Hydraulic system fault warning method and device based on digital twin

Abstract

The present invention discloses a hydraulic system fault warning method and device based on digital twins, and relates to the field of hydraulic warning technology. The method includes: constructing a digital twin model based on the physical structure and historical data of the hydraulic system; deploying sensors to collect monitoring data of the hydraulic system in real time and fitting it into the model; activating a fault warning module to determine the working cycle; integrating monitoring data to construct data of each cycle; configuring collaborative relationships according to workload and control strategy; fitting relationships to obtain monitoring fitting relationships; identifying anomalies based on fitting relationships and generating fault warning information. The method solves the technical problem that existing early warning methods often fail to accurately predict and promptly respond to potential faults within the hydraulic system, resulting in untimely fault handling and affecting the stability and operating efficiency of the hydraulic system. The method achieves the effect of fitting monitoring data based on collaborative relationships, ensuring the accuracy and timeliness of fault warnings, and improving the stability and operating efficiency of the hydraulic system.

Description

Technical Field

[0001] The present application relates to the technical field related to hydraulic early warning, and specifically to a hydraulic system fault early warning method and device based on digital twin. Background Art

[0002] With the continuous advancement of industrial automation, the stability and operating efficiency of hydraulic systems, as the core components of many key equipment and machinery, have an important impact on the overall production process. However, traditional hydraulic system fault warning methods often lack accuracy and timeliness when dealing with potential faults within the hydraulic system, resulting in untimely fault handling, which in turn affects the stability and operating efficiency of the hydraulic system. The complexity of the internal structure of the hydraulic system, including multiple components such as pumps, valves, actuators, and the interactions between them, makes it difficult to predict the occurrence of faults. At the same time, the working state of the hydraulic system is affected by a combination of factors, such as pressure, flow, temperature, vibration, etc., and even slight changes in these factors may indicate the occurrence of potential faults. Summary of the invention

[0003] The present application provides a hydraulic system fault warning method and device based on digital twins, which solves the technical problem that existing warning methods are often unable to accurately predict and promptly respond to potential faults within the hydraulic system, resulting in untimely fault handling and affecting the stability and operating efficiency of the hydraulic system. The application achieves the effect of fitting monitoring data based on collaborative relationships, ensuring the accuracy and timeliness of fault warnings, and improving the stability and operating efficiency of the hydraulic system.

[0004] The present application provides a hydraulic system fault warning method based on digital twins, the method comprising: constructing a digital twin model based on the physical structure and operation history data of the hydraulic system; deploying sensors to collect monitoring data of the hydraulic system in real time, including pressure, flow, temperature, vibration, and actuator position, and fitting the monitoring data into the digital twin model; activating a fault warning module in the digital twin model, comprising: determining each working cycle according to the actuator position; integrating pressure, flow, temperature, and vibration monitoring data based on each working cycle to construct monitoring data for each working cycle; configuring collaborative relationships for the monitoring data of each working cycle according to workload and control strategy; fitting the relationship of the monitoring data of each working cycle based on the collaborative relationship to obtain a monitoring fitting relationship; identifying anomalies according to the monitoring fitting relationship, and generating fault warning information based on the anomaly identification results.

[0005] The present application also provides a hydraulic system fault warning device based on digital twins, including: a model building unit, the model building unit is used to build a digital twin model based on the physical structure and operation history data of the hydraulic system; a data fitting unit, the data fitting unit is used to deploy sensors to collect real-time monitoring data of the hydraulic system, including pressure, flow, temperature, vibration, and actuator position, and fit the monitoring data into the digital twin model; a module activation unit, the module activation unit hydraulically activates the fault warning module in the digital twin model, including: determining each working cycle according to the actuator position; based on each working cycle, integrating the pressure, flow, temperature, and vibration monitoring data to construct each working cycle monitoring data; configuring the collaborative relationship of the monitoring data of each working cycle according to the workload and control strategy; based on the collaborative relationship, fitting the relationship of the monitoring data of each working cycle to obtain the monitoring fitting relationship; identifying anomalies according to the monitoring fitting relationship, and generating fault warning information based on the anomaly identification results.

