An on-line monitoring system for lubricating oil quality

Abstract

The present application relates to the technical field of gear box lubrication state monitoring, in particular to a kind of lubricating oil quality on-line monitoring system, comprising: data acquisition module obtains the disturbance signal and working condition parameter of large wind turbine main gear box;First processing module quantifies implicit disturbance characteristics;Second processing module identifies core chemical parameters, and core chemical parameters include: metal saponification reaction rate constant and colloid cluster electrophoretic mobility;Third processing module predicts macro rheological phase change, determines non-newtonian index;Working condition solution module calculates shear rate;Risk calculation module calculates abnormal wear risk index;State evaluation module determines evaluation result classification;Closed-loop control module responds to evaluation result classification, when evaluation result classification is secondary early warning, executes active inhibition strategy.The present application solves the problem that traditional technology cannot obtain micro-chemical kinetics parameters, and lays a foundation for accurate prediction.

Description

Technical Field

[0001] This invention relates to the field of gearbox lubrication condition monitoring technology, specifically to an online lubricating oil quality monitoring system. Background Technology

[0002] In the operation and maintenance of large wind turbine generators, the main gearbox, as a core transmission component, is directly affected by the condition of its lubricating oil, which in turn affects the safety and lifespan of the equipment. During operation, the gearbox operates in a complex electromechanical-chemical coupling environment. Traditional lubricating oil monitoring technologies typically focus on the oil's conventional physicochemical properties, such as viscosity, moisture content, and wear particles, and assess these through periodic sampling and offline analysis. However, in-service lubricating oils are subject to various physical field disturbances, including electromagnetic harmonics, stress waves, and micro-discharges. These disturbances can induce shear-thickening phase transitions in the oil, leading to anomalous wear phenomena where higher flow rates result in faster wear. Existing technologies are unable to detect risks arising from electrochemical self-organized critical states. Traditional lubrication models cannot describe this phenomenon, and monitoring systems struggle to identify the core chemical parameters initiating the phase transition and their cross-scale coupling relationships online.

[0003] Existing online monitoring systems lack a complete modeling link from multi-physics perturbations to macroscopic rheological property prediction, making it impossible to effectively warn of such anomalous wear risks. When anomalies occur, there is also a lack of active suppression strategies that can break the electro-chemical positive feedback loop at its root, resulting in the inability to effectively avoid catastrophic failures. Therefore, how to achieve online monitoring and early warning of anomalous wear caused by lubricating oil shear thickening, and provide a risk assessment and closed-loop control method that can couple multi-physics perturbations, chemical kinetics, and rheological properties, is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] To solve the above-mentioned technical problems, the present invention provides an online lubricating oil quality monitoring system. Specifically, the technical solution of the present invention includes:

[0005] The data acquisition module is used to acquire disturbance signals and operating parameters of the main gearbox of a large wind turbine. The disturbance signals include electromagnetic harmonics, stress wave signals and micro-discharge pulses; the operating parameters include real-time oil temperature and gear angular velocity.

[0006] The first processing module is used to quantify the latent perturbation characteristics based on the perturbation signal. The latent perturbation characteristics include: harmonic injection energy, acoustic emission spectrum entropy, and micro-area discharge power.

[0007] The second processing module is used to identify core chemical parameters based on micro-area discharge power, harmonic injection energy and real-time oil temperature. The core chemical parameters include: metal saponification reaction rate constant and colloidal cluster electrophoretic mobility.

[0008] The third processing module is used to predict macroscopic rheological phase transitions and determine non-Newtonian exponents based on core chemical parameters and acoustic emission spectrum entropy.

[0009] The working condition calculation module is used to calculate the shear rate based on the gear angular velocity and preset gear geometric factors;

[0010] The risk calculation module is used to calculate the abnormal wear risk index by combining shear rate and non-Newtonian exponent.

[0011] The condition assessment module is used to determine the classification of assessment results based on the abnormal wear risk index and the preset critical risk threshold.

[0012] The closed-loop control module is used to respond to the evaluation result classification. When the evaluation result classification is Level 2 warning, an active suppression strategy is executed.

[0013] Preferably, the first processing module is used to quantify latent perturbation features based on the perturbation signal, including:

[0014] Calculate the harmonic injection energy based on the energy spectral density of electromagnetic harmonics;

[0015] Calculate the acoustic emission spectral entropy based on the original acoustic emission spectrum of the stress wave signal;

[0016] The discharge power of the micro-region is calculated based on the total discharge energy, the total number of pulses, and the statistical time of the micro-discharge pulses.

[0017] Preferably, the second processing module is used to identify core chemical parameters, including:

[0018] Based on the modified Arrhenius equation, combined with real-time oil temperature and micro-area discharge power, the rate constant of metal saponification reaction is identified.

[0019] Based on an empirical degradation model, and combining the metal saponification reaction rate constant with harmonic injection energy, the electrophoretic mobility of colloidal clusters is identified.

[0020] Preferably, the third processing module is used to determine non-Newtonian exponents, including:

[0021] Based on a non-Newtonian rheological coupling model, the non-Newtonian index is determined by coupling the metal saponification reaction rate constant, the electrophoretic mobility of colloidal clusters, and the acoustic emission spectrum entropy.

[0022] Preferably, the working condition calculation module is used to calculate the shear rate, including:

[0023] The shear rate is determined by multiplying the gear angular velocity by the gear geometric factor according to the standard gear kinematics formula.

