A hydraulic system fault diagnosis method and system based on energy transmission efficiency

Abstract

The present application relates to a kind of hydraulic system fault diagnosis method and system based on energy transmission efficiency, belong to hydraulic system diagnosis technical field, specifically including acquisition hydraulic system operating data, calculate component pressure difference, pump operating frequency, vibration main frequency and band energy etc. Characteristic parameters;Combined with the structure and failure mode of hydraulic pump, valve, cylinder, construct failure mechanism model, calculate component leakage and vibration dissipation energy and system total failure dissipation energy;With component equivalent failure parameter as state variable, three kinds of energy transmission efficiency as observation, build nonlinear state space model, rely on unscented Kalman filtering recursion estimation failure parameter;Combined with parameter estimation result, efficiency degradation degree and vibration frequency characteristics, determine hydraulic system failure component, type and severity, output diagnosis result, the present application can improve the real-time performance, accuracy and explainability of hydraulic system fault diagnosis.

Description

Technical Field

[0001] This invention relates to a method and system for diagnosing hydraulic system faults based on energy transfer efficiency, belonging to the field of hydraulic system diagnostic technology. Background Technology

[0002] Hydraulic systems are widely used in engineering machinery, aerospace, metallurgical equipment, mining equipment, marine equipment, and automated production equipment. A hydraulic system typically consists of a hydraulic pump, hydraulic valves, hydraulic cylinders, pipelines, an oil tank, and a load-bearing mechanism, and is characterized by high power density, fast response speed, and strong load-bearing capacity. If a hydraulic system malfunctions, it can cause problems such as sluggish actuator movement, insufficient output force, pressure fluctuations, abnormal flow, increased vibration, and elevated temperature. In severe cases, it can lead to equipment shutdown and even safety risks.

[0003] Common hydraulic system faults include abnormal vibrations caused by factors such as wear of the piston pair, flow distribution pair, and slipper pair of the hydraulic pump (taking the piston pump as the research object), increased clearance between the valve core and valve body of the hydraulic valve, wear of the piston seal of the hydraulic cylinder, and leakage of the hydraulic cylinder.

[0004] Hydraulic systems are complex in structure, exhibit diverse failure modes, and their fault characteristics are difficult to measure directly, with numerous coupling factors during propagation. These faults are often not isolated; their coupling characteristics manifest on two levels. First, single-fault propagation coupling occurs when the energy of a single component failure propagates upstream and downstream along the hydraulic transmission path, triggering linked changes in system pressure, flow, and vibration. For example, a hydraulic pump failure affects the valve's operating state, thereby altering the hydraulic cylinder's action characteristics. Changes in the hydraulic cylinder load, in turn, affect the system pressure, forming a complex fault propagation path. Second, multi-fault coupling occurs when multiple components, such as hydraulic pumps, valves, and cylinders, fail simultaneously. The pressure pulsations, vibration components, and leakage flow generated by each fault source interact and superimpose within the system, causing sensor signals to present a mixture of multi-source fault characteristics, making it difficult to directly separate and locate individual fault sources. The combined effect of single-fault propagation coupling and multi-fault coupling, coupled with the unmeasurability of internal leakage, significantly increases the difficulty of fault location and severity assessment. Current hydraulic system fault diagnosis methods mainly include feature analysis methods based on pressure and vibration signals, and data-driven diagnostic methods based on machine learning. However, most of the above methods rely on a single type of signal and lack a unified description of the fault mechanism, making it difficult to achieve unified modeling and quantitative parameter inversion of multi-component coupled faults under complex operating conditions.

[0005] Energy transfer is the essence of hydraulic transmission, and energy transfer efficiency is a crucial indicator of the state of a hydraulic transmission system. Under normal operating conditions, the hydraulic pump converts mechanical energy into hydraulic energy, the hydraulic valve regulates and distributes the hydraulic energy, and the hydraulic cylinder converts the hydraulic energy back into mechanical energy, forming a complete energy transfer path. When a system fault occurs, the decreased energy conversion efficiency of local components will affect upstream and downstream components along the transfer path. Whether it is single-fault propagation coupling or multi-fault coupling, it will ultimately manifest as a comprehensive decrease in the system's energy transfer efficiency. Simultaneously, internal leakage faults manifest as abnormal hydraulic energy loss, and impact faults such as wear manifest as an abnormal increase in vibration energy; both are different forms of energy transfer efficiency loss. Therefore, faults of different types and locations can be uniformly characterized by energy transfer efficiency, thus providing a unified theoretical framework for diagnosing multi-component coupled faults. Summary of the Invention

[0006] To address the technical problems existing in the prior art, this invention provides a hydraulic system fault diagnosis method and system based on energy transfer efficiency. It unifies the leakage energy loss and vibration energy loss of key components such as hydraulic pumps, hydraulic valves and hydraulic cylinders into the fault dissipation energy framework, thereby improving the real-time performance, accuracy and interpretability of hydraulic system fault diagnosis.

[0007] To achieve the above objectives, the technical solution adopted by this invention is a hydraulic system fault diagnosis method based on energy transfer efficiency, comprising the following steps: S1. Collect hydraulic system operating status data, calculate the pressure difference between upstream and downstream hydraulic components, determine the hydraulic pump operating frequency and related characteristic frequencies, and extract the vibration main frequency and characteristic frequency band energy; S2. Establish a fault mechanism model based on the structural characteristics and fault modes of hydraulic pumps, hydraulic valves, and hydraulic cylinders; S3. Calculate the leakage dissipation energy and vibration dissipation energy of hydraulic components based on the fault mechanism model, and then combine them to obtain the total fault dissipation energy of the system. S4. Using the equivalent fault parameters of hydraulic pumps, hydraulic valves and hydraulic cylinders as state variables, and hydraulic energy transfer efficiency, mechanical energy transfer efficiency and system comprehensive energy transfer efficiency as observations, a nonlinear state-space model is established, and the unscented Kalman filter algorithm is used to recursively estimate the fault parameters. S5. Based on the fault parameter estimation results, the degree of degradation of hydraulic energy transmission efficiency, the degree of degradation of mechanical energy transmission efficiency, the degree of degradation of the overall system energy transmission efficiency, and the vibration characteristic frequency distribution, the faulty components, fault types, and fault severity in the hydraulic system are judged and the diagnostic results are output.

[0008] Preferably, in step S1, the hydraulic system operating status data includes the inlet and outlet pressures of the hydraulic pump, the inlet and outlet pressures of the hydraulic valve, the pressures of the two chambers of the hydraulic cylinder, the speed of the hydraulic pump, and the vibration acceleration signals at the hydraulic pump housing, the valve block of the hydraulic valve, and the mounting base of the hydraulic cylinder.

[0009] Preferably, in step S1, in step S2, when establishing the fault mechanism model, an internal leakage model is established based on the structural characteristics and fault modes of the hydraulic pump, hydraulic valve, and hydraulic cylinder. The internal leakage model includes a hydraulic pump leakage model, a hydraulic valve leakage model, and a hydraulic cylinder leakage model.

[0010] Preferably, in step S3, the leakage dissipation energy is calculated based on the pressure difference between the upstream and downstream of the hydraulic component and the leakage flow rate, the vibration dissipation energy is calculated based on the vibration acceleration amplitude, characteristic angular frequency and equivalent vibration mass, and the total system failure dissipation energy is the sum of the leakage dissipation energy and vibration dissipation energy of each hydraulic component.

