Diaphragm type energy accumulator air tightness detection method

Abstract

The invention relates to the technical field of air tightness detection, in particular to a diaphragm type energy accumulator air tightness detection method which is used for solving the problems that in the prior art, equipment operation characteristics cannot be accurately described, typical defect modes cannot be recognized in combination with spectral clustering, and a quantitative basis cannot be provided for equipment fault trend analysis and intelligent maintenance. The method comprises the following steps: constructing a low-dimensional state map to describe equipment operation characteristics, identifying typical defect modes in combination with spectral clustering, establishing a nonlinear correlation model to reveal a defect evolution relationship, optimizing classification model parameters by adopting an evolutionary algorithm, improving the identification accuracy, mining a most probable defect evolution path based on a transition probability matrix, and improving the identification efficiency. And a quantitative basis is provided for equipment fault trend analysis and intelligent maintenance.

Description

Technical Field

[0001] This invention relates to the field of airtightness testing technology, and more specifically, to a method for testing the airtightness of a diaphragm accumulator. Background Technology

[0002] The airtightness of diaphragm accumulators is a core guarantee for the efficiency of hydraulic systems, but existing testing technologies have three major bottlenecks: the traditional gas pressure decay method cannot quantify the pre-charge gas pressure loss caused by diaphragm fatigue, making it difficult to assess long-term reliability; the water immersion method is inefficient and easily damages products, failing to meet the needs of mass production; the helium mass spectrometry leak detection method is costly and complex to operate, making it difficult to promote in industrial scenarios, and the lack of a quantitative correlation model between diaphragm fatigue characteristics and airtightness decay results in a lack of data support for material selection and structural optimization.

[0003] Referring to patent application CN119164562A, a method, system, and storage medium for testing the airtightness of a diaphragm accumulator are disclosed. The method includes: acquiring preparation completion trigger information of the diaphragm accumulator; acquiring test plan information of the diaphragm accumulator based on the preparation completion trigger information; controlling a preset hydraulic system to perform a liquid filling test on the diaphragm accumulator according to the test plan information to determine gas pressure detection information; analyzing the gas pressure detection information and the test plan information to determine the liquid-gas pressure information of the hydraulic system; determining whether the gas pressure detection information meets the requirements of the liquid-gas pressure information; if it does not meet the requirements, issuing a prompt according to a preset non-compliance prompt; if it meets the requirements, issuing a prompt according to a preset compliance prompt. This application has the effect of improving the accuracy of airtightness testing of diaphragm accumulators. However, the aforementioned reference patent improves the airtightness and gas pressure detection accuracy by controlling the hydraulic system to test the filling of the diaphragm accumulator and comparing gas pressure and filling time information. When the test pattern does not conform to the step pattern, this method is used to fill the accumulator to improve filling convenience. However, it cannot accurately reconstruct the internal structural characteristics of the diaphragm accumulator, cannot accurately identify crack propagation behavior, reduces the sensitivity to defect features of different scales, and cannot provide technical support for equipment health status assessment and intelligent maintenance decision-making. At the same time, it cannot accurately characterize the equipment operating characteristics, cannot combine spectral clustering to identify typical defect patterns, and cannot provide quantitative basis for equipment failure trend analysis and intelligent maintenance.

[0004] To address the aforementioned issues, we propose a method for testing the airtightness of diaphragm accumulators. Summary of the Invention

[0005] The purpose of this invention is to provide a method for airtightness testing of diaphragm accumulators, which solves the problems of existing technologies that cannot accurately reconstruct the internal structural characteristics of diaphragm accumulators, cannot accurately identify crack propagation behavior, reduce sensitivity to defect features of different scales, and cannot provide technical support for equipment health status assessment and intelligent maintenance decision-making; at the same time, they cannot accurately characterize equipment operating characteristics, cannot combine spectral clustering to identify typical defect patterns, and cannot provide quantitative basis for equipment failure trend analysis and intelligent maintenance.

[0006] The objective of this invention is achieved through the following technical solution: A method for testing the airtightness of a diaphragm accumulator includes the following steps: Step 1: Collect status monitoring data during the operation of the diaphragm accumulator. The status monitoring data includes gas pressure data, temperature data, vibration data, acoustic emission signals, and strain data. Perform preprocessing operations on the collected status monitoring data. Step 2: Real-time monitoring and closed-loop control of gas flow rate, gas pressure, and gas path status during the inflation process, and automatic switching of gas path according to detection requirements; Step 3: By integrating gas pressure, temperature, vibration and acoustic emission signals, and using acoustic wave propagation inversion, dynamic crack identification and multi-scale analysis methods, the internal defects of the accumulator are preliminarily located and their evolution trend is analyzed. Step 4: Construct a high-dimensional map based on condition monitoring data to characterize the nonlinear correlation between defect patterns, and optimize model parameters by combining evolutionary algorithms to achieve defect identification and evolution path analysis.

