Lubricating oil aging monitoring digital twin system
By reconstructing the ideal baseline state of lubricating oil using a digital twin system and comparing its topological features, the accuracy problem of lubricating oil aging monitoring under complex operating conditions is solved, enabling precise assessment of the remaining life of lubricating oil and identification of abnormal conditions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- RITUI ENERGY TECH HEBEI CO LTD
- Filing Date
- 2026-03-13
- Publication Date
- 2026-06-12
AI Technical Summary
Existing technologies cannot accurately separate reversible parameter fluctuations caused by environmental factors from irreversible aging caused by the chemical degradation of lubricating oil under complex operating conditions of varying temperature and pressure. This leads to a decrease in the accuracy of aging monitoring results and makes it difficult to distinguish between normal operating condition fluctuations and actual equipment hazards.
A digital twin system for monitoring lubricating oil aging is constructed. Real-time operating conditions and sensor data are acquired through data acquisition terminals. The digital twin simulation server reconstructs the ideal baseline state and generates the simulated aging state. By comparing the topological features of the theoretical aging residual and the actual total residual, the true aging index is generated, enabling accurate identification of non-model-based abnormal states.
It effectively isolates viscosity changes caused by temperature fluctuations, distinguishes between normal chemical degradation and sudden contamination events or sensor failures, improves the accuracy of lubricant remaining life monitoring, and avoids unnecessary downtime or catastrophic wear caused by false alarms or missed alarms.
Smart Images
Figure CN122194884A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of digital twin technology, specifically to a digital twin system for monitoring lubricating oil aging. Background Technology
[0002] Currently, lubricant condition monitoring in industrial equipment is typically based on fixed threshold comparisons or single sensor data acquisition. When equipment is under complex operating conditions with varying temperatures and pressures, existing technologies cannot accurately separate reversible parameter fluctuations caused by environmental factors from irreversible aging caused by chemical degradation. This makes it difficult to distinguish between normal operating condition fluctuations and actual equipment hazards, thus reducing the accuracy of aging monitoring results. Summary of the Invention
[0003] The purpose of this invention is to provide a digital twin system for monitoring lubricating oil aging, which solves the problems existing in the background art.
[0004] To address the aforementioned technical problems, this invention provides a digital twin system for monitoring lubricating oil aging, comprising: a data acquisition terminal, a digital twin simulation server, and an interactive feedback terminal; wherein, the data acquisition terminal is used to acquire real-time operating condition time-series data of the target device and measured data from online lubricating oil sensors, and sends the real-time operating condition time-series data and measured data from online sensors to the digital twin simulation server; the digital twin simulation server is used to reconstruct the ideal baseline state at the current moment based on the real-time operating condition time-series data and a preset static physicochemical fingerprint, and to generate a simulated aging state including theoretical degradation based on the real-time operating condition time-series data and a preset cumulative damage dynamics model; the digital twin simulation server is also used to calculate the theoretical aging residual between the simulated aging state and the ideal baseline state, and the actual total residual between the measured data from online sensors and the ideal baseline state; the digital twin simulation server is also used to perform topological feature comparison on the theoretical aging residual and the actual total residual, and, when the theoretical aging residual and the actual total residual meet preset coupling conditions, generate a true aging index and send it to the interactive feedback terminal; the interactive feedback terminal is used to display the true aging index.
[0005] Preferably, the digital twin simulation server is also used to: determine that the lubricating oil is in a non-modeled abnormal state in response to the fact that the theoretical aging residual and the actual total residual do not meet the preset coupling conditions; and send an abnormal alarm command to the interactive feedback terminal, the abnormal alarm command being used to indicate a sudden pollution event or sensor failure.
[0006] Preferably, when reconstructing the ideal baseline state at the current moment, the digital twin simulation server is specifically used to: call a preset state equation model, and use the temperature and pressure parameters in the real-time operating condition time series data as input variables; combine the basic parameters of the new oil in the static physicochemical fingerprint to calculate the theoretical physical values that the lubricating oil should present under the assumption that it is in a brand new state, as the ideal baseline state; wherein, the ideal baseline state is used to isolate the reversible influence of environmental factors on the physical parameters of the lubricating oil.