[0011] Preferably, in step S4, the hydraulic energy transmission efficiency, mechanical energy transmission efficiency, and system comprehensive energy transmission efficiency are obtained by normalizing the fault dissipation energy and the system input hydraulic energy, respectively. The hydraulic energy transfer efficiency is used to represent the degree of hydraulic energy loss caused by leakage faults in the hydraulic system; the mechanical energy transfer efficiency is used to represent the degree of mechanical energy dissipation caused by wear and other faults in the hydraulic system; the overall system energy transfer efficiency is used to represent the overall energy transfer capability of the hydraulic system. The overall energy transfer efficiency of the system is obtained using a weighted fusion method: , in, For hydraulic energy transmission efficiency, For mechanical energy transfer efficiency, For the overall energy transfer efficiency of the system, , These are the weighting coefficients, and in , , , and All of these are dimensionless quantities. The weighted comprehensive evaluation index is used for fault diagnosis and is not directly equivalent to... ; , Health benchmarks should be calibrated using health data under the same working conditions.

[0012] Preferably, in step S4, the unscented Kalman filter algorithm includes the following steps: S41. Set the initial state estimate, initial covariance matrix, process noise covariance matrix, and observation noise covariance matrix: , , , , In the formula, The initial fault parameter estimates consist of the initial equivalent clearances of the hydraulic pump, valves, and cylinders. These are the equivalent clearances for the health benchmarks of hydraulic pumps, valves, and cylinders, respectively. The initial state covariance matrix; This represents the true state vector of the system at the initial moment. The process noise covariance matrix; This is process noise; To observe noise; The observation noise covariance matrix; the subtraction within parentheses represents the deviation between the true state and the estimated state, and the product of the two deviation vectors is the outer product; subscripts Indicates the initial time, subscript Indicates the first A diagnostic moment; Indicates the equivalent gap, subscript , , These represent hydraulic pumps, hydraulic valves, and hydraulic cylinders, respectively. and All have a zero mean; S42. Generate based on the current state estimate and covariance matrix. Sigma points: , , , In the formula, For the first One Sigma point; The first square root of the matrix List; For the first The state covariance matrix at time t, Let the dimension of the state variables be . Let be the scale parameter of the Sigma point distribution. The Sigma point index; the square root indicates the matrix. Find the square root of the matrix. This indicates taking the first square root of the matrix. List," "and" "Used to construct the positive and negative offset Sigma points on both sides of the mean, respectively, subscript..." The central Sigma point is represented; the scale parameter and weights are further defined as follows: , , , , ;in, For the distribution range parameters of Sigma points, Used to reflect the characteristics of the prior distribution of the state. As an auxiliary scale parameter, and satisfying ; S43. Substitute each Sigma point into the state transition function to obtain the predicted Sigma point: , And calculate the mean of the state prediction: In the formula, For the first One predicted Sigma point; For the first Predicted values ​​of fault parameters at any given time; This is the state transition function. For the first The system input or known control variable at a given time. For the previous version Sigma points, The first state obtained through state transition One predicted Sigma point; For the first The mean weight of each Sigma point, superscript " "Indicates the prior prediction value before the observation update; S44. Calculate the prediction covariance matrix: In the formula, The state prediction covariance matrix; For the first Covariance weights of Sigma points; This represents the outer product of the vectors representing the deviations of the predicted Sigma points from the predicted mean. The process noise covariance matrix is ​​represented by the superscript. Indicates transpose; S45. Substitute the predicted Sigma points into the energy transfer efficiency observation function, and first calculate the leakage dissipation energy based on the equivalent fault parameters corresponding to each predicted Sigma point. and vibration dissipation of energy Then, it is normalized into hydraulic energy transmission efficiency, mechanical energy transmission efficiency and system comprehensive energy transmission efficiency. The predicted value of hydraulic energy transmission efficiency is expressed as: , The predicted value of mechanical energy transfer efficiency is expressed as: , The predicted value of the system's overall energy transfer efficiency is expressed as follows: , Therefore, we obtain the first... Predicted observations corresponding to each Sigma point: , Calculate the observed and predicted mean: In the formula, This is the observation vector for predicting energy transfer efficiency; where, and The first The leakage dissipation energy and vibration dissipation energy corresponding to each predicted Sigma point. Input hydraulic energy into the system within the same diagnostic time window; , and These represent the hydraulic energy, mechanical energy, and overall system energy transfer efficiency corresponding to the Sigma point, all of which are dimensionless quantities. The vector is composed of the three efficiencies mentioned above, in the order of hydraulic energy, mechanical energy, and overall system energy. Square brackets indicate a vector. Weighted by mean, To predict the observed mean; S46. Calculate the observation-prediction covariance matrix: In the formula, To observe the covariance matrix; Calculate the cross-covariance matrix between the state variables and the observed variables: , In the formula, This indicates the correlation between changes in fault parameters and changes in energy transfer efficiency. For the first The predicted observation vector of Sigma points, and the aforementioned Same meaning; For the first The predicted observation mean at time, compared with the aforementioned Same meaning; To observe and predict the covariance matrix, To observe the noise covariance matrix; This is the cross-covariance matrix between the state and the observation; For covariance weights, Indicates to to Summation: Multiply the two deviation vectors and add a superscript. Indicates the outer product; S47. Calculate the Kalman gain: In the formula, This is the Kalman gain matrix, used to determine the degree of influence of real-time observations on state corrections; Represents the observation-prediction covariance matrix inverse matrix, matrix product Used to map observation residuals to state corrections. subscript Indicates the first A diagnostic moment; S48. Based on real-time observation vectors With predicted observation vector The deviation updates the state variables: , In the formula, For the first The fault parameter estimates after real-time correction; For the first The measured energy transfer efficiency observation vector at any given time. For the corresponding predicted observation vector, To observe the residuals, This is the state correction amount. This is the prior state estimate before correction; " indicates that the correction amount is added to the prior state estimate; If the real-time observed energy transfer efficiency is lower than the predicted energy transfer efficiency, it indicates that the system's energy transfer capability has decreased more than the predicted value, and the current fault parameters may be underestimated. The algorithm corrects this by increasing the Kalman gain. If the real-time observed energy transfer efficiency is higher than the predicted energy transfer efficiency, it indicates that the decrease in the system's energy transfer capacity is less than the predicted value, and the current fault parameters may be overestimated. The algorithm will then reduce and correct this. S49, Covariance Update: In the formula, For the first The state covariance matrix updated at each time step; For the first The predicted state covariance matrix before the time update. To observe and predict the covariance matrix, Kalman gain matrix Transpose of; The minus sign indicates that the uncertainty eliminated after this observation is used, and the superscript indicates that this part is subtracted from the predicted covariance. This indicates transpose.

[0013] Preferably, in step S5, when outputting the diagnostic results, the following steps are performed: S51. First, calculate the fault severity coefficient: In the formula, For the first Failure severity coefficient of each component For the first Estimated fault parameters for each component This is the initial design clearance or reference clearance for the component under healthy conditions; For hydraulic component serial numbers; It is a dimensionless ratio. This indicates that the estimated gap is equal to the healthy baseline gap. A value greater than 1 indicates that the gap is larger than that of a healthy state; S52, according to The fault severity is classified into normal, minor, moderate, and severe faults based on the threshold range in which it is located. S53. Based on the calculated hydraulic energy transmission efficiency, mechanical energy transmission efficiency and system comprehensive energy transmission efficiency, calculate the energy transmission efficiency degradation. When the degradation of hydraulic energy transmission efficiency is large, while the degradation of mechanical energy transmission efficiency is small, it indicates that hydraulic energy loss is dominant, and the fault is mainly internal leakage. When the mechanical energy transmission efficiency degrades significantly while the hydraulic energy transmission efficiency degrades only slightly, it indicates increased mechanical vibration dissipation, and the fault is primarily determined to be a wear fault. When both hydraulic energy transmission efficiency and mechanical energy transmission efficiency decrease significantly, it indicates that hydraulic energy loss and mechanical energy loss exist simultaneously, and it is judged as a combined leakage-wear fault. S54. Combining the fault severity coefficient, energy transfer efficiency degradation characteristics, and vibration characteristic frequency distribution, comprehensively determine the faulty component, fault type, and fault severity, and output the diagnostic results.