[0007] In a preferred embodiment of the present invention, the process of real-time monitoring and closed-loop control of gas flow rate, gas pressure, and gas path status during the inflation process in step two includes: Gas flow rate is collected using a thermal flow meter, and the collected gas flow rate data is transmitted to the control unit. A piezoelectric pressure sensor is used to collect gas pressure, and the collected data is transmitted to the control unit. The gas path conduction status is collected using a switch status sensor, the valve opening degree is collected using a valve position sensor, and the gas flow status is collected using a differential pressure flow detection module. The collected data is then connected to the control unit. The control system consists of sensors, a control unit, and an actuator. The control unit receives sensor signals from gas flow, gas pressure, and gas path status, compares the received actual measured values ​​with preset target values, calculates the deviation, and generates a control output signal based on the deviation to drive the actuator to adjust the system state. The control unit uses a proportional-integral-derivative control algorithm to perform closed-loop regulation of gas flow and gas pressure, and changes the gas flow rate, adjusts the gas pressure, or switches the gas path according to preset program instructions.

[0008] In a preferred embodiment of the present invention, the process of automatically switching the gas path according to the detection requirements in step two includes: The control system determines whether to perform a gas path switch based on the task identifier and operating parameters. After identifying the switching conditions, the control system generates a switching command. The control system selects the target air path and issues a switching command, and the air path switching valve responds to the command and changes the airflow direction. The various components of the control system work together, and the air path switching mechanism guides the airflow into different paths according to the task identifier; The control system enables path switching and parameter adjustment of the gas path system.

[0009] In a preferred embodiment of the present invention, the process of fusing gas pressure, temperature, vibration, and acoustic emission signals and performing acoustic wave propagation inversion analysis in step three includes: Preprocessed gas pressure data, temperature data, vibration data, and acoustic emission signals are acquired, and their features are extracted. The extracted features are then normalized and weighted and fused into a comprehensive feature vector F. The material's internal physical properties or structural information can be deduced from the acoustic wave signals measured at the boundary using the acoustic wave propagation inversion method. In an isotropic linear elastic medium, sound waves satisfy a three-dimensional second-order wave equation; When the location of the sound source and the parameters of the medium are known, the wave equation is solved using the finite difference method to obtain the theoretical wave field u(x,t); The goal of the inversion is to minimize the observed wavefield u gc (x i ,t) and the calculation of wave field u(x) i The difference between ,t;m) is used to optimize the medium parameter m(x); Introducing the adjoint field λ(x,t), the gradient of the objective function with respect to the medium parameters is calculated through the adjoint state equation; Using this gradient, the medium parameter m is iteratively updated using the L-BFGS algorithm until convergence.

[0010] In a preferred embodiment of the present invention, the process of employing dynamic crack identification and multi-scale analysis methods in step three includes: Dynamic crack propagation causes a decrease in material stiffness, which manifests as a decrease in structural modal frequency and an increase in damping ratio. The location of the crack can be identified by combining local energy density changes. The response signal is processed using the Hilbert-Huang transform to extract the nth instantaneous frequency f.n (t) and damping ratio ζ n (t); By introducing a local energy density function, the energy concentration increases during crack propagation, and the crack location is determined by combining the change in wave propagation path. Structural features at different scales are processed by multi-scale analysis methods, and cross-scale feature extraction is achieved by wavelet decomposition. Define micro variables ε is a small parameter that is much smaller than 1. The field quantity u has a multi-scale expansion form. Substitute the expansion into the control equation and separate the variables according to ε, and solve it step by step. Using wavelet basis function ψ j,k (x) Perform multi-scale decomposition on the signal.

[0011] In a preferred embodiment of the present invention, the process of preliminary location and evolution trend analysis of internal defects in the accumulator in step three includes: Defects in energy storage devices can be analyzed for their spatiotemporal evolution characteristics through multi-source data fusion. By inverting and reconstructing the internal reflection coefficient distribution r(x) of the accumulator, the distribution map of r(x) is obtained by inverse calculation of the measured echo s(t) to determine the location of the defect; The defect parameter D(t) is defined as the damage factor, and its evolution is defined as conforming to an exponential growth model. When the defect parameter D(t) reaches or exceeds the set threshold D sd When a defect is deemed to be in a dangerous state, repair or replacement measures must be taken.

[0012] In a preferred embodiment of the present invention, step four, which involves constructing a high-dimensional map based on condition monitoring data and characterizing the nonlinear correlation between defect modes, includes: Obtain the preprocessed state monitoring data, construct a time-series state vector from it, and build a state data matrix for the total observation time length T. Using a fixed window length and sliding step size, state information of consecutive time periods is extracted sequentially to form a series of trajectory segments. The total number of trajectories is calculated based on the set window length and sliding step size, and a trajectory set is generated. The high-dimensional trajectory is reduced using the diffusion mapping algorithm. The similarity between trajectories is measured by constructing a Gaussian kernel function and then normalized to obtain the transition probability matrix. The first few non-trivial eigenvectors of the matrix and their corresponding eigenvalues ​​are extracted to construct the node representation in the low-dimensional embedding space, and finally a set of low-dimensional nodes are obtained, which constitute the graph structure of the device state. For each pair of low-dimensional nodes, if one node is the nearest neighbor of the other node, then an edge is established between the two nodes, eventually constructing a weighted graph.

[0013] In a preferred embodiment of the present invention, the process of characterizing the nonlinear correlation between defect patterns in step four includes: The low-dimensional node set is classified using spectral clustering. Nodes with similar characteristics are grouped into the same category. Each category represents a defect pattern. The center point of each category is calculated as the representative state of the defect of that category. The squared Euclidean distance between the centers of various defects is calculated as a measure of their differences, and finally a nonlinear correlation matrix reflecting the relationship between the various defect modes is obtained. A threshold is set as the top 20 percentile of the similarity between all defect pairs. If the squared Euclidean distance between two defect categories is less than this threshold, a directed edge is established between the corresponding nodes, thereby constructing a directed association graph between defects, which represents the possible evolution path of the defects.