[0007] Preferably, when generating a simulated aging state containing theoretical degradation, the digital twin simulation server is specifically used to: convert real-time operating condition time-series data into aging stress factors; based on the aging stress factors, use a tribochemical evolution model containing shear sensitivity weights and oxidation sensitivity weights to calculate the theoretical viscosity degradation caused by mechanical shear and thermal oxidation as the theoretical damage amount; superimpose the theoretical damage amount onto the ideal baseline state to generate the simulated aging state; wherein, the simulated aging state is used to characterize the theoretical aging trend caused solely by the accumulation of operating conditions.
[0008] Preferably, when the digital twin simulation server converts real-time operating condition time-series data into aging stress factors, it is specifically used for: performing integral calculations on the heat load history in the real-time operating condition time-series data based on the Arrhenius equation to generate a thermal oxidation factor; and calculating the mechanical shear loss rate of polymer additives based on the rotational speed data and shear time data of the target equipment to generate a shear dilution factor.
[0009] Preferably, when performing topological feature comparison between the theoretical aging residual and the actual total residual, the digital twin simulation server specifically performs the following: within a sliding time window, it calculates the cross-correlation coefficient or dynamic time warping distance between the change patterns of the theoretical aging residual and the change patterns of the actual total residual; in response to the cross-correlation coefficient being higher than a preset similarity threshold, or the dynamic time warping distance being lower than a preset distance threshold, it determines that the theoretical aging residual and the actual total residual satisfy a preset coupling condition; in response to the cross-correlation coefficient being lower than a preset similarity threshold, or the dynamic time warping distance being higher than a preset distance threshold, it determines that the theoretical aging residual and the actual total residual do not satisfy the preset coupling condition.
[0010] Preferably, when generating the true aging index, the digital twin simulation server is specifically used to: extract the amplitude characteristics of the theoretical aging residual; and calculate the remaining service life of the lubricating oil based on the ratio of the amplitude characteristics to the preset aging limit threshold, which is then used as the true aging index.
[0011] Compared with existing technologies, this invention realizes the application of synthetic analysis in the field of oil monitoring by constructing a dual reference system of ideal baseline state and simulated aging state. It achieves the effect of separating reversible environmental influences from irreversible aging influences in a dimension-reducing manner. Compared with the fixed threshold monitoring method in existing technologies that cannot distinguish between normal operating condition fluctuations and actual equipment hazards, it can effectively separate viscosity changes caused by temperature fluctuations and solve the problem that the monitoring results of traditional methods cannot truly reflect the remaining life of lubricating oil.
[0012] This invention achieves accurate identification of non-model-based abnormal states by comparing the topological features of theoretical aging residuals and actual total residuals and using a decoupling judgment mechanism. It distinguishes between normal chemical degradation and sudden pollution events or sensor failures. Compared with existing technologies that easily misjudge external pollution as normal aging, this invention uses a digital twin as a benchmark truth value to build a defense mechanism, effectively identifying coolant leaks or sensor drift, and solving the problems of unnecessary downtime caused by false alarms or catastrophic wear caused by missed alarms. Attached Figure Description
[0013] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Figure 1 This is a logic block diagram of the system of the present invention. Detailed Implementation
[0014] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.
[0015] Please see Figure 1This invention provides a digital twin system for monitoring lubricating oil aging, comprising: a data acquisition terminal, a digital twin simulation server, and an interactive feedback terminal; wherein, the data acquisition terminal is used to acquire real-time operating condition time-series data of the target device and measured data from online lubricating oil sensors, and sends the real-time operating condition time-series data and measured data from online sensors to the digital twin simulation server; the digital twin simulation server is used to reconstruct the ideal baseline state at the current moment based on the real-time operating condition time-series data and a preset static physicochemical fingerprint, and to generate a simulated aging state including theoretical degradation based on the real-time operating condition time-series data and a preset cumulative damage dynamics model; the digital twin simulation server is also used to calculate the theoretical aging residual between the simulated aging state and the ideal baseline state, and the actual total residual between the measured data from online sensors and the ideal baseline state; the digital twin simulation server is also used to perform topological feature comparison on the theoretical aging residual and the actual total residual, and, when the theoretical aging residual and the actual total residual meet preset coupling conditions, generate a true aging index and send it to the interactive feedback terminal; the interactive feedback terminal is used to display the true aging index.