[0014] A hydraulic system fault diagnosis system based on energy transfer efficiency includes, The data acquisition unit is used to collect hydraulic system operating status data, including hydraulic pump inlet and outlet pressures, hydraulic valve inlet and outlet pressures, hydraulic cylinder two-chamber pressures, hydraulic pump speed, and vibration acceleration signals at the hydraulic pump housing, hydraulic valve block, and hydraulic cylinder mounting base. The unit performs low-pass filtering, time synchronization processing, and fixed time window segmentation processing on the pressure signals, speed signals, and vibration signals to obtain status detection data for fault diagnosis. The fault mechanism model unit is used to establish the correspondence between the fault parameters of hydraulic pumps, hydraulic valves and hydraulic cylinders and the leakage flow and vibration response based on the fault mechanism of key components of the hydraulic system. Among them, the fault parameters of hydraulic pumps include the equivalent wear clearance of piston pair, flow distribution pair and slipper pair; the fault parameters of hydraulic valves include the equivalent radial clearance between valve core and valve body; and the fault parameters of hydraulic cylinders include the equivalent leakage clearance between piston and cylinder. The fault dissipation energy calculation unit is used to calculate the leakage dissipation energy based on the pressure difference between the upstream and downstream of the hydraulic component and the leakage flow rate, calculate the vibration dissipation energy based on the vibration acceleration amplitude, characteristic angular frequency and equivalent vibration mass, and integrate the leakage dissipation energy and the vibration dissipation energy to obtain the total fault dissipation energy of the system. The fault parameter estimation unit based on energy transfer efficiency is used to establish a nonlinear state-space model with the equivalent fault parameters of hydraulic pumps, hydraulic valves and hydraulic cylinders as state variables, and the hydraulic energy transfer efficiency, mechanical energy transfer efficiency and system comprehensive energy transfer efficiency as observations. The unscented Kalman filter algorithm is then used to recursively estimate the fault parameters. The fault diagnosis output unit based on energy transfer efficiency is used to judge the faulty components, fault types and fault severity in the hydraulic system and output diagnostic results based on the estimated fault parameters, the degree of degradation of hydraulic energy transfer efficiency, the degree of degradation of mechanical energy transfer efficiency, the degree of degradation of the overall system energy transfer efficiency and the vibration characteristic frequency distribution.

[0015] Compared with existing technologies, this invention has the following technical advantages: It adopts the concept of energy transfer efficiency to uniformly describe the abnormal energy loss caused by internal leakage, wear, and other faults in hydraulic systems; it establishes a system-level fault dissipation energy model for hydraulic pumps, hydraulic valves, and hydraulic cylinders, which can reduce misjudgments caused by fluctuations in operating conditions caused by single-signal diagnostic methods; it uses an unscented Kalman filter algorithm to back-estimate fault parameters, achieving better real-time performance and stability under nonlinear models and measurement noise conditions; the diagnostic results include not only whether a fault exists, but also the fault location, fault type, estimated fault parameters, and fault severity, possessing strong engineering application value.

[0016] Furthermore, this invention can identify and estimate parameters of internal leakage faults in hydraulic systems without the need for flow sensors, and indirectly measure unmeasurable fault parameters through energy inversion methods, thereby reducing sensor costs and improving system-level diagnostic capabilities. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the steps of the method of the present invention.

[0018] Figure 2 This is a schematic diagram of a diagnostic framework for hydraulic systems based on energy transfer efficiency.

[0019] Figure 3 This is a schematic diagram of the failure mechanism of a hydraulic pump, hydraulic valve, and hydraulic cylinder. Detailed Implementation

[0020] To make the technical problems to be solved, the technical solutions, and the beneficial effects of the present invention clearer, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present invention and are not intended to limit the present invention.

[0021] like Figure 1 , Figure 2 As shown, a fault diagnosis method for hydraulic systems based on energy transfer efficiency includes the following five steps: S1. Collect hydraulic system operating status data, including pressure signals such as hydraulic pump inlet and outlet pressures, hydraulic valve inlet and outlet pressures, and hydraulic cylinder two-chamber pressures, as well as hydraulic pump speed signals and vibration acceleration signals at the hydraulic pump housing, hydraulic valve block, and hydraulic cylinder mounting base. The pressure signals are used to calculate the pressure difference between the upstream and downstream hydraulic components; the speed signals are used to determine the hydraulic pump's operating frequency and related characteristic frequencies; and the vibration acceleration signals are used to extract the dominant vibration frequency and characteristic frequency band energy.

[0022] S2. Based on the structural characteristics and failure modes of hydraulic pumps, hydraulic valves, and hydraulic cylinders, establish failure mechanism models. When establishing these models, hydraulic pump failures mainly consider increased leakage and enhanced vibration caused by wear of the plunger pair, flow distribution pair, and slipper pair; hydraulic valve failures mainly consider internal leakage caused by increased clearance between the valve core and valve sleeve, and abnormal valve core movement; hydraulic cylinder failures mainly consider oil leakage between the two chambers and low-frequency reciprocating impact caused by piston seal wear. An internal leakage model is also established based on the structural characteristics and failure modes of hydraulic pumps, hydraulic valves, and hydraulic cylinders. This internal leakage model includes a hydraulic pump leakage model, a hydraulic valve leakage model, and a hydraulic cylinder leakage model. Under the conditions that the hydraulic oil is in a stable laminar flow state, the compressibility of the oil is negligible, and the leakage models of the three major friction pairs are equivalent to radial disk leakage, parallel disk leakage, and eccentric ring leakage, respectively, an internal leakage model of the hydraulic pump is established. Internal leakage model of a plunger pump: , In the formula: It is the average clearance between the cylinder block and the distributor plate (characterized by the oil film thickness between them); It refers to the dynamic viscosity of hydraulic oil; and It refers to the inner and outer radii of the inner sealing oil strip; and These are the inner and outer radii of the external sealing oil strip; The coefficient representing the influence of tilt on leakage; Number of plungers; The difference between the piston cavity radius and the piston radius is... ; This is the average length of the piston entering the cylinder. The swashplate inclination angle; The average gap between the skid and the ramp (characterized by the thickness of the oil film between them). and These are the inner and outer diameters of the skate, respectively. for The total internal leakage flow rate of the plunger pump at any given time. , and These are the leakage flows of the slipper pair, the plunger pair, and the cylinder block and distributor plate distribution pair, respectively. and These are the high-pressure side pressure and the low-pressure side reference pressure of the pump, respectively, and the difference between the two is the pressure difference for driving leakage; The diameter of the plunger. This represents the diameter of the plunger cavity. (Subscript) Indicates a hydraulic pump. Indicates the side shoe, This indicates the cylinder block-distributor plate assembly; Under the conditions that there is a small radial gap between the valve core and the valve sleeve, the heat generation is small, and the leakage model is equivalent to the eccentric ring leakage, an internal leakage model of hydraulic valve is established. Valve internal leakage model: , In the formula: The diameter of the valve orifice; The diameter of the valve core; Radial clearance ( =(D−d) / 2); The working pressure difference between the upstream and downstream sides of the valve port; This refers to the sealing length of the mating surface between the valve core and the valve sleeve. Eccentricity , (where is the eccentricity), where for The leakage flow rate inside the hydraulic valve is constantly monitored. The value represents the dynamic viscosity of the hydraulic oil; the subscript V indicates a hydraulic valve. A hydraulic cylinder internal leakage model is established under the condition that there is a small radial clearance between the piston and the cylinder and the leakage model is equivalent to eccentric ring leakage. Leakage model of hydraulic cylinder: , in It is the piston diameter. It is the clearance between the piston and the hydraulic cylinder. It is the total length of the linear clearance between the piston and the hydraulic cylinder barrel. for Constantly monitor the leakage flow rate inside the hydraulic cylinder. The piston's speed. This refers to the pressure in the high-pressure chamber of the hydraulic cylinder. This is the reference pressure for the low-pressure chamber or the return oil chamber. The value represents the viscosity of the hydraulic oil, and the subscript C indicates the hydraulic cylinder. The first term on the right side of the equal sign represents shearing and dragging leakage caused by piston movement, and the second term represents clearance leakage caused by pressure difference.