[0014] In a preferred embodiment of the present invention, the process of optimizing model parameters by combining evolutionary algorithms in step four includes: Design an input-output model with the device state vector as input and the defect category number as output. The model parameters are variables to be optimized. Define a comprehensive evaluation function, which includes the following three items: classification accuracy, loss function value and parameter norm. Each item is combined into an overall fitness index according to a set weight. The evolutionary search steps are as follows: T1: Initialize a batch of individuals with random parameters and calculate the fitness value of each individual; T2: Select the individuals with the highest fitness as parents, perform crossover to generate new individuals, and add a small mutation perturbation: T3: Repeat the above process until the preset number of iterations or the convergence condition is reached; The final output is the optimal model parameters.

[0015] In a preferred embodiment of the present invention, the process of defect identification and evolution path analysis in step four includes: Each state vector collected during equipment operation is input into the trained model, and the corresponding defect category label is output to generate a complete defect label sequence. The frequency of defect transfer between adjacent defect labels is statistically analyzed, and the frequency ratio of each defect to another defect is calculated to form a defect transfer probability matrix. Define a path sequence consisting of multiple defect categories, where the probability of a path is the product of the transition probabilities of all its adjacent defects. Among all possible paths, identify the one with the highest probability as the most likely defect evolution path.

[0016] Compared with the prior art, the advantages of this invention are: (1) In this invention, by integrating gas pressure, vibration, temperature and acoustic emission signals, key features are extracted and a comprehensive feature vector is constructed. The internal structural characteristics of the diaphragm accumulator are reconstructed by combining the acoustic wave propagation inversion method. Based on the dynamic crack identification model, the changes in modal frequency, damping ratio and energy distribution are analyzed to identify crack propagation behavior. Multi-scale wavelet analysis method is introduced to enhance the sensitivity to defect features at different scales and improve the accuracy of defect location and the ability to predict evolution trends. The whole process integrates physical models and data-driven methods to provide technical support for equipment health status assessment and intelligent maintenance decision-making. (2) In this invention, a low-dimensional state map is constructed to characterize the equipment operation characteristics, combined with spectral clustering to identify typical defect patterns, a nonlinear correlation model is established to reveal the defect evolution relationship, an evolutionary algorithm is used to optimize the classification model parameters to improve the identification accuracy, and the most likely defect evolution path is mined based on the transition probability matrix to provide a quantitative basis for equipment failure trend analysis and intelligent maintenance. Attached Figure Description

[0017] Figure 1 This is a flowchart of the airtightness testing method for the diaphragm accumulator in this invention; Figure 2 This is a flowchart of the evolutionary search steps in this invention. Detailed Implementation

[0018] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0019] Example 1: As Figure 1 As shown, the present invention proposes a method for testing the airtightness of a diaphragm accumulator, comprising the following steps: Step 1: Collect status monitoring data during the operation of the diaphragm accumulator. The status monitoring data includes gas pressure data, temperature data, vibration data, acoustic emission signals and strain data. Perform preprocessing operations on the collected status monitoring data, including signal denoising, outlier removal and time synchronization. By collecting multi-dimensional status monitoring data such as gas pressure, temperature, vibration, acoustic emission, and strain during the operation of the diaphragm accumulator, and performing preprocessing operations such as signal denoising, outlier removal, and time synchronization, comprehensive equipment operating status information is obtained, effectively improving data quality and consistency. High-quality data not only provides a reliable foundation for subsequent status identification, fault diagnosis, and life prediction, but also helps to capture early subtle abnormal changes in the equipment and achieve early warning of faults.

