CN-121980979-A - Hydraulic pump health evaluation method, system and equipment based on digital twin

CN121980979ACN 121980979 ACN121980979 ACN 121980979ACN-121980979-A

Abstract

The application relates to state monitoring and health management of a hydraulic system, in particular to a hydraulic pump health evaluation method and system based on a digital twin model, wherein the method comprises the steps of determining an absolute health reference state vector of a hydraulic pump under a rated working condition based on an absolute health reference model; the method comprises the steps of collecting current physical abrasion state parameters, preprocessing the current physical abrasion state parameters, carrying out hierarchical updating on a preset digital twin model based on the preprocessed current physical abrasion state parameters, carrying out closed-loop simulation operation of preset constraint conditions on the updated preset digital twin model to obtain a current fault-free ideal state vector of the hydraulic pump, calculating the difference degree of the current fault-free ideal state vector and an absolute health reference state vector in a preset performance characteristic space, and quantitatively determining a health evaluation index of the hydraulic pump based on the difference degree. The application reduces the data dependence and improves the universality and the adaptability.

Inventors

WANG JIASHUAI
XU YI
LIU PEIJUN
ZHONG BO
CHENG LI

Assignees

杭州励贝电液科技有限公司

Dates

Publication Date: 20260505
Application Date: 20260409

Claims (10)

