is the normalization coefficient, is the state parameter value The normalized value of Determine the information entropy weight of each state parameter, and the calculation formula is as follows: ； in, is the information entropy weight of the i-th state parameter; The dynamic weight is obtained through the information entropy weight of each state parameter; ； in, is the dynamic weight of the i-th state parameter, , They represent the maximum and minimum values ​​of the i-th state parameter that can allow the wind turbine gearbox lubrication system to operate normally, represents the average value of the measured value of the i-th state parameter, represents the penalty coefficient; The fault warning module is used to generate warning information according to the lubricating oil status; The user interface module is used to display relevant warning information to operation and maintenance personnel in an intuitive manner, and also provides data query and report generation functions.

2. A wind turbine gearbox distributed intelligent lubrication monitoring and fault warning system according to claim 1, characterized in that: The sensors include temperature sensors, pressure sensors, oil sensor, moisture sensor, and particle size sensor; The state parameters of the lubrication system include the temperature, pressure, flow, viscosity, acid value, moisture, and wear particle size of the lubricating oil; Automatically acquire the latest lubrication system status parameters from each sensor according to the preset acquisition frequency; Controlled collection is carried out according to event triggers. Event triggers include manual collection by staff and increasing the collection frequency when the collection status parameters exceed the safety threshold range.

3. A wind turbine gearbox distributed intelligent lubrication monitoring and fault warning system according to claim 2, characterized in that: The method for determining the safety threshold range specifically includes: S1: The operating state parameter values ​​of the lubrication system are organized into a time state sequence according to the acquisition time; S2: Draw the probability density distribution diagram of each state parameter based on the collected state parameter values, and calculate the standard deviation of each state parameter value: ； in, represents the standard deviation of the i-th lubricant state parameter value, represents the jth measured value of the i-th lubricant state parameter, k is the total number of state parameter values ​​collected, is the average value of the measured value of the i-th lubricating oil state parameter; S3: Determine a safety threshold according to the standard deviation. The safety threshold includes a first safety threshold and a second safety threshold, which are respectively: ； ； in, represents the first safety threshold of the i-th state parameter, represents the second safety threshold of the i-th state parameter, is the threshold coefficient of the i-th state parameter; the safety threshold range is determined as .

4. A wind turbine gearbox distributed intelligent lubrication monitoring and fault warning system according to claim 3, characterized in that: The preprocessing specifically includes dividing the collected data into intervals, and removing abnormal points, filling missing values, and normalizing the data for the state parameter values ​​in each interval, wherein: Interval division is done by setting time intervals; The abnormal point elimination includes determining the safety threshold of each lubricating oil state parameter value in each interval, defining the state parameter value as an abnormal data point when the collected state parameter value exceeds the safety threshold of the interval, and replacing the abnormal data point with the mean value of the interval; Missing value filling includes replacing the missing value of the state parameter value that is missing during the collection process with the previous state parameter value; Data normalization is calculated using the normalization formula.

5. A wind turbine gearbox distributed intelligent lubrication monitoring and fault warning system according to claim 4, characterized in that: The lubrication monitoring module specifically includes: a model training unit, a real-time prediction unit, and a risk warning unit, wherein: The model training unit is used to train the state parameter prediction model through historical offline state parameter values; The real-time prediction unit is used to input the lubricating oil state parameter values ​​collected in real time according to the trained state parameter prediction model, and output the predicted value of each state parameter of the lubricating oil in the future period; The risk warning unit is used to provide early risk warnings based on the predicted values ​​of the state parameters output by the real-time prediction unit.

6. A wind turbine gearbox distributed intelligent lubrication monitoring and fault warning system according to claim 5, characterized in that: The risk warning unit provides a warning in the following ways: The total number of state parameters collected during the prediction period is N; When one or more lubricating oil state parameter prediction values ​​exceed the safety threshold range, the number of abnormal data points is ; Calculate the abnormal rate of state parameters , configure the status parameter abnormality rate threshold; When the abnormality rate is greater than or equal to the abnormality rate threshold, a risk warning is issued in advance; when the abnormality rate is less than the abnormality rate threshold, the model continues to predict the predicted value of each state parameter of the lubricant.

7. A wind turbine gearbox distributed intelligent lubrication monitoring and fault warning system according to claim 6, characterized in that: The fault warning module is used to detect the lubrication status level. When the alarm is triggered, the operator generates early warning information to the staff, and the early warning information includes the collected state parameters of the lubricating oil and the early warning level.

8. A wind turbine gearbox distributed intelligent lubrication monitoring and fault warning system according to claim 7, characterized in that: The user interface module specifically includes: Provides a real-time updated lubricant status parameter dashboard, showing the trend changes of key parameters in chart form; Fault case database, which records historical fault information, handling process, and effect evaluation, and provides historical fault data to operation and maintenance personnel; Provides custom alarm categories and priorities, and color-codes and labels different levels of warning information.

Images