[0016] This embodiment elaborates on the core architecture and operating mechanism of the aforementioned digital twin system for monitoring lubricating oil aging. Firstly, the system synchronously acquires real-time operating condition data that directly affects the physicochemical state of the lubricating oil through a data acquisition terminal deployed on the target device side. And measured data of lubricating oil from online sensors obtained directly through immersion sensors. Based on this, the digital twin simulation server, as the core computing unit, utilizes the benchmark reconstruction engine to call preset static physical and chemical fingerprints, combined with... Reconstruct the ideal baseline state assuming the lubricating oil is in a brand new state at the current moment. This state aims to establish a zero-aging reference system that dynamically changes with operating conditions; simultaneously, the server runs the damage evolution engine in parallel, based on... Using a pre-defined cumulative damage kinetic model, the chemical degradation caused solely by historical operating conditions is calculated to generate a simulated aging state. Subsequently, the server performs a dual-track differential calculation; during this process, in order to solve... Multidimensional vectors, i.e., problems involving matching the dimensions of viscosity, dielectric constant, etc., with the model output, will be handled by the server. and Constructed as with A dimensionally consistent vector, i.e., filling the non-dynamically calculated parameter components with static fingerprint baseline values, is used to derive the theoretical aging residuals characterizing the pure aging trend. And the actual total residual, which includes actual aging and external disturbances. Furthermore, the system... and Topological feature comparison is performed. In response to the two satisfying the preset coupling conditions, that is, when it is confirmed that the changing trend of the actual observation value is statistically consistent with the theoretical deduction, the real aging index is generated and sent to the interactive feedback terminal for display. This embodiment realizes the application of synthetic analysis in the field of oil monitoring by constructing a dual reference system of ideal baseline state and simulated aging state. The system can effectively separate reversible environmental influences, such as viscosity changes caused by temperature fluctuations, from irreversible aging influences, such as deterioration caused by oxidation, in complex operating scenarios such as wind turbines or heavy hydraulic equipment. This solves the technical pain point of traditional threshold monitoring methods in being unable to distinguish between normal operating condition fluctuations and actual equipment hazards, ensuring that the monitoring results can truly reflect the remaining life of the lubricating oil. To clarify the mathematical definition of the preset coupling conditions and meet the requirement of full disclosure, this embodiment specifically points out that the coupling conditions are not abstract concepts, but are defined as... The logical combination; that is, requiring that the theoretical residual and the actual residual have similarity in form. Above the threshold or path regularization distance Below the threshold When the condition is true, statistical consistency is quantified into a computable set of mathematical inequalities.
[0017] In this embodiment, the digital twin simulation server is also used to: determine that the lubricating oil is in a non-modeled abnormal state in response to the fact that the theoretical aging residual and the actual total residual do not meet the preset coupling conditions; and send an abnormal alarm command to the interactive feedback terminal, the abnormal alarm command being used to indicate a sudden pollution event or sensor failure. This embodiment further specifies the abnormal state discrimination logic; during the continuous operation of the system, the digital twin simulation server will continuously monitor the theoretical aging residual within the sliding time window. Total residual with respect to reality The system determines that the lubricating oil is in a non-modeled abnormal state when it detects that the two do not meet the preset coupling conditions, i.e., when a significant decoupling phenomenon occurs. This state specifically refers to the state change caused by sudden, non-cumulative factors that are not included in the cumulative damage kinetic model. Subsequently, the server immediately sends an abnormal alarm command to the interactive feedback terminal, which clearly points to two high-risk scenarios: one is a sudden contamination event, such as coolant leakage leading to oil emulsification or a large intrusion of external particles. There will be violent fluctuations. The system can maintain stability; the second possibility is sensor failure, such as zero-point drift or circuit failure; the system will then trigger an audible and visual alarm or send a notification to the mobile device of the maintenance personnel. This embodiment utilizes a digital twin as a baseline truth to construct a defense mechanism against non-natural aging. In scenarios such as internal leakage in coolers or sensor failure that often occur in industrial settings, this feature can accurately identify abnormal oil parameters caused by external contamination or equipment malfunction. This avoids misjudging such sudden physical contamination as normal chemical degradation of lubricating oil, thereby greatly improving the safety and accuracy of operation and maintenance decisions and preventing unnecessary downtime due to false alarms or catastrophic wear due to missed alarms.