[0020] Step 2: Real-time monitoring and closed-loop control of gas flow rate, gas pressure, and gas path status during the inflation process, and automatic switching of gas path according to detection requirements; Step two involves real-time monitoring and closed-loop control of gas flow rate, gas pressure, and gas path status during the inflation process, including: A thermal flow meter is used to collect gas flow. The sensor outputs an electrical signal based on the change in heat. The collected gas flow data is transmitted to the control unit, which adjusts the opening of the regulating valve according to the current flow value. A piezoelectric pressure sensor is used to collect gas pressure. The sensor outputs an electrical signal through voltage changes. The collected data is transmitted to the control unit, which compares the current gas pressure value with the set value and performs adjustment operations. A switch status sensor is used to collect the gas path continuity status. The sensor outputs a binary signal to indicate whether the gas path is open. A valve position sensor is used to collect the valve opening. The sensor outputs a continuous signal to indicate the valve position. A differential pressure flow detection module is used to collect the gas flow status. The sensor output signal is used to determine whether the gas is flowing. The collected data is connected to the control unit to determine whether there is an abnormality in the gas path. The control system consists of sensors, control units, and actuators. The control unit receives sensor signals from gas flow, gas pressure, and gas path status, compares the received actual measured values ​​with preset target values, calculates the deviation, generates a control output signal based on the deviation, and drives the actuators to adjust the system state. The adjustment process is completed based on a feedback mechanism. The control unit uses a proportional-integral-derivative control algorithm to perform closed-loop regulation of gas flow and gas pressure, and changes the gas flow rate, adjusts the gas pressure, or switches the gas path according to preset program instructions. Proportional term: Generates control quantity based on the current error; Integral term: Generates corrections based on accumulated errors; Differential term: Generates dynamic compensation amount based on the rate of change of error; The controller parameters are set according to preset rules; The actuators include airflow regulating valves, pressure regulating valves, and air path switching valves; Step two involves automatically switching the gas path according to the detection requirements, including: The control system determines whether to perform a gas path switch based on the task identifier and operating parameters. The task identifier is provided by external input, and the operating parameters include gas pressure, flow rate and timestamp. The control system generates a switch command after recognizing the switching conditions. The control system selects the target gas path and issues a switching command. The gas path switching valve responds to the command and changes the airflow direction. After the switching is completed, it continues to collect gas flow rate, gas pressure and gas path status data. The various components in the control system work together. The sensors are responsible for collecting data, the control unit is responsible for processing the data and generating control signals, the actuators adjust according to the control signals, and the air path switching mechanism guides the airflow into different paths according to the task identifier. The control system enables path switching and parameter adjustment of the gas circuit system, and the control unit executes subsequent operations based on the collected data. By collecting real-time gas flow, pressure, and gas path status information through sensors, and combining this with a closed-loop control mechanism based on PID algorithm, the system achieves dynamic adjustment and stable control of gas parameters, ensuring the safety and accuracy of the inflation process. Simultaneously, the system automatically switches gas paths based on task identifiers and operating parameters, enhancing its flexibility and adaptability. The collaborative work of all components forms a "perception-decision-execution" closed loop, which not only improves control response speed and adjustment accuracy but also supports automated operation under multiple working conditions, making it suitable for gas path management and intelligent control in complex detection tasks.