[0006] The proposed hydraulic system fault warning method and device based on digital twins in this application is intended to construct a digital twin model based on the physical structure and operation history data of the hydraulic system; deploy sensors to collect monitoring data of the hydraulic system in real time, including pressure, flow, temperature, vibration, and actuator position, and fit the monitoring data to the digital twin model; activate the fault warning module in the digital twin model, including: determining each working cycle according to the actuator position; integrating the pressure, flow, temperature, and vibration monitoring data based on each working cycle to construct the monitoring data of each working cycle; configuring the collaborative relationship of the monitoring data of each working cycle according to the workload and control strategy; fitting the relationship of the monitoring data of each working cycle based on the collaborative relationship to obtain the monitoring fitting relationship; performing abnormal identification based on the monitoring fitting relationship, and generating fault warning information based on the abnormal identification result. The technical problem that the existing warning methods often fail to accurately predict and promptly respond to potential faults in the hydraulic system, resulting in untimely fault handling and affecting the stability and operation efficiency of the hydraulic system is solved, and the monitoring data is fitted based on the collaborative relationship to ensure the accuracy and timeliness of the fault warning, and improve the stability and operation efficiency of the hydraulic system. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] In order to more clearly illustrate the technical solution of the embodiment of the present disclosure, the accompanying drawings of the embodiment of the present disclosure will be briefly introduced below. A flow chart is used in the present application to illustrate the operations performed by the device according to the embodiment of the present application. It should be understood that the previous or following operations are not necessarily performed precisely in order. On the contrary, various steps can be processed in reverse order or simultaneously as needed. At the same time, other operations can also be added to these processes, or one or more operations can be removed from these processes.

[0008] Figure 1 A schematic diagram of a hydraulic system fault warning method based on digital twins provided in an embodiment of the present application;

[0009] Figure 2 A schematic diagram of the execution flow of a fault warning module of a hydraulic system fault warning method based on digital twins provided in an embodiment of the present application;

[0010] Figure 3 A schematic diagram of the structure of a hydraulic system fault warning device based on digital twin provided in an embodiment of the present application.

[0011] Explanation of the reference numerals: model building unit 11, data fitting unit 12, module activation unit 13. DETAILED DESCRIPTION

[0012] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented in accordance with the contents of the specification. In order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are listed below.

[0013] In order to make the purpose, technical solutions and advantages of the present application clearer, the present application will be further described in detail below in conjunction with the accompanying drawings. The described embodiments should not be regarded as limiting the present application. All other embodiments obtained by ordinary technicians in the field without making creative work are within the scope of protection of this application.

[0014] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments, but it is understood that "some embodiments" may be the same subset or different subsets of all possible embodiments, and may be combined with each other without conflict, and the terms "first\second" involved are merely to distinguish similar objects and do not represent a specific ordering of objects. The terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, device, product, or server that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or modules that are not clearly listed or inherent to these processes, methods, products, or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of this application. The terms used herein are for the purpose of describing the embodiments of the present application only.

[0015] The embodiment of the present application provides a hydraulic system fault warning method based on digital twin, such as Figure 1 As shown, the method includes:

[0016] A digital twin model is built based on the physical structure and historical operation data of the hydraulic system.