[0023] S3. Calculate the leakage dissipation energy and vibration dissipation energy of the hydraulic components based on the fault mechanism model, and then combine them to obtain the total fault dissipation energy of the system. The leakage dissipation energy is calculated based on the pressure difference between the upstream and downstream of the hydraulic component and the leakage flow rate. The vibration dissipation energy is calculated based on the vibration acceleration amplitude, characteristic angular frequency and equivalent vibration mass. The total fault dissipation energy of the system is the sum of the leakage dissipation energy and vibration dissipation energy of each hydraulic component. Leakage dissipation energy reflects the hydraulic energy loss caused by internal leakage, while vibration dissipation energy reflects the mechanical energy dissipation caused by wear, loosening, cavitation, or impact. , In the formula, For the first Energy dissipation due to leakage of individual hydraulic components. For the first The pressure difference between the upstream and downstream sides of a hydraulic component Let i be the internal leakage flow rate caused by the fault clearance of the i-th hydraulic component. and The first The generalized potential variable and generalized flow variable of a leakage branch of a hydraulic component correspond to the pressure difference in the hydraulic domain, respectively. and leaked flow The product of the two is the instantaneous leakage power. , Combined with system stiffness Natural angular frequency Derivation , In the formula, For the first The hydraulic component dissipates energy through vibration, where A is the acceleration amplitude at the characteristic frequency. The characteristic angular frequency, For equivalent vibration mass, For the first The hydraulic component dissipates energy through vibration. In this formula... As an equivalent vibration energy characteristic of mechanical vibration dissipation within the diagnostic time window, it is applicable to single-degree-of-freedom linear simple harmonic motion and narrowband dominant frequency approximation, but is not equivalent to damped cumulative dissipation energy in a strict sense. When there are multiple components to be monitored in the system, the total fault dissipation energy of the system can be expressed as the sum of the leakage dissipation energy and vibration dissipation energy of each component. , , In the formula, To diagnose the total system failure energy dissipation within the diagnostic time window, This represents the total energy dissipated due to system leakage. This represents the total energy dissipated by the system's vibration. and The first Leakage energy dissipation and vibration energy dissipation of each component; The total number of hydraulic components to be monitored. The index is used for summation.

[0024] S4. Using the equivalent fault parameters of hydraulic pumps, hydraulic valves and hydraulic cylinders as state variables, and hydraulic energy transfer efficiency, mechanical energy transfer efficiency and system comprehensive energy transfer efficiency as observations, a nonlinear state-space model is established, and the unscented Kalman filter algorithm is used to recursively estimate the fault parameters. Hydraulic energy transfer efficiency, mechanical energy transfer efficiency, and overall system energy transfer efficiency are obtained by normalizing the energy dissipated due to faults and the hydraulic energy input to the system, respectively. Hydraulic energy transfer efficiency is used to represent the degree of hydraulic energy loss caused by leakage faults in the hydraulic system. Mechanical energy transfer efficiency is used to represent the degree of mechanical energy dissipation caused by wear and other faults in the hydraulic system. Overall system energy transfer efficiency is used to represent the overall energy transfer capacity of the hydraulic system. Let the system input hydraulic energy be The energy dissipated by hydraulic leakage is The energy dissipated by mechanical vibration is Then, the hydraulic energy transmission efficiency, mechanical energy transmission efficiency, and overall system energy transmission efficiency are expressed as follows: , , , In the formula, For hydraulic energy transmission efficiency, For mechanical energy transfer efficiency, For the overall energy transfer efficiency of the system, , These are the weighting coefficients, and The overall energy transfer efficiency of the system is calculated using a weighted fusion method, with the weights determined based on the type of hydraulic system, the importance of components, or empirical calibration. The fault parameters are then recursively estimated using an unscented Kalman filter algorithm, which includes the following steps: S41. Set the initial state estimate, initial covariance matrix, process noise covariance matrix, and observation noise covariance matrix: , , , , In the formula, The initial fault parameter estimates consist of the initial equivalent clearances of the hydraulic pump, valves, and cylinders. These are the equivalent clearances for the health benchmarks of hydraulic pumps, valves, and cylinders, respectively. The initial state covariance matrix; This represents the true state vector of the system at the initial moment. The process noise covariance matrix; This is process noise; To observe noise; The observation noise covariance matrix; the subtraction within parentheses represents the deviation between the true state and the estimated state, and the product of the two deviation vectors is the outer product; subscripts Indicates the initial time, subscript Indicates the first A diagnostic moment; Indicates the equivalent gap, subscript , , These represent hydraulic pumps, hydraulic valves, and hydraulic cylinders, respectively. and All have a zero mean; S42. Generate based on the current state estimate and covariance matrix. Sigma points: , , , In the formula, For the first One Sigma point; The first square root of the matrix Columns are typically obtained through Cholesky decomposition; For the first The state covariance matrix at time t, Let the dimension of the state variables be . Let be the scale parameter of the Sigma point distribution. The Sigma point index; the square root indicates the matrix. Find the square root of the matrix. This indicates taking the first square root of the matrix. List," "and" "Used to construct the positive and negative offset Sigma points on both sides of the mean, respectively, subscript..." The central Sigma point is represented; the scale parameter and weights are further defined as follows: , , , , ;in, For the distribution range parameters of Sigma points, Used to reflect the characteristics of the prior distribution of the state. As an auxiliary scale parameter, and satisfying ; S43. Substitute each Sigma point into the state transition function to obtain the predicted Sigma point: , And calculate the mean of the state prediction: In the formula, For the first One predicted Sigma point; For the first Predicted values ​​of fault parameters at any given time; S44. Calculate the prediction covariance matrix: In the formula, The state prediction covariance matrix; For the first Covariance weights of Sigma points; This represents the outer product of the vectors representing the deviations of the predicted Sigma points from the predicted mean. The process noise covariance matrix is ​​represented by the superscript. Indicates transpose; S45. Substitute the predicted Sigma points into the energy transfer efficiency observation function, and first calculate the leakage dissipation energy based on the equivalent fault parameters corresponding to each predicted Sigma point. and vibration dissipation of energy Then, it is normalized into hydraulic energy transmission efficiency, mechanical energy transmission efficiency and system comprehensive energy transmission efficiency. The predicted value of hydraulic energy transmission efficiency is expressed as: , The predicted value of mechanical energy transfer efficiency is expressed as: , The predicted value of the system's overall energy transfer efficiency is expressed as follows: , Therefore, we obtain the first... Predicted observations corresponding to each Sigma point: , Calculate the observed and predicted mean: In the formula, This is the observation vector for predicting energy transfer efficiency; where, and The first The leakage dissipation energy and vibration dissipation energy corresponding to each predicted Sigma point. Input hydraulic energy into the system within the same diagnostic time window; , and These represent the hydraulic energy, mechanical energy, and overall system energy transfer efficiency corresponding to the Sigma point, all of which are dimensionless quantities. The vector is composed of the three efficiencies mentioned above, in the order of hydraulic energy, mechanical energy, and overall system energy. Square brackets indicate a vector. Weighted by mean, To predict the observed mean; S46. Calculate the observation-prediction covariance matrix: In the formula, The observation covariance matrix is ​​used to describe the uncertainty of the observed quantities for predicting energy transfer efficiency. Calculate the cross-covariance matrix between the state variables and the observed variables: , In the formula, This indicates the correlation between changes in fault parameters and changes in energy transfer efficiency. For the first The predicted observation vector of Sigma points, and the aforementioned Same meaning; For the first The predicted observation mean at time, compared with the aforementioned Same meaning; To observe and predict the covariance matrix, To observe the noise covariance matrix; This is the cross-covariance matrix between the state and the observation; For covariance weights, Indicates to to Summation: Multiply the two deviation vectors and add a superscript. Indicates the outer product; S47. Calculate the Kalman gain: In the formula, This is the Kalman gain matrix, used to determine the degree of influence of real-time observations on state corrections; Represents the observation-prediction covariance matrix inverse matrix, matrix product Used to map observation residuals to state corrections. subscript Indicates the first A diagnostic moment; S48. Based on real-time observation vectors With predicted observation vector The deviation updates the state variables: , In the formula, For the first The fault parameter estimates after real-time correction; For the first The measured energy transfer efficiency observation vector at any given time. For the corresponding predicted observation vector, To observe the residuals, This is the state correction amount. This is the prior state estimate before correction; " indicates that the correction amount is added to the prior state estimate; If the real-time observed energy transfer efficiency is lower than the predicted energy transfer efficiency, it indicates that the system's energy transfer capability has decreased more than the predicted value, and the current fault parameters may be underestimated. The algorithm corrects this by increasing the Kalman gain. If the real-time observed energy transfer efficiency is higher than the predicted energy transfer efficiency, it indicates that the decrease in the system's energy transfer capacity is less than the predicted value, and the current fault parameters may be overestimated. The algorithm will then reduce and correct this. S49, Covariance Update: In the formula, For the first The state covariance matrix updated at each time step; For the first The predicted state covariance matrix before the time update. To observe and predict the covariance matrix, Kalman gain matrix Transpose of; The minus sign indicates that the uncertainty eliminated after this observation is used, and the superscript indicates that this part is subtracted from the predicted covariance. This indicates transpose.