[0021] Step 3: By integrating gas pressure, temperature, vibration and acoustic emission signals, and using acoustic wave propagation inversion, dynamic crack identification and multi-scale analysis methods, the internal defects of the accumulator are preliminarily located and their evolution trend is analyzed. Step three, which involves fusing gas pressure, temperature, vibration, and acoustic emission signals and performing acoustic wave propagation inversion analysis, includes: Preprocessed gas pressure data, temperature data, vibration data, and acoustic emission signals were acquired, and feature extraction was performed on them. The feature extraction process is as follows: Extract the mean, variance, and average of the absolute values ​​of the differences between adjacent time points from the gas pressure data; Extract the maximum, minimum, standard deviation, and rate of temperature change from the temperature data; Vibration data is used to extract peak value, RMS value, peak-to-peak value, kurtosis, dominant frequency, and amplitude. The acoustic emission signal is extracted by pulse count, number of impacts, maximum amplitude, dominant frequency, and energy distribution. The extracted features are normalized and weighted to form a comprehensive feature vector F. The weight coefficients are obtained from historical data training and are expressed as follows: F=[f1,f2,…f n ] T , where fi represents the i-th feature component in the comprehensive feature vector, n represents the feature dimension, and F represents the overall physical characteristics of the current system state; The material's internal physical properties or structural information can be deduced from the acoustic wave signals measured at the boundary using the acoustic wave propagation inversion method. In an isotropic linear elastic medium, sound waves satisfy a three-dimensional second-order wave equation, expressed by the following expression: ; Where u(x,t) represents the displacement field of the sound wave at position x and time t, and c is the wave velocity in the medium. 2 It is the Laplace operator; When the location of the sound source and the parameters of the medium are known, the wave equation is solved using the finite difference method to obtain the theoretical wave field u(x,t); The goal of the inversion is to minimize the observed wavefield u gc (x i ,t) and the calculation of wave field u(x) i The difference between ,t;m) is used to optimize the medium parameter m(x), and the comprehensive eigenvector F is used to set the range of values ​​of the medium parameter m(x) in the process of solving the wave equation, as the initial constraint condition for inversion; The objective function is expressed by the following expression: , where J(m) is the objective function, representing the energy of the difference between the observed and calculated wavefield, m(x) is the distribution of medium parameters, N is the number of sensors, and T is the total observation time; Introducing the adjoint field λ(x,t), the gradient of the objective function with respect to the medium parameters is calculated using the adjoint state equation: ; Using this gradient, the medium parameter m is iteratively updated using the L-BFGS algorithm until convergence; Step three, which employs dynamic crack identification and multi-scale analysis methods, includes the following: Dynamic crack propagation leads to a decrease in material stiffness, manifested as a decrease in structural modal frequencies and an increase in damping ratio. The location of the crack can be identified by combining local energy density changes. The structural dynamics equations are as follows: Where M is the mass matrix, C is the damping matrix, K(t) is the stiffness matrix that varies with crack propagation, f(t) is the external excitation, and u(t) is the displacement vector. These are the first and second time derivatives of the displacement; When the value of the comprehensive feature vector F changes abruptly or deviates from the historical average by more than a set threshold, the system is determined to be in an abnormal operating state, triggering the start of the dynamic crack identification module. This vector does not participate in the calculation of modal frequency or damping ratio, but is only used as a trigger condition for crack identification. The response signal is processed using the Hilbert-Huang transform to extract the nth instantaneous frequency f. n (t) and damping ratio ζ n (t): , where m n It is the nth-order quality coefficient, c n It is the nth order damping coefficient, k n (t) is the nth-order stiffness coefficient; Crack propagation causes stiffness coefficient k n The decrease in (t) leads to a decrease in modal frequency and an increase in damping ratio; Introducing the local energy density function: ,in It is a displacement-velocity vector; The energy concentration increases during crack propagation, and the crack location is determined by combining the changes in the wave propagation path. Discrete wavelet transform is used to perform a three-level decomposition of the signal. Daubechiesdb4 wavelet basis functions are selected to extract structural features at both macroscopic and microscopic scales, and microscopic variables are defined. Where ε is much smaller than 1 and is a small parameter, the field quantity u has a multi-scale expansion form: u(x,t)=u0(x,ξ,t)+εu1(x,ξ,t)+ε 2 u2(x,ξ,t)+…; Substitute the expansion into the control equation and separate the variables by expanding by ε, then solve step by step; Using wavelet basis functions ψ j,k (x) Perform multi-scale decomposition on the signal: Where j controls the scale, k controls the translation, and c j,k These are wavelet coefficients, reflecting local features at different scales; The comprehensive eigenvector F reflects the changing trend of the overall system state. During wavelet decomposition, analysis weights are set at different scales based on the changes in the eigenvector. For example, when the eigenvector shows enhanced high-frequency energy, the wavelet coefficients in the high-frequency band are analyzed first. Step three, which involves preliminary location and evolution trend analysis of internal defects in the accumulator, includes: Defects in accumulators (such as cavitation, fatigue cracks, or corrosion pits) can be analyzed for their spatiotemporal evolution characteristics through multi-source data fusion. Reconstructing the internal reflection coefficient distribution r(x) of the accumulator using inversion: , where G(x s (x,t) is the position of the sound source x sGreen's function to point x, G(x,x) r ,t) is the distance from point x to receiver position x. r The Green's function is given by s(t), where s(t) is the echo signal received at the receiver, and r(x) is the reflection coefficient distribution, representing the defect intensity. For the measured echo s(t), the distribution map of r(x) is obtained by inverse calculation to determine the location of the defect; Define the defect parameter D(t) as the damage factor, and define its evolution as conforming to an exponential growth model: D(t) = D0e θt Where D0 is the initial defect level and θ is the growth coefficient, which is obtained by experimental fitting. In practical applications, the value of θ is affected by the current working conditions and can be obtained by linear mapping of comprehensive feature vectors. When the defect parameter D(t) reaches or exceeds the set threshold D sd When a defect is deemed to be in a dangerous state, repair or replacement measures must be taken. Here is a simulation example of acoustic wave propagation inversion for defect localization based on the content of step three: Consider an application scenario of an energy storage device in a wind turbine generator set. This energy storage device is used to store and release energy to balance the demand of the power grid. Due to the long-term operation in an environment with high pressure and frequent temperature changes, the energy storage device is prone to internal defects such as fatigue cracks or corrosion points. In order to ensure the safe operation of the system, it is necessary to conduct health monitoring of the energy storage device regularly. First, multiple acoustic sensors are installed on the outside of the accumulator to collect acoustic signals at the boundary; at the same time, real-time data is obtained using built-in pressure, temperature and vibration sensors. After preprocessing the collected data, feature extraction is performed according to the method described in step three, and these features are fused into a comprehensive feature vector F; Based on the known location of the sound source and medium parameters (such as wave velocity c), the wave equation is solved using the finite difference method. By utilizing the difference between the observed wavefield and the calculated wavefield, the medium parameter m(x) is optimized through the process of minimizing the objective function, and the L-BFGS algorithm is used to iteratively update it until convergence. Finally, based on the reconstructed internal reflection coefficient distribution r(x) of the accumulator, the location of potential defects is determined; When a sudden change is detected in the comprehensive feature vector F, the dynamic crack identification module is activated to analyze the changes in frequency, mode and energy characteristics. Wavelet decomposition was used to analyze the structural features at different scales to determine the specific location and development trend of cracks. After the above steps, the location results of the internal defects of the accumulator and the curve of the defect parameter D(t) changing with time were obtained. Table 1 below shows the defect parameter values ​​at some key moments and their corresponding dangerous state judgments: Table 1 Defect Parameter Information The set danger threshold is 0.08. As can be seen from the table, on the 100th day, the defect parameter D(t) of the accumulator exceeded the set danger threshold, indicating that immediate repair or replacement measures are needed to avoid an accident. By fusing heterogeneous signals from multiple sources, such as gas pressure, temperature, vibration, and acoustic emission, key features are extracted and a comprehensive feature vector is constructed. Combined with acoustic wave propagation inversion methods, the internal structural characteristics of the diaphragm accumulator are reconstructed. A dynamic crack identification model is used to analyze changes in modal frequencies, damping ratios, and energy distribution to identify crack propagation behavior. At the same time, a multi-scale wavelet analysis method is introduced to enhance the sensitivity to defect features at different scales and improve the accuracy of defect location and the ability to predict evolution trends. The entire process is based on a combination of physical models and data-driven approaches, possessing strong robustness and engineering applicability, and can effectively support equipment health status assessment and intelligent maintenance decisions.