[0024] Preferably, the risk calculation module is used to calculate the abnormal wear risk index, including:

[0025] Calculate the difference between the non-Newtonian exponent and the value one to determine the degree of shear thickening;

[0026] The abnormal wear risk index is determined by multiplying the degree of shear thickening by the shear rate.

[0027] Preferably, the status assessment module is used to determine the classification of the assessment results, including:

[0028] When the abnormal wear risk index is less than or equal to zero, the assessment result is classified as a safe state.

[0029] When the abnormal wear risk index is greater than zero and less than or equal to the critical risk threshold, the assessment result is classified as a Level 1 warning.

[0030] When the abnormal wear risk index is greater than the critical risk threshold, the assessment result is classified as a Level II early warning.

[0031] Preferably, the active inhibition strategy includes:

[0032] Send commands to the wind turbine's main control system to reduce operating power in order to suppress the shear rate;

[0033] The active filter of the pitch system power supply is activated to suppress harmonic injection energy.

[0034] Start the backup oil circulation loop and inject antistatic additives to suppress micro-area discharge power.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] 1. This invention can deeply perceive the root causes of abnormal wear; by real-time monitoring of electromagnetic harmonics, stress waves and micro-discharge pulses, and quantifying them into latent perturbation characteristics such as harmonic injection energy, acoustic emission spectrum entropy and micro-area discharge power; and then, combined with models such as the modified Arrhenius equation, it has for the first time realized the online identification of core chemical parameters such as the metal saponification reaction rate constant, solving the problem that traditional technologies cannot obtain microscopic chemical kinetic parameters, and laying the foundation for accurate prediction;

[0037] 2. This invention accurately predicts the macroscopic rheological phase transition of lubricating oil and quantifies the risk of anomalous wear. By constructing a non-Newtonian rheological coupling model, online identified microscopic chemical parameters are coupled with mechanical perturbations to determine a non-Newtonian index that accurately characterizes the transition of lubricating oil from shear thinning to shear thickening. Based on this, an anomalous wear risk index is constructed, whose definition is completely consistent with the physical mechanism, successfully quantifying the phenomenon that higher flow rates lead to faster wear, thus solving the technical problem of traditional model failure.

[0038] 3. This invention constructs a complete technical chain from multi-physics perturbation monitoring, cross-scale feature identification, electrochemical-rheological coupling modeling to risk calculation; the system innovatively collects perturbation signals such as electromagnetic harmonics and stress waves, quantifies implicit features, identifies core chemical parameters, and finally predicts non-Newtonian exponents through coupling models, calculating a risk index that is highly consistent with the physical mechanism; it solves the technical problem that existing technologies cannot identify shear thickening phase transitions and anomalous wear caused by electro-mechanical-chemical coupling online;

[0039] 4. This invention provides a closed-loop active suppression strategy to break the positive feedback of faults, which significantly improves the reliability of equipment operation. The system performs graded early warning based on the abnormal wear risk index. Once the second-level early warning is triggered, it will coordinate the execution of active suppression strategies such as reducing operating power, activating active filters, and injecting antistatic additives. This strategy starts from the root cause and simultaneously suppresses mechanical stress, electromagnetic disturbances, and chemical catalysis, effectively breaking the electro-chemical positive feedback loop that leads to abnormal wear and avoiding catastrophic failures. Attached Figure Description

[0040] The present invention will be further explained below with reference to the accompanying drawings and embodiments:

[0041] Figure 1 This is a structural diagram of the system of the present invention. Detailed Implementation

[0042] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to specific embodiments.

[0043] Example 1:

[0044] Please see Figure 1 An online lubricating oil quality monitoring system, comprising:

[0045] The data acquisition module is used to acquire disturbance signals and operating parameters of the main gearbox of a large wind turbine. The disturbance signals include electromagnetic harmonics, stress wave signals and micro-discharge pulses; the operating parameters include real-time oil temperature and gear angular velocity.

[0046] The first processing module is used to quantify the latent perturbation characteristics based on the perturbation signal. The latent perturbation characteristics include: harmonic injection energy, acoustic emission spectrum entropy, and micro-area discharge power.

[0047] The second processing module is used to identify core chemical parameters based on micro-area discharge power, harmonic injection energy and real-time oil temperature. The core chemical parameters include: metal saponification reaction rate constant and colloidal cluster electrophoretic mobility.

[0048] The third processing module is used to predict macroscopic rheological phase transitions and determine non-Newtonian exponents based on core chemical parameters and acoustic emission spectrum entropy.

[0049] The working condition calculation module is used to calculate the shear rate based on the gear angular velocity and preset gear geometric factors;

[0050] The risk calculation module is used to calculate the abnormal wear risk index by combining shear rate and non-Newtonian exponent.

[0051] The condition assessment module is used to determine the classification of assessment results based on the abnormal wear risk index and the preset critical risk threshold.

[0052] The closed-loop control module is used to respond to the evaluation result classification. When the evaluation result classification is Level 2 warning, an active suppression strategy is executed.

[0053] The system provided in this embodiment is an online lubricating oil quality monitoring system. This system is specifically designed to solve the abnormal wear problem of large wind turbine main gearboxes, which is caused by the shear thickening phase change of lubricating oil due to the high flow rate and fast wear under complex electromechanical-chemical coupling environment. The system constructs a complete technical link from multi-physics field disturbance monitoring, cross-scale feature identification, electrochemical-rheological coupling modeling, abnormal wear risk calculation to closed-loop active suppression.