[0025] S5. Based on the fault parameter estimation results, the degree of degradation of hydraulic energy transmission efficiency, the degree of degradation of mechanical energy transmission efficiency, the degree of degradation of the overall system energy transmission efficiency, and the vibration characteristic frequency distribution, the faulty components, fault types, and fault severity in the hydraulic system are judged and the diagnostic results are output. When outputting diagnostic results, follow these steps: S51. First, calculate the fault severity coefficient: In the formula, For the first Failure severity coefficient of each component For the first Estimated fault parameters for each component This is the initial design clearance or reference clearance for the component under healthy conditions; For hydraulic component serial numbers; It is a dimensionless ratio. This indicates that the estimated gap is equal to the healthy baseline gap. A value greater than 1 indicates that the gap is larger than that of a healthy state; S52, according to The fault severity is classified into normal, minor, moderate, and severe faults based on the threshold range in which it is located. S53. Based on the calculated hydraulic energy transmission efficiency, mechanical energy transmission efficiency and system comprehensive energy transmission efficiency, calculate the energy transmission efficiency degradation. When the degradation of hydraulic energy transmission efficiency is large, while the degradation of mechanical energy transmission efficiency is small, it indicates that hydraulic energy loss is dominant, and the fault is mainly internal leakage. When the mechanical energy transmission efficiency degrades significantly while the hydraulic energy transmission efficiency degrades only slightly, it indicates increased mechanical vibration dissipation, and the fault is primarily determined to be a wear fault. When both hydraulic energy transmission efficiency and mechanical energy transmission efficiency decrease significantly, it indicates that hydraulic energy loss and mechanical energy loss exist simultaneously, and it is judged as a combined leakage-wear fault. S54. Combining the fault severity coefficient, energy transfer efficiency degradation characteristics, and vibration characteristic frequency distribution, comprehensively determine the faulty component, fault type, and fault severity, and output the diagnostic results.

[0026] The diagnostic results are shown in the table below: Table 1 Fault Type Determination Rules The efficiency of hydraulic energy transmission degrades significantly. Abnormal pressure difference, increased leakage energy Internal leakage fault Mechanical energy transfer efficiency degrades significantly. Increased dominant vibration frequency and higher vibration energy Wear and tear Overall energy transfer efficiency has decreased significantly. Simultaneous abnormal leakage and vibration Leakage-wear combined fault

[0027] Table 2 Fault Severity Classification Table

[0028] To achieve the above diagnostic method, the present invention also provides a hydraulic system fault diagnosis system based on energy transfer efficiency, comprising: The data acquisition unit is used to collect hydraulic system operating status data, including hydraulic pump inlet and outlet pressures, hydraulic valve inlet and outlet pressures, hydraulic cylinder two-chamber pressures, hydraulic pump speed, and vibration acceleration signals at the hydraulic pump housing, hydraulic valve block, and hydraulic cylinder mounting base. The unit performs low-pass filtering, time synchronization processing, and fixed time window segmentation processing on the pressure signals, speed signals, and vibration signals to obtain status detection data for fault diagnosis. The fault mechanism model unit is used to establish the correspondence between the fault parameters of hydraulic pumps, hydraulic valves and hydraulic cylinders and the leakage flow and vibration response based on the fault mechanism of key components of the hydraulic system. Among them, the fault parameters of hydraulic pumps include the equivalent wear clearance of piston pair, flow distribution pair and slipper pair; the fault parameters of hydraulic valves include the equivalent radial clearance between valve core and valve body; and the fault parameters of hydraulic cylinders include the equivalent leakage clearance between piston and cylinder. The fault dissipation energy calculation unit is used to calculate the leakage dissipation energy based on the pressure difference between the upstream and downstream of the hydraulic component and the leakage flow rate, calculate the vibration dissipation energy based on the vibration acceleration amplitude, characteristic angular frequency and equivalent vibration mass, and integrate the leakage dissipation energy and the vibration dissipation energy to obtain the total fault dissipation energy of the system. The fault parameter estimation unit based on energy transfer efficiency is used to establish a nonlinear state-space model with the equivalent fault parameters of hydraulic pumps, hydraulic valves and hydraulic cylinders as state variables, and the hydraulic energy transfer efficiency, mechanical energy transfer efficiency and system comprehensive energy transfer efficiency as observations. The unscented Kalman filter algorithm is then used to recursively estimate the fault parameters. The fault diagnosis output unit based on energy transfer efficiency is used to judge the faulty components, fault types and fault severity in the hydraulic system and output diagnostic results based on the estimated fault parameters, the degree of degradation of hydraulic energy transfer efficiency, the degree of degradation of mechanical energy transfer efficiency, the degree of degradation of the overall system energy transfer efficiency and the vibration characteristic frequency distribution.