[0018] When reconstructing the ideal baseline state at the current moment, the digital twin simulation server is specifically used to: call the preset state equation model, and use the temperature and pressure parameters in the real-time operating condition time series data as input variables; combine the basic parameters of the new oil in the static physicochemical fingerprint to calculate the theoretical physical values that the lubricating oil should present under the assumption that it is in a brand new state, as the ideal baseline state; wherein, the ideal baseline state is used to isolate the reversible influence of environmental factors on the physical parameters of the lubricating oil. This embodiment further specifies the ideal baseline state reconstruction process, supplementing the physical acquisition path of key model parameters to meet the requirement of full disclosure; the digital twin simulation server calls the preset state equation model to achieve mathematical stripping of the reversible effects of environmental factors; firstly, the server extracts the oil tank temperature from the real-time operating condition time series data. and pipeline pressure As input variables; then, combining the pre-stored basic parameters of the new oil in the static physicochemical fingerprint, the modified Walther-MacCoull equation combined with the Barus viscosity-pressure equation is used for calculation; in order to correct the physical principle deviation of the original linear superposition model in predicting high-pressure viscosity under variable temperature conditions, that is, the viscosity-pressure effect changes exponentially with the basic viscosity rather than being nonlinearly superimposed, this embodiment adopts a step-by-step coupling model, that is, first calculates the temperature-viscosity relationship under normal pressure, and then superimposes the pressure-viscosity correction, as shown in the following formula:
[0019] Derived from calculations, its physical meaning is the ideal baseline kinematic viscosity reconstructed at the current moment, with units of... ; Derived from the dimensional analysis rules, its physical meaning is the dimensionally normalized reference viscosity, with a value of 1 and a unit of . ; The data is sourced from the data acquisition terminal and its physical meaning is the real-time oil tank temperature, measured in Kelvin (K). Derived from the physical definition, its physical meaning is a reference temperature constant, with a value of 1K; The data is sourced from the data acquisition terminal and its physical meaning is real-time pipeline pressure, measured in MPa. Derived from laboratory calibration, dimensionless, specifically through standard viscosity testing of new oil samples, a system of two equations is constructed. The inverse solution yields that v is the standard kinematic viscosity of the new oil sample at the calibration temperature T, in mm² / s. Derived from industry standards, this is a shift constant used to prevent negative values in logarithmic calculations. It has a value of 0.7 and is dimensionless. Derived from high-pressure rheological testing, its physical meaning is the viscosity-pressure coefficient, and its unit is... This parameter corrects for the increase in viscosity index caused by the reduction in the distance between fluid molecules under high pressure. Its physical necessity lies in resolving the mapping contradiction between dimensionless numerical values and dimensional physical quantities: the exponential term in the above formula The calculated value is a relative value of viscosity with respect to a unit dimension, i.e., dimensionless; if the physical quantity is directly expressed as... Equal to this value will cause the dimensions on both sides of the physical equation to be non-conserved, that is, the dimensions on the left side of the physical equation will be... The dimensions on the right side of the equation are Therefore, we introduce dimensional units. Participating in computation essentially involves constructing... The physical relationship ensures that the mathematical deduction results can be rigorously converted into kinematic viscosity values with physical units; The server will then solve the problem. As an ideal reference state The core components; it should be noted that, in order to integrate with online sensor measured data containing multiple parameters, namely viscosity, dielectric constant, moisture content, and ferrographic density. Perform consistency calculations to achieve the ideal baseline state. In practice, it is constructed as a state vector of the same dimension; the kinematic viscosity component is solved in real time by the above state equation, while parameters such as dielectric constant and moisture content, which are insensitive to temperature and pressure or have no operating condition correlation, are directly called from the new oil reference value in the static physicochemical fingerprint, such as the new oil reference moisture value. Perform filling to ensure and Strict matching in dimensions.
[0020] In this embodiment, when generating a simulated aging state containing theoretical degradation, the digital twin simulation server specifically performs the following: converting real-time operating condition time-series data into an aging stress factor; based on the aging stress factor, using a tribochemical evolution model including shear sensitivity weights and oxidation sensitivity weights, calculating the theoretical viscosity degradation caused by mechanical shear and thermal oxidation as the theoretical damage amount; superimposing the theoretical damage amount onto an ideal baseline state to generate a simulated aging state; wherein, the simulated aging state is used to characterize the theoretical aging trend caused solely by accumulated operating conditions; when converting real-time operating condition time-series data into an aging stress factor, the digital twin simulation server specifically performs the following: based on the Arrhenius equation, performing an integral calculation on the thermal load history in the real-time operating condition time-series data to generate a thermal oxidation factor; based on the target equipment's rotational speed data and shear time data, calculating the mechanical shear loss rate of the polymer additive to generate a shear dilution factor; This embodiment provides an in-depth analysis of the simulation aging state generation process and experimentally defines the physical source and computational mapping of the model coefficients to meet the requirement of sufficient disclosure.