[0054] The system includes a data acquisition module, the purpose of which is to comprehensively and in real-time capture multi-source physical field data reflecting the electro-chemical-mechanical coupling state of the gearbox. In this embodiment, the module is used to acquire disturbance signals and operating parameters of the main gearbox of a large wind turbine. Specifically, the disturbance signals refer to unsteady physical field excitations, including: electromagnetic harmonics monitored in real-time by a high-frequency electromagnetic interference (EMI) antenna deployed outside the gearbox housing, for example, in a specific frequency band of 10-50kHz; stress wave signals acquired by a broadband acoustic emission (AE) sensor installed near the gearbox; and micro-discharge pulses monitored by a high-voltage micro-discharge sensor integrated in the oil circulation loop. The operating parameters refer to quasi-static parameters characterizing the system's operating state, including real-time oil temperature acquired by thermocouples or similar sensors. The gear angular velocity obtained by the speed sensor ;

[0055] The system includes a first processing module, the purpose of which is to reduce the dimensionality and extract features from the acquired raw, high-dimensional disturbance signal, quantifying it into latent disturbance features that can characterize the core causes of system instability. This module quantifies the latent disturbance features based on the disturbance signal. In this embodiment, the latent disturbance features are necessary inputs for subsequent coupling modeling, specifically including: harmonic injection energy characterizing the intensity of electromagnetic disturbance. Acoustic emission spectral entropy characterizing the mechanical stirring effect And the micro-area discharge power characterizing the intensity of electrochemical catalysis inside the oil. ;

[0056] In addition, the system includes a second processing module, the purpose of which is to construct an electrochemical positive feedback coupling model, utilizing observable physical field characteristics ( ) and working conditions This module dynamically identifies the core chemical parameters of oil under microscopic conditions; it is used to identify the core chemical parameters of oil under microscopic conditions based on micro-area discharge power. Harmonic injection energy and real-time oil temperature Identify core chemical parameters; these parameters serve as a bridge connecting microscopic chemical reactions and macroscopic rheological properties, and in this embodiment include: the metal saponification reaction rate constant. Electrophoretic mobility of colloidal clusters ;

[0057] The system includes a third processing module, the purpose of which is to establish a cross-scale coupled model to incorporate microscopic chemical kinetic parameters ( ) and mesoscopic mechanical disturbances Phase fusion to predict the macroscopic rheological phase transition of lubricating oils; this module is used to determine the phase transition based on core chemical parameters ( ) and acoustic emission spectrum entropy Predicting macroscopic rheological phase transitions and determining non-Newtonian exponents. ; It is a key dimensionless parameter, and the degree to which it deviates from the reference value of 1 characterizes the transformation of lubricating oil from shear thinning to shear thickening.

[0058] Meanwhile, the system includes a working condition calculation module, the purpose of which is to convert measurable macroscopic mechanical motion parameters into microscopic physical effects experienced by the lubricating oil in the meshing zone; this module is used to calculate the gear angular velocity based on real-time acquisition. With preset gear geometry factors Calculate the shear rate ; It is a dimensionless constant, which is determined in advance based on the geometric parameters of the gearbox design drawings;

[0059] The system also includes a risk calculation module, designed to quantify a specific type of anomalous wear risk arising from an electrochemically self-organized critical state, which is invisible in traditional lubrication models; this module is used to incorporate the non-Newtonian index determined by the third processing module. The shear rate calculated by the working condition solution module Calculate the abnormal wear risk index This index is specifically designed to characterize the anomaly that higher flow rates lead to faster wear.

[0060] The system also includes a status assessment module, the purpose of which is to evaluate continuous risk indices. Mapped to discrete, decision-guiding early warning levels; this module is used based on the abnormal wear risk index. Compared with the preset critical risk threshold Determine the classification of the assessment results; The data is not derived from empirical settings, but rather from bench tests, which calibrated the parameters that lead to a significant reduction in the evolution cycle of micropitting on tooth surfaces. The value has a clear physical meaning;

[0061] The system includes a closed-loop control module designed to execute counterintuitive proactive suppression strategies based on the assessed risk level, thereby breaking the electrochemical positive feedback loop that leads to anomalous wear at its source. This module responds to the assessment result classification; when the assessment result is classified as a Level 2 warning, i.e., a risk index... At that time, an active suppression strategy is implemented; this strategy differs from the traditional enhanced cooling, and aims to suppress mechanical stress, electromagnetic disturbance and chemical catalysis simultaneously;

[0062] This embodiment, through the collaborative work of the aforementioned modules, constructs a complete technical chain from multi-physics disturbance signal acquisition, cross-scale implicit feature quantification, electro-chemical-rheological coupling modeling, to anomalous wear risk assessment and closed-loop suppression. It solves the major technical problem that existing technologies cannot identify the shear thickening phase transition caused by electrochemical self-organized critical state online, and thus cannot provide early warning and control of the anomalous wear mode where higher flow rates lead to faster wear. This system can assess the true rheological state of lubricating oil in real time and accurately, and break the positive feedback loop of the fault through an active suppression strategy before a catastrophic failure occurs, greatly improving the operational reliability and maintenance predictability of the main gearbox of large wind turbines.