[0022] Example 2: The technical solution of this embodiment of the invention differs from that of Example 1 in that: like Figure 1 and Figure 2 As shown, step four: construct a high-dimensional map based on condition monitoring data to characterize the nonlinear correlation between defect patterns, and optimize model parameters by combining evolutionary algorithms to achieve defect identification and evolution path analysis; Step four, which involves constructing a high-dimensional map based on condition monitoring data and characterizing the nonlinear correlations between defect modes, includes: Obtain the preprocessed state monitoring data and construct a time-series state vector from it: x(t k )=[x (1) (t k ),x (2) (t k ),…,x (n) (t k )] T ∈R n ; Where t k It is the kth sampling time, and n is the number of data collected. Each component represents the real-time measurement value of a type of physical quantity. For the total observation time T, construct the state data matrix: X=[x(t1),x(t2),…,x(t T )]∈R n×T; Using a fixed window length and sliding step size, the state information of consecutive time periods is extracted sequentially to form a series of trajectory segments. Each trajectory is composed of data spliced ​​from multiple time points. The total number of trajectories is calculated based on the set window length and sliding step size, and a trajectory set is generated. The diffusion mapping algorithm is used to reduce the dimensionality of high-dimensional trajectories. A Gaussian kernel function is constructed to measure the similarity between trajectories, and the transition probability matrix is ​​obtained by normalization. The first few non-trivial eigenvectors of this matrix and their corresponding eigenvalues ​​are extracted to construct the node representation in the low-dimensional embedding space. z i =[λ1φ1(i),λ2φ2(i),…,λ d φ d (i)] T ∈R d , where λ j is the j-th feature value, and d is the dimension of the embedding space; Ultimately, a set of low-dimensional nodes is obtained, which constitute the graph structure of the device status; For each pair of low-dimensional nodes, if one node is the nearest neighbor of the other, then an edge is created between the two nodes. The weight of the edge is determined by the distance between the nodes; the smaller the distance, the higher the weight. This ultimately constructs a weighted graph. , where Z is the set of nodes, each node corresponds to a low-dimensional embedded state, β is the set of edges, representing the connection relationship between states, and W is the weighted adjacency matrix; Step four, which describes the nonlinear correlation between defect modes, includes: The low-dimensional node set is classified using spectral clustering. Nodes with similar characteristics are grouped into the same category. Each category represents a defect pattern. The center point of each category is calculated as the representative state of the defect of that category. The squared Euclidean distance between the centers of various defects is calculated as a measure of their difference. A nonlinear function is introduced to convert the distance into the correlation strength. The smaller the distance, the stronger the correlation. Finally, a nonlinear correlation matrix reflecting the relationship between the defect patterns is obtained. A threshold is set as the top 20% quantile of the similarity between all defect pairs. If the squared Euclidean distance between two defect categories is less than this threshold, a directed edge is established between the corresponding nodes, thereby constructing a directed association graph between defects. Where C is the total number of defect categories, β c It is a set of directed edges, representing the possible transition directions of defect categories. Q is a nonlinear correlation strength matrix. This graph represents the possible evolution path of defects. Step four, which involves optimizing model parameters using evolutionary algorithms, includes: Design an input-output model with the device state vector as input and the defect category number as output. The model parameters are variables to be optimized. Define a comprehensive evaluation function, which includes the following three terms: Classification accuracy: The degree of consistency between the model's predictions and the true labels; Loss function value: reflects the magnitude of prediction error; Parameter norm: controls model complexity and prevents overfitting; Each item is combined according to a set weight to form an overall fitness index: ; Where η is the parameter to be optimized in the model, Acc(η) is the classification accuracy, and L(η) is the loss function value. All are weighting coefficients; The evolutionary search steps are as follows: T1: Initialize a batch of individuals with random parameters and calculate the fitness value of each individual; T2: Select the individuals with the highest fitness as parents, perform crossover to generate new individuals, and add a small mutation perturbation: T3: Repeat the above process until the preset number of iterations or the convergence condition is reached; Once the population converges or reaches the maximum number of iterations, the individual with the highest fitness is selected as the final model parameter set and saved. Step four, which involves defect identification and evolution path analysis, includes: Each state vector collected during equipment operation is input into the trained model, and the corresponding defect category label is output: ,in For time t k The predicted defect category label, f(·) is the defect identification model function, x(t) k (t) represents time t k The state vector, These are the optimal model parameters after training. Generate a complete defect label sequence and record the changes in defect types during equipment operation; The frequency of defect labels transitioning between adjacent time points is statistically analyzed, and the frequency ratio of each defect transitioning to another defect is calculated to form a defect transition probability matrix, which describes the possible transition trends between defects. Define a path sequence consisting of multiple defect categories, where the probability of a path is the product of the transition probabilities of all its adjacent defects: ; in From the i-thk Class defect transfer to the i-th k+1 The probability of a defect type, where K is the total number of defect types in the path; Among all possible paths, find the path with the highest probability: ,in The optimal evolution path; As the most likely defect evolution path; Here is a simulation example of high-dimensional graph modeling and defect evolution path prediction based on the content of step four: Consider an industrial manufacturing environment where a critical production piece of equipment undergoes continuous condition monitoring. This equipment may develop various types of defects due to prolonged operation, such as mechanical wear and fatigue cracks. To ensure the safe operation of the equipment and extend its service life, it is necessary to construct a high-dimensional map using condition monitoring data, analyze the nonlinear correlations between these defect patterns, and predict potential defect evolution paths. The equipment collects state monitoring data of various physical quantities (such as vibration, temperature, pressure, etc.) to form a time-series state vector; Based on the total observation time T, a state data matrix X is constructed, which contains data from all sampling times; Using a fixed window length and sliding step size, a series of trajectory segments are extracted from X. The diffusion mapping algorithm is then applied to reduce the dimensionality of these high-dimensional trajectories, resulting in a low-dimensional node representation z. i ; A weighted graph is built based on the distance between low-dimensional nodes to reflect the graph structure of the device status; The low-dimensional node set is classified using spectral clustering to determine each defect pattern. The squared Euclidean distance between each defect center is calculated and converted into nonlinear correlation strength. A directed correlation graph is constructed to show the transfer probability between different defect patterns. Design an input-output model and define a comprehensive evaluation function. Use an evolutionary algorithm to search for the optimal model parameters to maximize classification accuracy and minimize the loss function value. The real-time collected state vectors are input into the trained model to obtain the corresponding defect category labels. The transition frequency between defect labels at adjacent time points is analyzed to form a defect transition probability matrix. Based on the transition probability, the probability of all possible paths is calculated, and the most likely defect evolution path is found. After the above steps, the defect evolution path analysis results of the equipment over a period of time were obtained. Table 2 below is a simplified example of a defect transition probability matrix: Table 2 Defect Transfer Status Each cell represents the probability of transitioning from one defect type to another; By mapping multidimensional condition monitoring data into a low-dimensional spectral structure, the global evolution characteristics of equipment operating status are characterized. Typical defect patterns are extracted using spectral clustering methods, and the potential evolutionary relationships between defects are revealed through nonlinear correlation modeling. An evolutionary algorithm is introduced to optimize the parameters of the input-output model, which improves the classification accuracy and enhances the model's generalization ability. Finally, the most likely evolutionary path is mined based on the defect transition probability matrix, providing an interpretable quantitative analysis tool for equipment failure evolution trends.