[0021] The thermal oxidation factor calculation and parameter acquisition server, based on the Arrhenius equation, integrates the historical heat load in real-time operating data to generate the thermal oxidation factor. :
[0022] Considering that practical digital twin systems process discrete sampled data, in order to meet the requirements of programmability, this embodiment explicitly adopts the rectangular numerical integration method to transform the above continuous integral into a discrete accumulation form:
[0023] For the first Real-time oil tank temperature at each sampling point ; For the first Time step of each sampling point ; This represents the total number of sampling points up to the current time. The key model parameters are obtained in the following ways: Activation energy and Pre-exponential factor: derived from dual-temperature accelerated oxidation experiments; the specific steps are: selecting lubricating oil samples and subjecting them to two different constant high temperatures. like and like The following were performed using rotating oxygen bomb (RPVOT) tests, and the oxidation induction periods were measured as follows: and A system of equations was constructed based on the principles of chemical kinetics. This allows for the calculation of the specific oil product. And then substitute get This step ensures that the model parameters are derived from real physical tests rather than being arbitrarily preset. Ideal gas constant: takes the value of , used for thermodynamic energy calculations; The shear dilution factor calculation and parameter acquisition server is based on the principle of mechanical work and calculates the cumulative shear dilution factor. :
[0024] To clarify the meaning of each variable in the formula and its correspondence with real-time time series data, the following definitions are provided: The physical meaning is the current moment, that is, the cumulative running time calculated from the monitoring start point; The physical meaning is up to the current moment. The total number of accumulated real-time operating condition data sampling points; the right side of the formula approximates the continuous time by accumulating discrete time steps. The integration process, namely ; : Derived from real-time operating condition time series data, the physical meaning is the first time... Real-time rotational speed data of the target device collected at each sampling time point; : The sampling settings originate from the data acquisition terminal, and their physical meaning is the first The time interval between sampling points, i.e., the sampling step size; through Calculate the shear stroke within a single step; Shear degradation coefficient: derived from equipment-level geometric calibration experiments; this coefficient is an inherent property characterizing the shear strength of the gearbox mechanical structure, independent of the type of lubricating oil, aiming to decouple the equipment's geometric characteristics from the oil's physical properties; specifically, it is obtained by using a reference oil with industry-standard shear stability data, such as RL-203 standard oil, on a test bench of the same model of equipment, and recording its known shear sensitivity weight as... For example, take the standard shear loss rate of RL-203 as 0.15; run Duration, viscosity decrease rate measured Based on this reverse calculation, the following is obtained: ; For the known constant operating speed in the bench calibration experiment; this step ensures It only includes equipment factors and can be directly applied to subsequent calculations for different target oils; The tribochemical evolution model and the superposition server use the component mapping method to generate simulated aging states. For the kinematic viscosity component Nonlinear superposition is performed using the following formula:
[0025] The mapping logic for the sensitivity coefficient is as follows: Shear sensitivity weight: derived from KRL shear test DIN51350-6; the specific calculation formula is as follows:
[0026] To measure the percentage of viscosity loss of the target oil after the standard test, for example, if the loss is 15%, take 15. It needs to be normalized by dividing by 100 in the formula. The denominator is the total shear revolutions under the standard test conditions, so as to convert the laboratory index into the dimensionless coefficient of the model. The oxidation sensitivity weight is derived from the RPVOT test ASTM D2272; set This means that when the accumulated thermal oxidizing factor When the value reaches 1.0, the cumulative heat load is equal to one complete RPVOT life, and the theoretical viscosity increases due to oxidation to the reference ratio, thereby achieving a normalized mapping between laboratory life and operating life. Through the clearly defined parameter acquisition and mapping formulas described above, this embodiment ensures complete traceability from laboratory physicochemical data to digital twin model parameters, achieving model reproducibility.