[0063] Example 2:

[0064] The first processing module is used to quantify the latent perturbation characteristics based on the perturbation signal, including:

[0065] Calculate the harmonic injection energy based on the energy spectral density of electromagnetic harmonics;

[0066] Calculate the acoustic emission spectral entropy based on the original acoustic emission spectrum of the stress wave signal;

[0067] The discharge power of the micro-region is calculated based on the total discharge energy, the total number of pulses, and the statistical time of the micro-discharge pulses.

[0068] This embodiment provides a more specific implementation method for how the first processing module quantizes the latent perturbation features based on the perturbation signal;

[0069] This module calculates harmonic injection energy based on the energy spectral density of electromagnetic harmonics. To quantify the energy input of electromagnetic harmonics in a specific frequency band to the oil system, this embodiment introduces harmonic injection energy. The calculation formula is as follows:

[0070]

[0071] in: Harmonic injection energy, unit: joules (J), calculated using this formula; Energy spectral density (ESD) of electromagnetic harmonic signals, in joules per hertz (J / Hz), is calculated from the raw electromagnetic harmonic signals acquired by the data acquisition module using signal processing methods such as standard Fourier transform. : Integral frequency lower and upper limits, unit: Hertz (Hz), preset according to the characteristics of the EMI signal of the wind turbine pitch system, for example Through this calculation, this module transforms the time-domain EMI signal into a scalar characterizing the intensity of the disturbance source. This is to facilitate the subsequent identification of the electrophoretic mobility of colloidal clusters. It provides the crucial electric field disturbance input;

[0072] This module calculates the acoustic emission spectral entropy based on the original acoustic emission spectrum of the stress wave signal. To quantify the spectral complexity of the gear meshing stress field, i.e., the intensity of the mechanical stirring effect, this embodiment introduces the acoustic emission spectral entropy based on the Shannon information entropy definition. The calculation formula is as follows:

[0073]

[0074] in, Acoustic emission spectrum entropy, a dimensionless parameter, is calculated using this formula;

[0075] Represents the first in the spectrum One frequency component;

[0076] The first digit in the original acoustic emission spectrum The proportion or probability of the energy of the first frequency component to the total energy is dimensionless and is obtained by normalizing the stress wave signal acquired by the data acquisition module after spectral analysis. Specifically, it is calculated by... The ratio of the energy of each frequency component to the total energy of the spectrum determines the... value; The magnitude of the value directly reflects the degree of disorder in the stress field; high The value represents a wide and chaotic distribution of the stress field spectrum, corresponding to a strong mechanical stirring effect, which helps to dissociate colloidal clusters generated by electrochemical reactions; this parameter is a predictor of the non-Newtonian exponent. Key factors for inhibiting shear thickening;

[0077] This module calculates the micro-area discharge power based on the total discharge energy, total number of pulses, and statistical time of the micro-discharge pulses. To simultaneously characterize the frequency and intensity of micro-discharges within the oil and accurately assess their catalytic acceleration effect on chemical reactions, this embodiment introduces the micro-area discharge power. The calculation formula is as follows:

[0078]

[0079] in, The average power of the micro-area discharge, in watts (W) or joules per second (J / s), is calculated using this formula. The total discharge energy accumulated within the statistical time window, in joules (J), determined by the micro-discharge sensor. The cumulative value of the energy of all micro-discharge pulses within the time period; Statistical time window, unit: seconds (s), preset calculation period, e.g., 1 second; this formula is computationally equivalent to the average single discharge energy. Multiply by the discharge frequency ,in The total number of pulses; using Rather than frequency or energy alone, it can more accurately characterize the overall catalytic contribution of micro-discharge to the metal saponification reaction;

[0080] This embodiment transforms the elusive original physical field signal into three implicit perturbation features with clear physical meaning that can be used for subsequent coupling modeling through the three specific quantization models described above. This quantification method ensures the accuracy and interpretability of subsequent chemical parameter identification and rheological phase change prediction, and is the foundation for realizing cross-scale modeling.

[0081] Example 3:

[0082] The second processing module is used to identify core chemical parameters, including:

[0083] Based on the modified Arrhenius equation, combined with real-time oil temperature and micro-area discharge power, the rate constant of metal saponification reaction is identified.

[0084] Based on an empirical degradation model, and combining the metal saponification reaction rate constant with harmonic injection energy, the electrophoretic mobility of colloidal clusters is identified.

[0085] This embodiment provides a more specific implementation method for how the second processing module identifies core chemical parameters;

[0086] To identify the rate constant of metal saponification reactions, this module uses a modified Arrhenius equation, combined with real-time oil temperature and micro-area discharge power, to identify the rate constant of metal saponification reactions; and to identify the rate constant of metal saponification reactions accelerated under electric field catalysis. This embodiment employs a modified Arrhenius equation, which innovatively introduces a catalytic term for the micro-area discharge power. Its identification model is as follows:

[0087]