[0023] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and its improved concept, should be covered within the scope of protection of the present invention.

Claims

1. A method for testing the airtightness of a diaphragm accumulator, characterized in that, Includes the following steps: Step 1: Collect status monitoring data during the operation of the diaphragm accumulator. The status monitoring data includes gas pressure data, temperature data, vibration data, acoustic emission signals, and strain data. Perform preprocessing operations on the collected status monitoring data. Step 2: Real-time monitoring and closed-loop control of gas flow rate, gas pressure, and gas path status during the inflation process, and automatic switching of gas path according to detection requirements; Step 3: By integrating gas pressure, temperature, vibration and acoustic emission signals, and using acoustic wave propagation inversion, dynamic crack identification and multi-scale analysis methods, the internal defects of the accumulator are preliminarily located and their evolution trend is analyzed. Step 4: Construct a high-dimensional map based on condition monitoring data to characterize the nonlinear correlation between defect patterns, and optimize model parameters by combining evolutionary algorithms to achieve defect identification and evolution path analysis.

2. The method for testing the airtightness of a diaphragm accumulator according to claim 1, characterized in that, The process of real-time monitoring and closed-loop control of gas flow rate, gas pressure, and gas path status during the inflation process in step two includes: Gas flow rate is collected using a thermal flow meter, and the collected gas flow rate data is transmitted to the control unit. A piezoelectric pressure sensor is used to collect gas pressure, and the collected data is transmitted to the control unit. The gas path conduction status is collected using a switch status sensor, the valve opening degree is collected using a valve position sensor, and the gas flow status is collected using a differential pressure flow detection module. The collected data is then connected to the control unit. The control system consists of sensors, a control unit, and an actuator. The control unit receives sensor signals from gas flow, gas pressure, and gas path status, compares the received actual measured values ​​with preset target values, calculates the deviation, and generates a control output signal based on the deviation to drive the actuator to adjust the system state. The control unit uses a proportional-integral-derivative control algorithm to perform closed-loop regulation of gas flow and gas pressure, and changes the gas flow rate, adjusts the gas pressure, or switches the gas path according to preset program instructions.

3. The method for testing the airtightness of a diaphragm accumulator according to claim 2, characterized in that, The process of automatically switching the gas path according to the detection requirements in step two includes: The control system determines whether to perform a gas path switch based on the task identifier and operating parameters. After identifying the switching conditions, the control system generates a switching command. The control system selects the target air path and issues a switching command, and the air path switching valve responds to the command and changes the airflow direction. The various components of the control system work together, and the air path switching mechanism guides the airflow into different paths according to the task identifier; The control system enables path switching and parameter adjustment of the gas path system.

4. The method for testing the airtightness of a diaphragm accumulator according to claim 1, characterized in that, The process of fusing gas pressure, temperature, vibration, and acoustic emission signals and performing acoustic wave propagation inversion analysis in step three includes: Preprocessed gas pressure data, temperature data, vibration data, and acoustic emission signals are acquired, and their features are extracted. The extracted features are then normalized and weighted and fused into a comprehensive feature vector F. The material's internal physical properties or structural information can be deduced from the acoustic wave signals measured at the boundary using the acoustic wave propagation inversion method. In an isotropic linear elastic medium, sound waves satisfy a three-dimensional second-order wave equation; When the location of the sound source and the parameters of the medium are known, the wave equation is solved using the finite difference method to obtain the theoretical wave field u(x,t); The goal of the inversion is to minimize the observed wavefield u gc (x i ,t) and the calculation of wave field u(x) i The difference between ,t;m) is used to optimize the medium parameter m(x); Introducing the adjoint field λ(x,t), the gradient of the objective function with respect to the medium parameters is calculated through the adjoint state equation; Using this gradient, the medium parameter m is iteratively updated using the L-BFGS algorithm until convergence.