[0027] When performing topological feature comparison between theoretical aging residuals and actual total residuals, the digital twin simulation server is specifically used for: calculating the cross-correlation coefficient or dynamic time warping distance between the change patterns of theoretical aging residuals and actual total residuals within a sliding time window; determining that theoretical aging residuals and actual total residuals satisfy a preset coupling condition when the cross-correlation coefficient is higher than a preset similarity threshold or the dynamic time warping distance is lower than a preset distance threshold; and determining that theoretical aging residuals and actual total residuals do not satisfy a preset coupling condition when the cross-correlation coefficient is lower than a preset similarity threshold or the dynamic time warping distance is higher than a preset distance threshold.
[0028] This embodiment is an algorithmic concretization of the topological feature comparison step; the server first defines a sliding time window. And perform feature dimension projection extraction within that window; because and As the system is a multidimensional vector sequence, directly calculating the correlation would introduce irrelevant variables, such as moisture noise. Therefore, the system focuses on the kinematic viscosity dimension, which is strongly coupled with the aforementioned cumulative damage kinetic model, from which the theoretical aging residual sequence is extracted. and the actual total residual sequence ;Right now Next, the system calculates the synchronicity of the changes in both, using cross-correlation coefficients. or dynamic time warping distance As a quantitative indicator, the cross-correlation coefficient is calculated using the following formula:
[0029] The average value of the theoretical aging residual sequence within the sliding time window; The average value of the total residual sequence within the sliding time window; Derived from calculation, its physical meaning is the morphological similarity between two sequences, and it is dimensionless. : Derived from the viscosity residual scalar sequence after dimensional projection, its physical meaning is the th The residual values of each sampling point; : Derived from the sliding window sampling count, its physical meaning is the time step index within the window, and its value range is arrive ,in This represents the total length of the sequence within the sliding time window; For dynamic time warping distance The system first constructs a distance matrix, and then calculates the cumulative distance matrix based on the idea of dynamic programming. The recursive formula is:
[0030] Derived from sequence index, its physical meaning is sequence. and The time point index in the text has a range of values. arrive Used to iterate and calculate the alignment cost of two time series at different time steps; Originating from computation, its physical meaning is sequence. In the Time and Sequence In the Local distance at any given moment; Finally, calculate the normalized path distance. ;in The length of the sequence within the sliding time window. The sum of steps in the optimal regular path derived from backtracking of dynamic programming is used to eliminate the influence of path length on distance metric. To address the issue of the universality of preset distance thresholds across different viscosity grades of oils, this embodiment specifically defines a threshold. This dynamic acquisition method avoids misjudgments caused by differences in the base viscosity of oil products.
[0031] The static physicochemical fingerprint recorded for the equipment's oil in The nominal kinematic viscosity, for example ; The preset allowable deviation coefficient, for example, a value of This means that a 5% mean absolute error is allowed. This definition ensures that the distance threshold has scale adaptability; Meanwhile, for the preset similarity threshold In pseudocode compilation, this is a key decision parameter, and the source of its value is clearly defined in this embodiment: Set as The selection of this value is based on the statistical characteristics of the Pearson correlation coefficient, within a sliding window. Under the conditions, Able to The confidence level eliminates spurious correlations caused by random noise, ensuring that coupling decisions are triggered only when the theoretical aging trend and the actual total residual are highly consistent with the physical evolution law; this explicit parameter definition eliminates the problem of undefined variables in the algorithm logic; This embodiment establishes a rigorous causal logic verification mechanism through explicit dimensional projection and dynamic threshold definition; in particular, it addresses the potential overlap in judgments that may arise from OR logic. and In scenarios where results conflict, the system employs a sequential priority decision logic at the algorithm implementation layer: it prioritizes decisions that satisfy the coupling condition, i.e., decisions based on morphological similarity. or proximity If any indicator meets the standard, the model is deemed effective. The design of this logic is to tolerate single-dimensional sensor drift. For example, when a constant deviation drift occurs, the morphological similarity remains high, thus prioritizing the maintenance of the availability of the monitoring system. Only when none of the above conditions are met will the determination of non-coupling condition be triggered, and then the abnormal alarm process of Example 2 will be executed.
[0032] When generating the true aging index, the digital twin simulation server is specifically used to: extract the amplitude characteristics of the theoretical aging residual; and calculate the remaining service life of the lubricating oil based on the ratio of the amplitude characteristics to the preset aging limit threshold, which serves as the true aging index.