[0088] in, Rate constant for metal saponification reaction, unit: It is identified by this model; Basic reaction rate, unit: ; Activation energy of reaction, unit: J / mol; Ideal gas constant; Real-time oil temperature, unit: Kelvin (K), acquired in real time by the data acquisition module; Micro-area discharge power, unit: W or J / s, calculated by the first processing module; Discharge catalysis weighting coefficient, unit: To ensure The unit of the item is The first term of the model It is the classical thermodynamic fundamental reaction rate, the second term. This is a correction term introduced in this invention for the micro-area discharge catalytic effect; These are all material-related inherent weighting coefficients; to adhere to the principle of variable independence in the parameter calibration process, these coefficients were determined through offline calibration experiments: a calibration dataset containing multiple working conditions was constructed. ,in Representing the The calibration temperature of the group experiment, Representing the The calibrated discharge power applied in the group experiment, This represents the rate constant of the metal saponification reaction observed experimentally under the calibration conditions. Based on this dataset, regression analysis methods such as nonlinear least squares are used to... The model is fitted to identify The optimal value was found; the model accurately decouples the temperature effect and the electrocatalytic effect, thus achieving the optimal value. The identification results are more consistent with the reality of complex working conditions; it should be noted that the linear catalytic term here... It is an approximate description of the main effects within a specific operating range. Under extreme discharge conditions, its catalytic effect may exhibit nonlinear saturation characteristics.

[0089] To identify the electrophoretic mobility of colloidal clusters, this module uses an empirical degradation model, combined with the metal saponification reaction rate constant and harmonic injection energy, to identify the electrophoretic mobility of colloidal clusters. Furthermore, it aims to identify the electrophoretic mobility of colloidal clusters degraded due to saponification product contamination and electric field perturbation. This embodiment employs an empirical degradation model calibrated based on experimental data, and its identification model is as follows:

[0090]

[0091] in, Electrophoretic mobility of colloidal clusters, unit: m² / (V·s), obtained from this model; : Basic migration rate of clean oil, unit: m² / (V·s); Rate constant for metal saponification reaction, unit: This was identified in the previous step of this module; Harmonic injection energy, unit: J, calculated by the first processing module; : Influence coefficient of saponification products, unit: s; Harmonic injection influence coefficient, unit: The denominator of this model It characterizes the degree of degradation in mobility; The larger the value, the faster the saponification products are generated and the more serious the pollution. The larger the value, the stronger the electric field disturbance; both will lead to... reduce; These are also intrinsic parameters determined through offline calibration experiments: construct a calibration dataset. ,in Representing the The known saponification contamination levels in the group of experiments correspond to specific rate constants. This represents the applied calibration harmonic injection energy. This represents the electrophoretic mobility observed in the experiment; based on this dataset, [the data is analyzed / conducted]. The model is fitted and identified The value of the chemical parameter; this model successfully converts the chemical parameter and electrical disturbance parameters Unify and share a common representation Dynamic changes;

[0092] This model simplifies the saponification contamination effect and the harmonic injection effect to a linear superposition in the denominator, reflecting the degradation relationship under the first-order approximation. In actual physical processes, these two effects may involve complex coupling, but this model is sufficient to characterize the key trends leading to rheological phase transitions.

[0093] This embodiment establishes for the first time the perturbation of macroscopic physical fields ( From microscopic core chemical parameters ( The dynamic mapping relationship of the oil is solved, which solves the problem that traditional oil monitoring cannot obtain key chemical kinetic parameters online, and provides the necessary chemical input, which has been corrected by the physical model, for subsequent accurate prediction of macroscopic rheological phase changes.

[0094] Example 4:

[0095] The third processing module is used to determine non-Newtonian exponents, including:

[0096] Based on a non-Newtonian rheological coupling model, the non-Newtonian index is determined by coupling the metal saponification reaction rate constant, the electrophoretic mobility of colloidal clusters, and the acoustic emission spectrum entropy.

[0097] This embodiment provides a more specific implementation method for how the third processing module determines non-Newtonian exponents;

[0098] This module, based on a non-Newtonian rheological coupling model, couples the metal saponification reaction rate constant, the electrophoretic mobility of colloidal clusters, and the acoustic emission spectral entropy to determine the non-Newtonian index. To predict the macroscopic rheological phase transition of lubricating oil from shear thinning to shear thickening, this embodiment constructs a non-Newtonian rheological coupling model based on an electrochemical positive feedback mechanism to determine the non-Newtonian index. Its prediction model is:

[0099]

[0100] in, Non-Newtonian exponents, dimensionless, determined by this model, and the benchmark value for Newtonian fluids; Rate constant for metal saponification reaction, unit: It is identified by the second processing module; : Electrophoretic mobility of colloidal clusters, unit: m² / (V·s), obtained by the second processing module; Acoustic emission spectrum entropy, dimensionless, calculated by the first processing module; Rheological coupling coefficient, unit: m² / V; Basic mechanical mixing factor, dimensionless;

[0101] The second term in the model is the deviation term; in the molecule The term characterizes the strength of the electrochemical positive feedback, i.e., the rate of saponification product formation. With colloidal migration / removal rate The imbalance between them; when Much larger When products accumulate, this term increases; the denominator... The terms collectively characterize the deagglomeration effect of mechanical stirring; among which This reflects the real-time changes in stirring intensity; the higher the value, the more effectively it suppresses agglomeration. As a fundamental mechanical stirring factor, it not only characterizes the inherent minimum stirring effect of the system, but also ensures the model's stability. Completeness of physical meaning and numerical robustness when approaching zero;

[0102] and It is an intrinsic factor characterizing rheological coupling properties, and its source is: a calibration dataset constructed in a high-shear rheometer. The first three represent the first three. Calibration parameters applied simultaneously in the group of experiments to simulate electrochemical states and mechanical stirring, for example, through the simulation of specific chemical additions. Imbalance, simulated by ultrasonic excitation effect, This represents the non-Newtonian exponents observed experimentally under the calibration conditions; based on this dataset, The model is fitted and identified and The value;

[0103] This model uses a simplified proportional relationship to characterize the competition between electrochemical and mechanical effects, which clearly reveals the core mechanism of anomalous wear. Future research could introduce nonlinear functions based on broader experimental data to further improve the model's predictive accuracy across the entire operating range.