[0029] To verify the closed-loop diagnostic process of "data acquisition - mechanism modeling - fault dissipation energy calculation - energy transfer efficiency evaluation - parameter estimation - fault diagnosis" proposed in this invention, a system-level fault diagnosis is performed using a certain type of valve-controlled hydraulic system as an example, and the parameter estimation process of the unscented Kalman filter is explained in detail to verify the feasibility and effectiveness of the method of this invention.

[0030] I. Data Collection Steps 1. The experimental object is a certain type of valve-controlled hydraulic system, which consists of an axial piston pump, an electro-hydraulic directional valve, a double-acting hydraulic cylinder and a load mechanism.

[0031] 2. The hydraulic pump has a rated pressure of 25 MPa and a rated speed of 1500 r / min.

[0032] 3. Pressure sensors are installed at the hydraulic pump outlet, hydraulic valve inlet and outlet, and both chambers of the hydraulic cylinder; speed sensors are installed on the hydraulic pump spindle; and triaxial acceleration sensors are installed on the hydraulic pump housing, hydraulic valve block, and hydraulic cylinder mounting base.

[0033] 4. The system sampling frequency is set to 10 kHz, and the length of a single diagnostic time window Δt is 0.1 s.

[0034] The system collected the following operational data under combined fault conditions: Hydraulic pump outlet pressure 21.6 MPa Pump outlet pressure Hydraulic valve outlet pressure 18.9 MPa Post-valve pressure Hydraulic cylinder return chamber pressure 2.4 MPa Hydraulic cylinder return oil pressure Hydraulic pump speed 1480 r / min System operating speed Hydraulic pump vibration amplitude 14.2 m / s² acceleration amplitude at the plunger's passing frequency Hydraulic valve vibration amplitude 8.6 m / s² acceleration amplitude at the valve core operating frequency Hydraulic cylinder vibration amplitude 5.1 m / s² acceleration amplitude at reciprocating impact frequency

[0035] II. Fault Mechanism Modeling Steps Based on the leakage mechanism of key components of the hydraulic system, leakage models for hydraulic pumps, hydraulic valves, and hydraulic cylinders are established.

[0036] Hydraulic pump leakage flow model: , Hydraulic valve annular gap leakage model: , Hydraulic cylinder leakage model: , in , , These represent the equivalent wear clearances of the hydraulic pump, hydraulic valve, and hydraulic cylinder, respectively. Since leakage flow has a cubic relationship with wear clearance, changes in clearance will lead to significant changes in energy loss. , and These represent the leakage flow rates of the hydraulic pump, hydraulic valve, and hydraulic cylinder, respectively, with Δp being the pressure difference between the upstream and downstream of the corresponding component. The dynamic viscosity of hydraulic oil. This refers to the sealing length of the mating surface between the valve core and the valve sleeve of the hydraulic valve; the proportional coefficients omitted in various formulas are determined by the structural dimensions of the corresponding components and the form of the leakage path.

[0037] III. Fault Dissipation Energy Calculation Steps Based on the structural parameters of the piston pump, hydraulic valve, and hydraulic cylinder, and combined with the leakage model, the following leakage flow rate and leakage dissipation energy are calculated: hydraulic pump <![CDATA[3.1×10 -5 ]]> 2.7 8.37 hydraulic valve <![CDATA[1.7×10 -6 ]]> 16.5 2.81 hydraulic cylinder <![CDATA[1.2×10 -6 ]]> 2.4 0.29

[0038] Total energy dissipated by system leakage: , Based on the vibration dissipation energy model: , hydraulic pump 0.42 The acceleration amplitude of the plunger at the frequency is increased. hydraulic valve 0.21 Increased acceleration amplitude at the valve core operating frequency hydraulic cylinder 0.09 Low-frequency reciprocating impact Total vibrational energy dissipation of the system: , Total energy dissipation due to system failure: + , The overall energy transfer efficiency of the system is significantly lower than that in a healthy state: = 0.9068, The overall energy transfer efficiency of the system under healthy conditions is obtained through experimental calibration or simulation and is used to calculate the degree of efficiency degradation. In this embodiment, the system input is assumed to be hydraulic energy. The weighting coefficient is 100 J. =0.8、 =0.2, then the hydraulic energy transfer efficiency, mechanical energy transfer efficiency, and overall system energy transfer efficiency can be obtained by normalizing the leakage dissipation energy and vibration dissipation energy, where This represents the total energy dissipated during system failures. and These represent the total energy dissipated by leakage and the total energy dissipated by vibration, respectively.

[0039] IV. Fault Parameter Estimation Steps Based on Unscented Kalman Filtering Equivalent wear clearance of hydraulic pump Equivalent radial clearance of hydraulic valve and the equivalent leakage gap of the hydraulic cylinder As a state variable: ,in For the first The state vector at each diagnostic moment The hydraulic energy transfer efficiency, mechanical energy transfer efficiency, and overall system energy transfer efficiency are calculated based on leakage dissipation energy and vibration dissipation energy, with energy transfer efficiency used as the observable metric. .in, For the first The observation vector at each diagnostic time. , and In order, they are hydraulic energy, mechanical energy, and overall system energy transfer efficiency.

[0040] 1. Initialization Let the initial parameters be as follows under healthy conditions:

[0041] Set the initial state vector: , Set the initial covariance matrix: , Process noise covariance matrix: , Observation noise covariance matrix: , in, The three components are arranged in the order of hydraulic pump, hydraulic valve, and hydraulic cylinder, and the numerical unit is μm; and The elements represent the initial state estimation error variance and the process noise variance, respectively, measured in squared gaps. The elements represent the variance of the three dimensionless efficiency observation noise.

[0042] 2. Sigma point generation State variable dimension Therefore, it generates There are Sigma points. Among them, The dimension of the state vector is given. In this embodiment, the three state components correspond to the hydraulic pump, hydraulic valve, and hydraulic cylinder, respectively. Represent a central Sigma point and along The state dimensions are distributed in positive and negative directions. There are Sigma points. To ensure that the initial Sigma points align with the dimensions in the table... The offset is consistent with μm, and in this embodiment, we take... , , ,thus and The corresponding weights are , as well as ( The sum of the mean weights is 1.

[0043]

[0044] The Sigma points mentioned above represent a set of possible states near the current state estimate, used to describe the uncertainty of the fault parameters. Centered at Sigma point, to For the remaining Sigma points, the three values ​​within the square brackets represent, in order: , and The unit is μm.

[0045] 3. State Prediction After completing several diagnostic cycle iterations, take the current time. The posterior state estimate is μm; the following example uses a recursion at this moment.

[0046] Since the wear parameters of hydraulic components change slowly, the state equation is taken as follows: ,in, For the first Time-state vector The state vector from the previous time step. For from the first Time to the The process noise vector at each time step; the subscript indicates the discrete diagnostic time step. "" indicates that a random process disturbance is superimposed on the slowly changing state of the previous moment.

[0047] Therefore, the predicted mean state is: This indicates that the clearances of the hydraulic pump, hydraulic valves, and hydraulic cylinders have all increased significantly compared to their healthy state. Among them, Indicates the use of the first Time and previous information for the first The mean of the predicted state at time t.