[0033] This embodiment further specifies the logic for generating the true aging index, focusing on the equivalent temperature projection algorithm and asymmetric threshold normalization processing to solve the problem of lifetime quantification under variable temperature conditions and shear / oxidation competition mechanisms. Once the system confirms that the theoretical aging residual and the actual total residual satisfy the coupling condition, it indicates that the current theoretical deduction is consistent with the actual physical change trend. The server then performs the following steps to calculate the actual aging index: Equivalent temperature projection: to eliminate real-time oil temperature The nonlinear effect of fluctuations on viscosity amplitude necessitates that the server uniformly convert the current state to the industry standard reference temperature. ; First, the viscosity-temperature coefficient from the static physicochemical fingerprint is called. ; Calculate the standard reference value Standard temperature ,Right now and standard pressure Substituting into the state equation model, the solution is obtained assuming that the new oil is in The theoretical viscosity is as follows; Calculate the standard aging value The above As a base, the accumulated shear dilution factor at the current moment is added. and thermal oxidizing factor The formula is calculated in real time by the model: ; Amplitude feature extraction: Calculate the theoretical aging residual amplitude after temperature normalization. :
[0034] This feature Physically represents in Below is the absolute increase in kinematic viscosity (cSt) solely due to oil aging; note It can be positive, meaning oxidation-dominated, or negative, meaning shear-dominated; Asymmetric threshold normalization and remaining life calculation: Considering that the aging limit of industrial lubricating oils is usually asymmetric, to avoid misjudgments caused by a single threshold, the system calls the alarm boundary field in the pre-stored static physical and chemical fingerprint to obtain the threshold parameter; specifically, and This is not a fixed value that is arbitrarily preset, but rather based on the ASTM / ISO oil change index standard corresponding to the lubricant. For example, for industrial gear oils, it is usually taken as the upper limit of viscosity increase +20% according to IEC60422 or OEM specifications. The lower limit of shear descent is -15%. Write to the fingerprint database; first, according to The sign determines the current absolute physical threshold. :
[0035] Subsequently, the remaining useful life is calculated based on the dynamically matched absolute threshold. Health status expressed as a percentage:
[0036] This embodiment uses a rigorous temperature projection algorithm and industry-standard asymmetric threshold logic to freeze the aging state under dynamic operating conditions to standard laboratory conditions for evaluation. This index is defined as the true aging index, which essentially uses a digital twin that has been measured and calibrated as a denoised virtual sensor to output a highly reliable lifespan prediction.
[0037] The data acquisition terminal includes a working condition data interface module and an oil sensor module. The working condition data interface module is used to collect speed, torque, oil tank temperature and pipeline pressure from the equipment control unit as real-time working condition time-series data. The oil sensor module is used to collect online viscosity value, dielectric constant, trace moisture and ferrographic density as online sensor measured data.
[0038] This embodiment further specifies the hardware configuration of the data acquisition terminal; the terminal integrates two core modules in its hardware architecture: firstly, the operating condition data interface module, which is directly connected to the device's control unit (ECU / PLC) via an industrial fieldbus, such as CAN-Open or ModbusTCP, to collect speed data at high frequency. Torque Oil tank temperature and pipeline pressure These data constitute real-time operating condition time-series data, providing physical boundary conditions for subsequent benchmark reconstruction and damage evolution. Secondly, there is the oil sensor module, integrated into the main return oil pipeline of the equipment, used to collect online viscosity values, reflecting rheological properties, dielectric constant, polar oxidation products, trace moisture, monitoring emulsification risk and ferrographic density, and monitoring wear particles. These data constitute the online sensor measured data. This embodiment achieves the hardware implementation of a physical-digital bidirectional verification mechanism by clearly defining the specific acquisition path for multi-source heterogeneous data. In particular, the fusion of operating condition data from the equipment controller and oil data from physical sensors ensures that the digital twin model has complete input variables. At the data source, the system can sense both what the equipment is doing (operating condition) and what changes have occurred in the oil (measurement), laying a solid data foundation for subsequent high-precision digital twin simulation.
[0039] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention in any other way. Any person skilled in the art may make changes or modifications to the above-disclosed technical content to create equivalent embodiments that can be applied to other fields. However, any simple modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the protection scope of the present invention.