5. The method for testing the airtightness of a diaphragm accumulator according to claim 4, characterized in that, The process of using dynamic crack identification and multi-scale analysis methods in step three includes: Dynamic crack propagation causes a decrease in material stiffness, which manifests as a decrease in structural modal frequency and an increase in damping ratio. The location of the crack can be identified by combining local energy density changes. The response signal is processed using the Hilbert-Huang transform to extract the nth instantaneous frequency f. n (t) and damping ratio ζ n (t); By introducing a local energy density function, the energy concentration increases during crack propagation, and the crack location is determined by combining the change in wave propagation path. Structural features at different scales are processed by multi-scale analysis methods, and cross-scale feature extraction is achieved by wavelet decomposition. Define micro variables ε is a small parameter that is much smaller than 1. The field quantity u has a multi-scale expansion form. Substitute the expansion into the control equation and separate the variables according to ε, and solve it step by step. Using wavelet basis functions ψ j,k (x) Performs multi-scale decomposition on the signal.

6. The method for testing the airtightness of a diaphragm accumulator according to claim 5, characterized in that, The process of preliminary location and evolution trend analysis of internal defects in the accumulator in step three includes: Defects in energy storage devices can be analyzed for their spatiotemporal evolution characteristics through multi-source data fusion. By inverting and reconstructing the internal reflection coefficient distribution r(x) of the accumulator, the distribution map of r(x) is obtained by inverse calculation of the measured echo s(t) to determine the location of the defect; The defect parameter D(t) is defined as the damage factor, and its evolution is defined as conforming to an exponential growth model. When the defect parameter D(t) reaches or exceeds the set threshold D sd When a defect is deemed to be in a dangerous state, repair or replacement measures must be taken.

7. The method for testing the airtightness of a diaphragm accumulator according to claim 1, characterized in that, Step four, which involves constructing a high-dimensional map based on condition monitoring data and characterizing the nonlinear correlations between defect modes, includes: Obtain the preprocessed state monitoring data, construct a time-series state vector from it, and build a state data matrix for the total observation time length T. Using a fixed window length and sliding step size, state information of consecutive time periods is extracted sequentially to form a series of trajectory segments. The total number of trajectories is calculated based on the set window length and sliding step size, and a trajectory set is generated. The high-dimensional trajectory is reduced using the diffusion mapping algorithm. The similarity between trajectories is measured by constructing a Gaussian kernel function and then normalized to obtain the transition probability matrix. The first few non-trivial eigenvectors of the matrix and their corresponding eigenvalues ​​are extracted to construct the node representation in the low-dimensional embedding space, and finally a set of low-dimensional nodes are obtained, which constitute the graph structure of the device state. For each pair of low-dimensional nodes, if one node is the nearest neighbor of the other node, then an edge is established between the two nodes, eventually constructing a weighted graph.

8. The method for testing the airtightness of a diaphragm accumulator according to claim 7, characterized in that, The process of characterizing the nonlinear correlation between defect patterns in step four includes: The low-dimensional node set is classified using spectral clustering. Nodes with similar characteristics are grouped into the same category. Each category represents a defect pattern. The center point of each category is calculated as the representative state of the defect of that category. The squared Euclidean distance between the centers of various defects is calculated as a measure of their differences, and finally a nonlinear correlation matrix reflecting the relationship between the various defect modes is obtained. A threshold is set as the top 20 percentile of the similarity between all defect pairs. If the squared Euclidean distance between two defect categories is less than this threshold, a directed edge is established between the corresponding nodes, thereby constructing a directed association graph between defects, which represents the possible evolution path of the defects.

9. The method for testing the airtightness of a diaphragm accumulator according to claim 8, characterized in that, The process of optimizing model parameters using evolutionary algorithms in step four includes: Design an input-output model with the device state vector as input and the defect category number as output. The model parameters are variables to be optimized. Define a comprehensive evaluation function, which includes the following three items: classification accuracy, loss function value and parameter norm. Each item is combined into an overall fitness index according to a set weight. The evolutionary search steps are as follows: T1: Initialize a batch of individuals with random parameters and calculate the fitness value of each individual; T2: Select the individuals with the highest fitness as parents, perform crossover to generate new individuals, and add a small mutation perturbation: T3: Repeat the above process until the preset number of iterations or the convergence condition is reached; The final output is the optimal model parameters.

10. The method for testing the airtightness of a diaphragm accumulator according to claim 9, characterized in that, The process of defect identification and evolution path analysis in step four includes: Each state vector collected during equipment operation is input into the trained model, and the corresponding defect category label is output to generate a complete defect label sequence. The frequency of defect transfer between adjacent defect labels is statistically analyzed, and the frequency ratio of each defect to another defect is calculated to form a defect transfer probability matrix. Define a path sequence consisting of multiple defect categories, where the probability of a path is the product of the transition probabilities of all its adjacent defects. Among all possible paths, identify the one with the highest probability as the most likely defect evolution path.

Images