1. The hydraulic pump health evaluation method based on the digital twin model is characterized by comprising the following steps of: Constructing an absolute health reference model of the hydraulic pump, and determining an absolute health reference state vector of the hydraulic pump under a rated working condition based on the absolute health reference model; Collecting at least one current physical abrasion state parameter of the hydraulic pump, and preprocessing the current physical abrasion state parameter; carrying out hierarchical updating on a preset digital twin model based on the preprocessed current physical abrasion state parameters; performing closed-loop simulation operation of preset constraint conditions on the updated preset digital twin model to obtain a current fault-free ideal state vector of the hydraulic pump; And calculating the difference degree of the current fault-free ideal state vector and the absolute health reference state vector in a preset performance characteristic space, and quantitatively determining the health evaluation index of the hydraulic pump based on the difference degree.
2. The method for evaluating the health of a hydraulic pump based on a digital twin model according to claim 1, wherein the determining of the absolute health reference state vector of the hydraulic pump under the rated working condition based on the absolute health reference model comprises the following steps: Constructing the absolute health reference model based on original design parameters and core working principles of the hydraulic pump; Determining a nondestructive performance output equation by combining a plunger motion rule, a flow distribution mechanism and an energy transfer characteristic of a hydraulic pump; and generating an absolute health reference state vector based on the lossless performance output equation, wherein the absolute health reference state vector represents a set of theoretical performance output parameter values when the hydraulic pump is in an ideal lossless and assembly error-free state and stably operates under a rated working condition.
3. The method for evaluating the health of a hydraulic pump based on a digital twin model according to claim 2, wherein the hierarchical step-up comprises a first layer updating mechanism and a second layer updating mechanism, wherein the hierarchical step-up updating is performed on the digital twin model of the hydraulic pump parameters constructed in advance, When the first layer is updated, the physical abrasion state parameters are utilized to directly correct the physical structure parameters of the digital twin model; collecting actual measurement operation data of a hydraulic pump in a second layer of updating mechanism, calculating output deviation of the actual measurement operation data and the digital twin model after the first layer of updating mechanism, and adjusting implicit degradation parameters of the digital twin model on line according to the output deviation; dynamically adjusting update parameters of the first layer and the second layer according to the parameter sensitivity and the implicit degradation parameters; If the digital twin model after the second layer updating mechanism continuously outputs the deviation after inputting new actually measured operation data, generating a feedback trigger signal, and calibrating the measurement accuracy of the physical mapping function or the physical abrasion state parameter of the first layer updating mechanism based on the feedback trigger signal.
4. The method for evaluating the health of a hydraulic pump based on a digital twin model according to claim 3, wherein the update parameters include a first layer trigger frequency and a second layer iteration frequency, and the update frequencies of the first layer and the second layer are dynamically adjusted according to parameter sensitivity and implicit degradation parameters, comprising the steps of: recording a time stamp of a first layer updating mechanism, calculating a corresponding first-order sensitivity coefficient based on a parameter sensitivity analysis library, and improving the triggering frequency of the first layer updating mechanism for physical structure parameters with high first-order sensitivity coefficients; aiming at the implicit degradation parameters with slow convergence of output deviation in the second layer updating mechanism, the iteration frequency of the second layer updating mechanism of the implicit degradation parameters is improved.
5. The method for evaluating the health of a hydraulic pump based on a digital twin model according to claim 1, characterized in that it comprises the steps of: Constructing a coupling relation model of core wear parameters based on a physical mechanism of a hydraulic pump, wherein the core wear parameters at least comprise a fit clearance between a plunger and a cylinder hole, wear of a joint surface of a sliding shoe and a swash plate, wear or roughness of a friction surface of a flow distribution pair, and equivalent rigidity or damping attenuation coefficient of a key sealing element; Inversion estimation is carried out on the abrasion loss of the fit clearance between the plunger and the cylinder hole and the abrasion loss of the joint surface between the slipper and the sloping cam plate based on the first layer of fusion framework, so as to obtain estimated fit clearance between the plunger and the cylinder hole and estimated abrasion loss of the joint surface between the slipper and the sloping cam plate; Inputting the estimated fit clearance between the plunger and the cylinder hole and the estimated wear amount of the joint surface of the slipper and the sloping cam plate to a second layer of fusion framework, and deducing a cooperative parameter by combining system-level performance parameters; And carrying out uncertainty quantization on all the abrasion parameters deduced by the synergy, outputting estimated values and confidence intervals of all the parameters, and combining to form a comprehensive abrasion state vector.
6. The method for evaluating the health of a hydraulic pump based on a digital twin model according to claim 5, wherein the cooperative parameters include the wear or roughness of friction surfaces of a flow distribution pair, the equivalent stiffness or damping coefficient of a key seal, and the cooperative parameters are deduced by combining system-level performance parameters, comprising the following steps: And (3) taking system-level performance parameters such as pressure, flow and vibration of the hydraulic pump as target output, and carrying out collaborative inference on the friction surface abrasion amount or roughness of the flow distribution pair, the equivalent rigidity or damping attenuation coefficient of the key sealing element based on a Kalman filtering algorithm and a coupling relation model of core abrasion parameters.
7. The digital twin model-based hydraulic pump health evaluation method according to claim 1, wherein calculating the degree of difference between the current fault-free ideal state vector and the absolute health reference state vector in a preset performance feature space comprises the steps of: And calculating the vector distance between the current fault-free ideal state vector and the absolute health reference state vector in a preset performance characteristic space, and taking the vector distance as the difference degree, wherein the vector distance adopts a weighted Euclidean distance or cosine distance.
8. The method for evaluating the health of a hydraulic pump based on a digital twin model according to claim 1, further comprising the steps of, after quantitatively determining the health evaluation index of the hydraulic pump based on the degree of difference: continuously collecting health evaluation indexes obtained by calculation at different moments, and arranging time sequence data sequences forming the health evaluation indexes according to time sequence; Training and predicting the time sequence data sequence of the health evaluation index by using a time sequence prediction model so as to obtain a predicted performance degradation track; Predicting the residual usable life of the hydraulic pump according to the performance degradation track and a preset failure threshold; And outputting a confidence interval of the residual usable life by analyzing the prediction error of the time sequence prediction model and the fluctuation characteristic of the physical abrasion parameter, wherein the confidence interval explicitly comprises the prediction range of the residual usable life and the corresponding confidence.
9. A digital twin model-based hydraulic pump health evaluation system for performing the digital twin model-based hydraulic pump health evaluation method of any one of claims 1-8, the system comprising: The reference construction module is used for constructing an absolute health reference model of the hydraulic pump and determining an absolute health reference state vector of the hydraulic pump under a rated working condition based on the absolute health reference model; The parameter sensing processing module is used for collecting at least one current physical abrasion state parameter of the hydraulic pump and preprocessing the current physical abrasion state parameter; The digital twin engine module performs layered stepwise updating on the preset digital twin model based on the preprocessed current physical abrasion state parameters, and performs closed-loop simulation operation of preset constraint conditions on the updated preset digital twin model to acquire a current fault-free ideal state vector of the hydraulic pump; The health evaluation decision module calculates the difference degree of the current fault-free ideal state vector and the absolute health reference state vector in a preset performance characteristic space, and quantitatively determines the health evaluation index of the hydraulic pump based on the difference degree.
10. An electronic device comprising a processor and a memory coupled to each other, the memory having stored thereon a computer program executable on the processor; the computer program, when executed by the processor, implements the digital twin model-based hydraulic pump health assessment method of any one of claims 1-8.