[0104] This embodiment creatively integrates microscopic chemical dynamics through this non-Newtonian rheological coupling model. With mesoscopic mechanical disturbance Unified under a single framework for predicting macro rheological indices When electrochemical positive feedback It dominated and overwhelmed mechanical mixing. hour, A value greater than 1 indicates shear thickening in the system; this model is the core of diagnosing anomalous wear, and its output... The value is the direct basis for subsequent risk calculations.

[0105] Example 5:

[0106] The working condition calculation module is used to calculate the shear rate, including:

[0107] The shear rate is determined by multiplying the gear angular velocity by the gear geometric factor according to the standard gear kinematics formula.

[0108] This embodiment provides a more specific implementation method for how the working condition calculation module calculates the shear rate;

[0109] This module determines the shear rate by multiplying the gear angular velocity by the gear geometric factor according to the standard gear kinematics formula; in order to obtain the shear rate of the lubricating oil in the gear meshing elastohydrodynamic lubrication zone. In this embodiment, the standard gear kinematics formula is used for calculation. The calculation formula is as follows:

[0110]

[0111] in, Shear rate, unit: It is calculated using this formula; Gear geometric factor, dimensionless, is a preset value determined based on the geometric parameters of the gearbox design drawings. Gear angular velocity, unit: rad / s or The data is collected in real time by the data acquisition module; this formula incorporates easily measurable macroscopic operating parameters. With inherent equipment structural parameters By combining these methods, the necessary solutions for calculating abnormal wear risk were accurately obtained. value;

[0112] This embodiment uses a simple and physically clear kinematic formula to achieve control over the key operating parameter, namely shear rate. The real-time and accurate calculation provides the necessary operating condition input for the subsequent risk calculation module, ensuring the accuracy of the risk index. It can accurately reflect the abnormal wear intensity under current operating conditions.

[0113] Example 6:

[0114] The risk calculation module is used to calculate the abnormal wear risk index, including:

[0115] Calculate the difference between the non-Newtonian exponent and the value one to determine the degree of shear thickening;

[0116] The abnormal wear risk index is determined by multiplying the degree of shear thickening by the shear rate.

[0117] This embodiment provides a more specific implementation method for how the risk calculation module calculates the abnormal wear risk index;

[0118] The difference between the non-Newtonian exponent and the numerical value is calculated to determine the degree of shear thickening; the degree of shear thickening is defined as... ; The non-Newtonian exponent is determined by the third processing module; when When the value is zero or negative, it indicates that the lubricating oil is in a Newtonian or shear-thinned state, with no risk of abnormal wear; when... When the value is positive, it indicates that the lubricating oil has entered the shear thickening phase transition state, and its magnitude reflects the severity of thickening.

[0119] Multiplying the degree of shear thickening by the shear rate determines the anomalous wear risk index; to quantify the higher the flow rate... The abnormal phenomenon of larger wear rates leading to faster wear is addressed in this embodiment by defining an abnormal wear risk index. The calculation formula is as follows:

[0120]

[0121] in, Abnormal wear risk index, unit: This can be understood as the rate of occurrence of abnormal wear, which is calculated by this formula; : Shear thickening degree, dimensionless, calculated in the previous step of this module; Shear rate, unit: The result is calculated by the working condition solution module; the logic of this model is completely consistent with the physical mechanism of anomalous wear: only when the system is in a shear thickening state, hour, Only then can it be positive; and under this premise, the shear rate The larger the fan speed, the higher the risk index. The higher the value, the more perfectly the anomalous characteristics are reproduced;

[0122] This embodiment defines... The degree of shear thickening and its relationship with the shear rate. By multiplying, a risk index was constructed that can accurately quantify the intensity of anomalous wear under the critical state of electrochemical self-organization. The index Provided with working conditions The highly correlated diagnostic indicators completely solve the technical problem of traditional lubrication models failing under such working conditions.

[0123] Example 7:

[0124] The status assessment module is used to determine the classification of assessment results, including:

[0125] When the abnormal wear risk index is less than or equal to zero, the assessment result is classified as a safe state.

[0126] When the abnormal wear risk index is greater than zero and less than or equal to the critical risk threshold, the assessment result is classified as a Level 1 warning.

[0127] When the abnormal wear risk index is greater than the critical risk threshold, the assessment result is classified as a Level II early warning.

[0128] This embodiment provides a more specific implementation method for how the status assessment module determines the classification of assessment results;

[0129] This module is based on the abnormal wear risk index. and critical risk threshold Comparison:

[0130] When the abnormal wear risk index is less than or equal to zero, that is The assessment result was classified as a safe state; this state corresponds to a non-Newtonian exponent. This indicates that the lubricating oil exhibits Newtonian or shear-thinning characteristics, the system is in a safe operating range, and the traditional lubrication model is effective.

[0131] When the abnormal wear risk index is greater than zero and less than or equal to the critical risk threshold, that is... The assessment result was determined to be classified as a Level 1 warning; this state corresponds to a non-Newtonian exponent. This indicates that the lubricating oil has exhibited slight shear thickening characteristics, the traditional lubrication model has begun to fail, and the system is at risk of electrochemical instability. This needs attention but does not require immediate strong intervention.