[0048] 4. Observation and Prediction Substituting the predicted Sigma point into the fault mechanism model, we first obtain the predicted leakage dissipation energy and the predicted vibration dissipation energy. Then, based on the input hydraulic energy, we perform normalization calculations to obtain the predicted energy transfer efficiency observation.

[0049] The predicted observation mean was calculated as follows: .in, Indicates the use of the first The first time and previous information obtained Predict the average value of observations at any given time.

[0050] 5. Kalman Gain and State Update The actual observations are as follows: ,in, For the first The actual measured observation vector at any given time has three components: hydraulic energy transfer efficiency, mechanical energy transfer efficiency, and overall system energy transfer efficiency, all of which are dimensionless.

[0051] The observation residuals are: ; , The three residual components correspond to hydraulic energy, mechanical energy, and the overall energy transfer efficiency of the system, respectively, and the residuals are dimensionless.

[0052] After calculating the Kalman gain, the state variables are corrected, and the final state is estimated. , in, For the first The fault state estimation vector after observation correction at any time, with three components 24, 15 and 14 representing the equivalent fault gaps of the hydraulic pump, hydraulic valve and hydraulic cylinder respectively.

[0053] This indicates that the hydraulic pump is the most severely worn, the hydraulic valve has obvious uneven wear, and the hydraulic cylinder has slight leakage.

[0054] V. Fault Diagnosis Output Steps Based on the fault parameter estimation results and energy transfer efficiency degradation characteristics analysis: , The overall energy transfer efficiency of the system is significantly lower than that in a healthy state: , Hydraulic energy transmission efficiency degradation: , Degradation of mechanical energy transfer efficiency: , Based on the fault parameter estimation results and energy transfer efficiency degradation characteristics analysis: the equivalent wear clearance of the hydraulic pump has increased significantly; the degradation of hydraulic energy transfer efficiency is dominant, indicating that leakage loss is the main source of the decrease in system energy transfer efficiency; the acceleration amplitude at the hydraulic pump plunger passing frequency and the acceleration amplitude at the hydraulic valve core actuation frequency have increased significantly.

[0055] Therefore, it can be determined that: the hydraulic pump has a moderate internal leakage fault; the hydraulic valve has a valve core wear fault; and the hydraulic cylinder has a slight internal leakage fault.

[0056] The final diagnosis results are consistent with the actual composite fault conditions, indicating that the hydraulic system fault diagnosis method based on energy transfer efficiency proposed in this invention can achieve unified modeling and dynamic parameter inversion of multi-component coupled faults in hydraulic systems.

[0057]

[0058] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of the present invention.

Claims

1. A fault diagnosis method for hydraulic systems based on energy transfer efficiency, characterized in that: Includes the following steps, S1. Collect hydraulic system operating status data, calculate the pressure difference between upstream and downstream of hydraulic components, determine the hydraulic pump operating frequency and related characteristic frequencies, and extract the vibration main frequency and characteristic frequency band energy. S2. Establish a fault mechanism model based on the structural characteristics and fault modes of hydraulic pumps, hydraulic valves, and hydraulic cylinders; S3. Calculate the leakage dissipation energy and vibration dissipation energy of hydraulic components based on the fault mechanism model, and then combine them to obtain the total fault dissipation energy of the system. S4. Using the equivalent fault parameters of hydraulic pumps, hydraulic valves and hydraulic cylinders as state variables, and hydraulic energy transfer efficiency, mechanical energy transfer efficiency and system comprehensive energy transfer efficiency as observations, a nonlinear state-space model is established, and the unscented Kalman filter algorithm is used to recursively estimate the fault parameters. S5. Based on the fault parameter estimation results, the degree of degradation of hydraulic energy transmission efficiency, the degree of degradation of mechanical energy transmission efficiency, the degree of degradation of the overall system energy transmission efficiency, and the vibration characteristic frequency distribution, the faulty components, fault types, and fault severity in the hydraulic system are judged and the diagnostic results are output.

2. The hydraulic system fault diagnosis method based on energy transfer efficiency according to claim 1, characterized in that: In step S1, the hydraulic system operating status data includes the inlet and outlet pressures of the hydraulic pump, the inlet and outlet pressures of the hydraulic valve, the pressures of the two chambers of the hydraulic cylinder, the speed of the hydraulic pump, and the vibration acceleration signals at the hydraulic pump housing, the valve block of the hydraulic valve, and the mounting base of the hydraulic cylinder.

3. The hydraulic system fault diagnosis method based on energy transfer efficiency according to claim 1, characterized in that: In step S1 and step S2, when establishing the fault mechanism model, an internal leakage model is established based on the structural characteristics and fault modes of the hydraulic pump, hydraulic valve, and hydraulic cylinder. The internal leakage model includes a hydraulic pump leakage model, a hydraulic valve leakage model, and a hydraulic cylinder leakage model.

4. The hydraulic system fault diagnosis method based on energy transfer efficiency according to claim 1, characterized in that: In step S3, the leakage dissipation energy is calculated based on the pressure difference between the upstream and downstream of the hydraulic component and the leakage flow rate. The vibration dissipation energy is calculated based on the vibration acceleration amplitude, characteristic angular frequency, and equivalent vibration mass. The total system failure dissipation energy is the sum of the leakage dissipation energy and vibration dissipation energy of each hydraulic component.

5. The hydraulic system fault diagnosis method based on energy transfer efficiency according to claim 1, characterized in that: In step S4, the hydraulic energy transmission efficiency, mechanical energy transmission efficiency, and system comprehensive energy transmission efficiency are obtained by normalizing the fault dissipation energy and the system input hydraulic energy, respectively. The hydraulic energy transfer efficiency is used to represent the degree of hydraulic energy loss caused by leakage faults in the hydraulic system; the mechanical energy transfer efficiency is used to represent the degree of mechanical energy dissipation caused by wear and other faults in the hydraulic system; the overall system energy transfer efficiency is used to represent the overall energy transfer capability of the hydraulic system. The overall energy transfer efficiency of the system is obtained using a weighted fusion method: ， in, For hydraulic energy transmission efficiency, For mechanical energy transfer efficiency, For the overall energy transfer efficiency of the system, , These are the weighting coefficients, and .