Claims
1. A digital twin system for monitoring lubricating oil aging, characterized in that, include: The system comprises a data acquisition terminal, a digital twin simulation server, and an interactive feedback terminal. The data acquisition terminal acquires real-time operating condition data and measured data from the online lubricating oil sensor of the target device, and sends these data to the digital twin simulation server. The digital twin simulation server reconstructs the ideal baseline state at the current moment based on the real-time operating condition data and a preset static physicochemical fingerprint, and generates a simulated aging state including theoretical degradation based on the real-time operating condition data and a preset cumulative damage kinetic model. The digital twin simulation server also calculates the theoretical aging residual between the simulated aging state and the ideal baseline state, as well as the actual total residual between the measured data from the online sensor and the ideal baseline state. Furthermore, the digital twin simulation server performs topological feature comparison on the theoretical aging residual and the actual total residual, and generates a true aging index that is sent to the interactive feedback terminal when the theoretical aging residual and the actual total residual meet preset coupling conditions. The interactive feedback terminal displays the true aging index.
2. The digital twin system for monitoring lubricating oil aging according to claim 1, characterized in that, The digital twin simulation server is also used to: determine that the lubricating oil is in a non-modeled abnormal state in response to the fact that the theoretical aging residual and the actual total residual do not meet the preset coupling conditions; and send an abnormal alarm command to the interactive feedback terminal, which is used to indicate a sudden pollution event or sensor failure.
3. The digital twin system for monitoring lubricating oil aging according to claim 1, characterized in that, When reconstructing the ideal baseline state at the current moment, the digital twin simulation server is specifically used to: call the preset state equation model, take the temperature and pressure parameters in the real-time operating condition time series data as input variables; combine the basic parameters of the new oil in the static physicochemical fingerprint, calculate the theoretical physical values that the lubricating oil should present under the assumption that it is in a brand new state, and use it as the ideal baseline state.
4. The digital twin system for monitoring lubricating oil aging according to claim 1, characterized in that, When generating simulated aging states that include theoretical degradation, the digital twin simulation server is specifically used to: convert real-time operating condition time series data into aging stress factors; and based on the aging stress factors, use a tribochemical evolution model that includes shear sensitivity weights and oxidation sensitivity weights to calculate the theoretical viscosity degradation caused by mechanical shear and thermal oxidation as the theoretical damage amount. The theoretical damage amount is superimposed onto the ideal baseline state to generate the simulated aging state.
5. The digital twin system for monitoring lubricating oil aging according to claim 4, characterized in that, When converting real-time operating condition time-series data into aging stress factors, the digital twin simulation server is specifically used for: integrating the thermal load history in the real-time operating condition time-series data based on the Arrhenius equation to generate a thermal oxidation factor; and calculating the mechanical shear loss rate of polymer additives based on the rotational speed data and shear time data of the target equipment to generate a shear dilution factor.
6. The digital twin system for monitoring lubricating oil aging according to claim 1, characterized in that, When performing topological feature comparison between theoretical aging residuals and actual total residuals, the digital twin simulation server is specifically used for: calculating the cross-correlation coefficient or dynamic time warping distance between the change patterns of theoretical aging residuals and actual total residuals within a sliding time window; determining that theoretical aging residuals and actual total residuals satisfy a preset coupling condition when the cross-correlation coefficient is higher than a preset similarity threshold or the dynamic time warping distance is lower than a preset distance threshold; and determining that theoretical aging residuals and actual total residuals do not satisfy a preset coupling condition when the cross-correlation coefficient is lower than a preset similarity threshold or the dynamic time warping distance is higher than a preset distance threshold.
7. The digital twin system for monitoring lubricating oil aging according to claim 1, characterized in that, When generating the true aging index, the digital twin simulation server is specifically used to: extract the amplitude characteristics of the theoretical aging residual; and calculate the remaining service life of the lubricating oil based on the ratio of the amplitude characteristics to the preset aging limit threshold, which serves as the true aging index.
8. The digital twin system for monitoring lubricating oil aging according to any one of claims 1-7, characterized in that, The data acquisition terminal includes a working condition data interface module and an oil sensor module. The working condition data interface module is used to collect speed, torque, oil tank temperature and pipeline pressure from the equipment control unit as real-time working condition time-series data. The oil sensor module is used to collect online viscosity value, dielectric constant, trace moisture and ferrographic density as online sensor measured data.