[0132] When the abnormal wear risk index is greater than the critical risk threshold, that is... The assessment result was classified as a Level II warning; this state indicates that the lubricating oil is in a strong shear thickening phase transition state, has entered the electrochemical self-organization criticality, and abnormal wear is occurring with intensity exceeding the acceptable critical point. The closed-loop control module must be triggered immediately to execute the active suppression strategy.

[0133] This embodiment sets three distinct levels: safety status, Level 1 warning, and Level 2 warning, and associates them with risk indices that have clear physical meaning. and calibration threshold This hierarchical approach provides a clear, accurate, and unambiguous basis for subsequent closed-loop control, avoiding a black-and-white, crude diagnosis and enabling refined and progressive management of abnormal wear risks.

[0134] Example 8:

[0135] Active inhibition strategies include:

[0136] Send commands to the wind turbine's main control system to reduce operating power in order to suppress the shear rate;

[0137] The active filter of the pitch system power supply is activated to suppress harmonic injection energy.

[0138] Start the backup oil circulation loop and inject antistatic additives to suppress micro-area discharge power.

[0139] This embodiment provides a more specific implementation of the active suppression strategy executed by the closed-loop control module when responding to a secondary warning; the core of this strategy is to break the electrochemical positive feedback loop, rather than the traditional method of enhancing cooling.

[0140] Active inhibition strategies include the following coordinated actions:

[0141] A command is sent to the wind turbine's main control system to reduce operating power in order to suppress the shear rate. This command can be sent via an industry-standard communication protocol, and the command content is a specific derating percentage or target power value. The purpose of this action is to suppress operating parameters. According to the risk index formula ,reduce In other words, reducing the amount of risk directly and quickly lowers the risk index. The most effective means;

[0142] The active filter of the pitch system power supply is activated to suppress harmonic injection energy; this action aims to suppress the input of the first processing module. By activating active filters targeting specific frequency bands, electromagnetic disturbances are reduced at the source; reducing... Will improve This reduces The term leads to non-Newtonian exponents. Decrease, inhibiting shear thickening;

[0143] The backup oil circulation loop is activated and an antistatic additive is injected to suppress micro-area discharge power; this action aims to suppress the input to the first processing module. By injecting antistatic additives to improve the conductivity of the oil, static electricity accumulation and micro-discharge can be suppressed, thereby reducing... ;reduce Will decrease This reduces This term also leads to non-Newtonian exponents. The decline breaks the positive feedback loop.

[0144] This embodiment provides a counterintuitive but mechanistically clear active suppression strategy; it does not passively respond to faults, but rather actively and multidimensionally ( Simultaneously intervening in three core factors leading to anomalous wear: mechanical conditions, electromagnetic disturbances, and electrochemical catalysis; through this synergistic inhibition, this strategy can effectively break the electrochemical positive feedback loop, shifting the system from... The shear thickening state is pulled back to This ensures a safe state, thereby preventing micro-pitting on the tooth surface from evolving into macro-scraping and achieving effective control over failure modes that would otherwise fail with traditional strategies.

[0145] Example 9:

[0146] This embodiment provides a more specific implementation of the offline parameter calibration method necessary for the operation of the online monitoring system; the calibration process is completed once before system deployment, ensuring the accuracy and reliability of the online model input.

[0147] 1. Parameters of the metal saponification model calibration

[0148] This calibration depends on constructing a calibration dataset. and fit the model The specific steps are as follows:

[0149] Experimental setup: Construct a laboratory testing platform, including: a sealed oil bath with a heating element for precise control of the oil temperature. A pair of electrodes connected to a high-voltage pulse power supply are used to simulate controllable micro-area discharge power. ; and a port for oil sampling;

[0150] Data acquisition: Design an operating condition matrix covering different temperatures and discharge powers. At each operating point... The experiment was conducted, and oil samples were extracted periodically.

[0151] Observation rate Measurement: The concentration of metal soaps in each oil sample was measured using Fourier transform infrared spectroscopy, specifically by tracking the characteristic absorption peaks of carboxylate groups. The rate constant of the metal saponification reaction was calculated based on the rate of change of this concentration over time. ;

[0152] Parameter fitting: Based on the collected dataset, a nonlinear least squares regression algorithm is used in software environments such as MATLAB or Python to fit the observed data with the model equation, thereby identifying the pre-exponential factors. Activation energy of reaction and discharge catalysis weighting coefficient The optimal value;

[0153] 2. Electrophoretic mobility model parameters calibration

[0154] This calibration utilizes a dataset. Fitting Model The specific steps are as follows:

[0155] Sample preparation: Using the conditions of the previous experiment, the oil samples were artificially accelerated to prepare a series of samples with different and known saponification levels.

[0156] Experimental setup: A micro-electric pool equipped with a laser Doppler velocimeter was used. The electric pool was placed inside a Helmholtz coil, and a uniform and quantifiable harmonic magnetic field was applied through a signal generator to simulate harmonic injection energy. ;

[0157] Observational mobility Measurement: for each with a known The sample and the given A DC electric field is applied to both ends of the electrophoretic pool; the velocity of colloidal clusters in the oil is measured using LDV; based on this velocity and the known DC electric field strength, the electrophoretic mobility of the colloidal clusters is calculated. ;

[0158] Parameter fitting: Using the collected dataset, the model is fitted with a curve fitting tool to identify the basic mobility. Influence coefficient of saponification products Harmonic Injection Influence Coefficient The value of .