6. The hydraulic system fault diagnosis method based on energy transfer efficiency according to claim 1, characterized in that: In step S4, the unscented Kalman filter algorithm includes the following steps: S41. Set the initial state estimate, initial covariance matrix, process noise covariance matrix, and observation noise covariance matrix: ， ， ， ， In the formula, These are the initial estimated fault parameters; These are the equivalent clearances for the health benchmarks of hydraulic pumps, valves, and cylinders, respectively. The initial state covariance matrix; This represents the true state vector of the system at the initial moment. The process noise covariance matrix; This is process noise; To observe noise; To observe the noise covariance matrix; S42. Generate based on the current state estimate and covariance matrix. Sigma points: ， ， ， In the formula, For the first One Sigma point; The first square root of the matrix List; For the first The state covariance matrix at time t, Let the dimension of the state variables be . Let be the scale parameter of the Sigma point distribution. For Sigma point number; This indicates taking the first square root of the matrix. List; S43. Substitute each Sigma point into the state transition function to obtain the predicted Sigma point: ， And calculate the mean of the state prediction: In the formula, For the first One predicted Sigma point; For the first Predicted values ​​of fault parameters at any given time; This is the state transition function. For the first The system input or known control variable at a given time. For the previous version Sigma points, The first state obtained through state transition One predicted Sigma point; For the first The mean weight of each Sigma point; S44. Calculate the prediction covariance matrix: In the formula, The state prediction covariance matrix; For the first Covariance weights of Sigma points; This represents the outer product of the vectors representing the deviations of the predicted Sigma points from the predicted mean. The process noise covariance matrix is ​​represented by the superscript. Indicates transpose; S45. Substitute the predicted Sigma points into the energy transfer efficiency observation function, and first calculate the leakage dissipation energy based on the equivalent fault parameters corresponding to each predicted Sigma point. and vibration dissipation of energy Then, it is normalized into hydraulic energy transmission efficiency, mechanical energy transmission efficiency and system comprehensive energy transmission efficiency. The predicted value of hydraulic energy transmission efficiency is expressed as: , The predicted value of mechanical energy transfer efficiency is expressed as: , The predicted value of the overall system energy transfer efficiency is expressed as follows: , Therefore, we obtain the first... Predicted observations corresponding to each Sigma point: , Calculate the observed and predicted mean: In the formula, This is the observation vector for predicting energy transfer efficiency; where, and The first The leakage dissipation energy and vibration dissipation energy corresponding to each predicted Sigma point. Input hydraulic energy into the system within the same diagnostic time window; , and These represent the hydraulic energy, mechanical energy, and overall system energy transfer efficiency corresponding to the Sigma point, all of which are dimensionless quantities. The vector is composed of the three efficiencies mentioned above, in the order of hydraulic energy, mechanical energy, and overall system energy. Square brackets indicate a vector. Weighted by mean, To predict the observed mean; S46. Calculate the observation-prediction covariance matrix: In the formula, To observe the covariance matrix; Calculate the cross-covariance matrix between the state variables and the observed variables: ， In the formula, This indicates the correlation between changes in fault parameters and changes in energy transfer efficiency. For the first The predicted observation vector of Sigma points, and the aforementioned Same meaning; For the first The predicted observation mean at time, compared with the aforementioned Same meaning; To observe and predict the covariance matrix, To observe the noise covariance matrix; This is the cross-covariance matrix between the state and the observation; For covariance weights, Indicates to to Summation: Multiply the two deviation vectors and add a superscript. Indicates the outer product; S47. Calculate the Kalman gain: In the formula, This is the Kalman gain matrix, used to determine the degree of influence of real-time observations on state corrections; Represents the observation-prediction covariance matrix inverse matrix, matrix product Used to map observation residuals to state corrections. subscript Indicates the first A diagnostic moment; S48. Based on real-time observation vectors With predicted observation vector The deviation updates the state variables: ， In the formula, For the first The fault parameter estimates after real-time correction; For the first The measured energy transfer efficiency observation vector at any given time. For the corresponding predicted observation vector, To observe the residuals, This is the state correction amount. This is the prior state estimate before correction; " indicates that the correction amount is added to the prior state estimate; If the real-time observed energy transfer efficiency is lower than the predicted energy transfer efficiency, it indicates that the system's energy transfer capability has decreased more than the predicted value, and the current fault parameters may be underestimated. The algorithm corrects this by increasing the Kalman gain. If the real-time observed energy transfer efficiency is higher than the predicted energy transfer efficiency, it indicates that the decrease in the system's energy transfer capacity is less than the predicted value, and the current fault parameters may be overestimated. The algorithm will then reduce and correct this. S49, Covariance Update: In the formula, For the first The state covariance matrix updated at each time step; For the first The predicted state covariance matrix before the time update. To observe and predict the covariance matrix, Kalman gain matrix transpose; The minus sign indicates that the uncertainty was subtracted from the predicted covariance to represent the uncertainty eliminated after this observation. The superscript indicates that the uncertainty was removed from the predicted covariance. This indicates transpose.

7. A hydraulic system fault diagnosis method based on energy transfer efficiency according to claim 5, characterized in that: In step S5, when outputting the diagnostic results, the following steps are performed. S51. First, calculate the fault severity coefficient: In the formula, For the first Failure severity coefficient of each component For the first Estimated fault parameters for each component This is the initial design clearance or reference clearance for the component under healthy conditions; For hydraulic component serial numbers; It is a dimensionless ratio. This indicates that the estimated gap is equal to the healthy baseline gap. A value greater than 1 indicates that the gap is larger than that of a healthy state; S52, according to The fault severity is classified into normal, minor, moderate, and severe faults based on the threshold range in which it is located. S53. Based on the calculated hydraulic energy transmission efficiency, mechanical energy transmission efficiency and system comprehensive energy transmission efficiency, calculate the energy transmission efficiency degradation. When the degradation of hydraulic energy transmission efficiency is large, while the degradation of mechanical energy transmission efficiency is small, it indicates that hydraulic energy loss is dominant, and the fault is mainly internal leakage. When the mechanical energy transmission efficiency degrades significantly while the hydraulic energy transmission efficiency degrades only slightly, it indicates increased mechanical vibration dissipation, and the fault is primarily determined to be a wear fault. When both hydraulic energy transmission efficiency and mechanical energy transmission efficiency decrease significantly, it indicates that hydraulic energy loss and mechanical energy loss exist simultaneously, and it is judged as a combined leakage-wear fault. S54. Combining the fault severity coefficient, energy transfer efficiency degradation characteristics, and vibration characteristic frequency distribution, comprehensively determine the faulty component, fault type, and fault severity, and output the diagnostic results.

8. A fault diagnosis system for a hydraulic system fault diagnosis method based on energy transfer efficiency according to any one of claims 1-7, characterized in that: include, The data acquisition unit is used to collect hydraulic system operating status data, including hydraulic pump inlet and outlet pressures, hydraulic valve inlet and outlet pressures, hydraulic cylinder two-chamber pressures, hydraulic pump speed, and vibration acceleration signals at the hydraulic pump housing, hydraulic valve block, and hydraulic cylinder mounting base. It also performs low-pass filtering, time synchronization processing, and fixed time window segmentation processing on the pressure signal, speed signal, and vibration signal to obtain status detection data for fault diagnosis. The fault mechanism model unit is used to establish the correspondence between the fault parameters of hydraulic pumps, hydraulic valves and hydraulic cylinders and the leakage flow and vibration response based on the fault mechanism of key components of the hydraulic system. Among them, the fault parameters of hydraulic pumps include the equivalent wear clearance of piston pair, flow distribution pair and slipper pair; the fault parameters of hydraulic valves include the equivalent radial clearance between valve core and valve body; and the fault parameters of hydraulic cylinders include the equivalent leakage clearance between piston and cylinder. The fault dissipation energy calculation unit is used to calculate the leakage dissipation energy based on the pressure difference between the upstream and downstream of the hydraulic component and the leakage flow rate, calculate the vibration dissipation energy based on the vibration acceleration amplitude, characteristic angular frequency and equivalent vibration mass, and integrate the leakage dissipation energy and the vibration dissipation energy to obtain the total fault dissipation energy of the system. The fault parameter estimation unit based on energy transfer efficiency is used to establish a nonlinear state-space model with the equivalent fault parameters of hydraulic pumps, hydraulic valves and hydraulic cylinders as state variables, and the hydraulic energy transfer efficiency, mechanical energy transfer efficiency and system comprehensive energy transfer efficiency as observations. The unscented Kalman filter algorithm is then used to recursively estimate the fault parameters. The fault diagnosis output unit based on energy transfer efficiency is used to judge the faulty components, fault types and fault severity in the hydraulic system and output diagnostic results based on the estimated fault parameters, the degree of degradation of hydraulic energy transfer efficiency, the degree of degradation of mechanical energy transfer efficiency, the degree of degradation of the overall system energy transfer efficiency and the vibration characteristic frequency distribution.

Images