Description

Hydraulic pump health evaluation method, system and equipment based on digital twin Technical Field The invention relates to the technical field of hydraulic system state monitoring and health management, in particular to a hydraulic pump health evaluation method, system and equipment based on digital twinning. Background The hydraulic pump refers to all devices for converting mechanical energy into hydraulic energy, and the axial plunger pump is a high-performance and high-efficiency structural form in the hydraulic pump. The axial plunger pump is widely applied to key fields of engineering machinery, aerospace, industrial hydraulic pressure and the like due to the characteristics of high pressure and high efficiency, and the health state of the axial plunger pump directly influences the reliability, safety and operation efficiency of the whole hydraulic system. Currently, state evaluation of axial plunger pumps in industry mainly depends on signal processing technologies such as vibration analysis, oil monitoring, pressure pulsation analysis and the like, and statistical analysis models based on historical operation data. The core logic of the method is to compare the actual data (such as vibration signals, pressure, temperature and the like) in the running process of the equipment with a preset fault threshold value or a degradation model obtained through historical data training, so as to judge the current state of the equipment or predict the residual service life. However, these methods are highly dependent on large, high quality historical data, and models are often built for specific equipment or specific conditions. The generalization capability of the model is limited by the quality and range of data, and the effect may be reduced when the model is transplanted to new equipment or different working conditions. Disclosure of Invention In order to reduce the dependence of evaluation accuracy on historical data of the axial plunger pump and accurately improve the accuracy of health evaluation of the axial plunger pump, the application provides a digital twin-based health evaluation method, a digital twin-based health evaluation system and digital twin-based health evaluation equipment for the axial plunger pump. In a first aspect, the application provides a hydraulic pump health evaluation method based on a digital twin model, which adopts the following technical scheme: a hydraulic pump health evaluation method based on a digital twin model, the method comprising the steps of: Constructing an absolute health reference model of the hydraulic pump, and determining an absolute health reference state vector of the hydraulic pump under a rated working condition based on the absolute health reference model; Collecting at least one current physical abrasion state parameter of the hydraulic pump, and preprocessing the current physical abrasion state parameter; carrying out hierarchical updating on a preset digital twin model based on the preprocessed current physical abrasion state parameters; performing closed-loop simulation operation of preset constraint conditions on the updated preset digital twin model to obtain a current fault-free ideal state vector of the hydraulic pump; And calculating the difference degree of the current fault-free ideal state vector and the absolute health reference state vector in a preset performance characteristic space, and quantitatively determining the health evaluation index of the hydraulic pump based on the difference degree. In another embodiment, the determining the absolute health reference state vector of the hydraulic pump under the rated working condition based on the absolute health reference model includes the following steps: Constructing the absolute health reference model based on original design parameters and core working principles of the hydraulic pump; Determining a nondestructive performance output equation by combining a plunger motion rule, a flow distribution mechanism and an energy transfer characteristic of a hydraulic pump; and generating an absolute health reference state vector based on the lossless performance output equation, wherein the absolute health reference state vector represents a set of theoretical performance output parameter values when the hydraulic pump is in an ideal lossless and assembly error-free state and stably operates under a rated working condition. In another embodiment, the hierarchical updating comprises a first layer updating mechanism and a second layer updating mechanism, wherein the hierarchical updating comprises the steps of performing hierarchical updating on a digital twin model of a pre-constructed hydraulic pump parameter, When the first layer is updated, the physical abrasion state parameters are utilized to directly correct the physical structure parameters of the digital twin model; collecting actual measurement operation data of a hydraulic pump in a second layer of updating mechanism, calculating output