[0159] 3. Parameters of non-Newtonian models calibration

[0160] This calibration utilizes a dataset. Fitting Model The specific steps are as follows:

[0161] Operating condition simulation: High shear rheometer used; to simulate the intensity of electro-chemical positive feedback. A specific concentration of oleic acid and metal nanoparticles was added to fresh oil samples to induce controllable cluster formation. This was done to simulate the mechanical stirring effect. The ultrasonic transducer is coupled to the sample chamber of the rheometer, and the acoustic emission spectral entropy is calculated by analyzing the spectrum of the ultrasonic signal.

[0162] Observation Index Measurement: Under each simulated operating condition, the rheometer was driven to run at a series of shear rates, and the corresponding shear stress was measured. This was achieved by using a power-law model (…). The non-Newtonian index under this operating condition is determined by fitting the measured flow curve data with the data. ;

[0163] Parameter fitting: All experimental data points were summarized, and nonlinear regression analysis was performed to identify the rheological coupling coefficient. and basic mechanical mixing factors The value;

[0164] This embodiment clarifies that the constants and coefficients used in the online system are not preset based on experience, but rather derived from rigorous offline experiments based on physical mechanisms, thereby ensuring the accuracy and reliability of the online monitoring model.

[0165] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. An online lubricating oil quality monitoring system, characterized in that... ,include: The data acquisition module is used to acquire disturbance signals and operating parameters of the main gearbox of a large wind turbine. The disturbance signals include electromagnetic harmonics, stress wave signals and micro-discharge pulses; the operating parameters include real-time oil temperature and gear angular velocity. The first processing module is used to quantify the latent perturbation characteristics based on the perturbation signal. The latent perturbation characteristics include: harmonic injection energy, acoustic emission spectrum entropy, and micro-area discharge power. The second processing module is used to identify core chemical parameters based on micro-area discharge power, harmonic injection energy and real-time oil temperature. The core chemical parameters include: metal saponification reaction rate constant and colloidal cluster electrophoretic mobility. The third processing module is used to predict macroscopic rheological phase transitions and determine non-Newtonian exponents based on core chemical parameters and acoustic emission spectrum entropy. The working condition calculation module is used to calculate the shear rate based on the gear angular velocity and preset gear geometric factors; The risk calculation module is used to calculate the abnormal wear risk index by combining shear rate and non-Newtonian exponent. The condition assessment module is used to determine the classification of assessment results based on the abnormal wear risk index and the preset critical risk threshold. The closed-loop control module is used to respond to the evaluation result classification. When the evaluation result classification is a level 2 warning, an active suppression strategy is executed. The risk calculation module is used to calculate the abnormal wear risk index, including: Calculate the difference between the non-Newtonian exponent and the value one to determine the degree of shear thickening; The abnormal wear risk index is determined by multiplying the degree of shear thickening by the shear rate.

2. The online lubricating oil quality monitoring system according to claim 1, characterized in that... The first processing module is used to quantify the latent perturbation characteristics based on the perturbation signal, including: Calculate the harmonic injection energy based on the energy spectral density of electromagnetic harmonics; Calculate the acoustic emission spectral entropy based on the original acoustic emission spectrum of the stress wave signal; The discharge power of the micro-region is calculated based on the total discharge energy, the total number of pulses, and the statistical time of the micro-discharge pulses.

3. The online lubricating oil quality monitoring system according to claim 1, characterized in that... The second processing module is used to identify core chemical parameters, including: Based on the modified Arrhenius equation, combined with real-time oil temperature and micro-area discharge power, the rate constant of metal saponification reaction is identified. Based on an empirical degradation model, and combining the metal saponification reaction rate constant with harmonic injection energy, the electrophoretic mobility of colloidal clusters is identified.

4. The online lubricating oil quality monitoring system according to claim 1, characterized in that... The third processing module is used to determine non-Newtonian exponents, including: Based on a non-Newtonian rheological coupling model, the non-Newtonian index is determined by coupling the metal saponification reaction rate constant, the electrophoretic mobility of colloidal clusters, and the acoustic emission spectrum entropy.

5. The online lubricating oil quality monitoring system according to claim 1, characterized in that... The working condition calculation module is used to calculate the shear rate, including: The shear rate is determined by multiplying the gear angular velocity by the gear geometric factor according to the standard gear kinematics formula.

6. The online lubricating oil quality monitoring system according to claim 1, characterized in that... The status assessment module is used to determine the classification of the assessment results, including: When the abnormal wear risk index is less than or equal to zero, the assessment result is classified as a safe state. When the abnormal wear risk index is greater than zero and less than or equal to the critical risk threshold, the assessment result is classified as a Level 1 warning. When the abnormal wear risk index is greater than the critical risk threshold, the assessment result is classified as a Level II early warning.

7. The online lubricating oil quality monitoring system according to claim 6, characterized in that... The active inhibition strategy includes: Send commands to the wind turbine's main control system to reduce operating power in order to suppress the shear rate; The active filter of the pitch system power supply is activated to suppress harmonic injection energy. Start the backup oil circulation loop and inject antistatic additives to suppress micro-area discharge power.

Images