[0017] In the embodiment of the present application, the system terminal defines the various components of the hydraulic system, including power elements (hydraulic pumps), actuators (hydraulic cylinders and hydraulic motors), control elements (various hydraulic valves), auxiliary elements (oil tanks, oil filters, etc.) and working media (hydraulic oil). Each part has its specific functions and roles, and they work together to ensure the normal operation of the hydraulic system. Subsequently, the operation history data of the hydraulic system is collected. These data include parameters such as pressure, flow, temperature, speed, load, etc. of the hydraulic system, as well as information such as fault records and maintenance records of the hydraulic system. These data are the basis for building a digital twin model, and they provide important information about the actual operation status of the hydraulic system. After collecting enough physical structure information and operation history data, the system terminal begins to build a digital twin model. This model is a virtual hydraulic system that can simulate the operation process of an actual hydraulic system. In order to build this model, the system terminal analyzes the functions, structures and working principles of each component, and clarifies the connection relationship and interaction between them. Then collect the physical structure data of these components, such as the size and material of each component, and extract a part from the operation history data as a verification data set. After that, a 3D model of the hydraulic system is constructed based on the physical structure data using 3D modeling software such as SolidWorks and AutoCAD. The model can accurately reflect the physical structure and spatial layout of the hydraulic system. Based on the 3D model, combined with the operation history data, a simulation model of the hydraulic system is constructed using simulation software such as ANSYS, MATLAB / Simulink, etc. The model can simulate the operation process of the hydraulic system, including the working status of each component, the flow of the liquid, the pressure distribution, etc. Then, the simulation model is verified using the verification data set, and the difference between the simulation results and the actual operation data is compared to ensure the accuracy of the model. Then, the model is optimized based on the verification results, and the model parameters and simulation conditions are adjusted to improve the accuracy and reliability of the model. Finally, the verified simulation model is output to generate a digital twin model. Through this digital twin model, the system terminal can simulate the operation status of the hydraulic system in different scenarios, better understand the performance and behavior of the hydraulic system, and provide support for the early warning of the hydraulic system.

[0018] Deploy sensors to collect monitoring data of the hydraulic system in real time, including pressure, flow, temperature, vibration, and actuator position, and fit the monitoring data into the digital twin model.

[0019] In one embodiment, sensors are deployed in the hydraulic system to obtain key parameters of the hydraulic system operation in real time, which are crucial to ensure the normal operation of the hydraulic system and prevent potential failures. After the sensors in the hydraulic system are installed, the system terminal uses these sensors to monitor and record the pressure, flow, temperature, vibration and actuator position of the hydraulic system in real time. These real-time collected monitoring data are of great value to the digital twin model. By fitting the real-time monitoring data to the digital twin model, it can be ensured that the model can accurately reflect the actual state of the hydraulic system. Specifically, this process includes inputting the data collected by the sensor into the digital twin model, and the digital twin model updates its internal state based on this data and simulates the real-time operation of the hydraulic system. In summary, deploying sensors and collecting monitoring data of the hydraulic system in real time, and then fitting this data to the digital twin model, is a key step to ensure that the digital twin model can accurately reflect the actual state of the hydraulic system. This process not only helps to better understand the operation of the hydraulic system, but also provides strong support for the maintenance of the hydraulic system.

[0020] like Figure 2 As shown, the fault warning module in the digital twin model is activated.

[0021] In one embodiment, activating the fault warning module in the digital twin model is a key step in performing hydraulic system fault warning. After the digital twin model is constructed and verified, and the fitting of the monitoring data is completed. The system terminal activates the fault warning module in the digital twin model, and the fault warning module can monitor and evaluate the mode of the hydraulic system status in real time. This module will continuously analyze the monitoring data, perform collaborative relationship configuration, relationship fitting, anomaly identification and other steps, and determine whether the parameters of the current hydraulic system deviate from the normal range or show some abnormal pattern. If there is an abnormality, the fault warning module generates fault warning information and outputs it to the system terminal. Through the fault warning module, potential problems can be discovered in advance, the impact of hydraulic system failures on production can be reduced, and users can have more time to maintain equipment.

[0022] Each working cycle is determined according to the actuator position.

[0023] In one embodiment, in a hydraulic system, the position data of the actuator is the key information for determining each working cycle. The working cycle refers to the hydraulic cycle working process, which is the complete process of the hydraulic system starting from the initial state, undergoing a series of operations and changes, and finally returning to the initial state or entering the next working state. Actuators, such as hydraulic cylinders or motors, play the role of converting hydraulic energy into mechanical energy in the hydraulic system. Their position changes reflect the different stages of the hydraulic cycle. For example, when the hydraulic cylinder pushes forward, it indicates the beginning of a working cycle; when the hydraulic cylinder is fully extended or retracted to a predetermined position, it indicates the end of this working cycle and is ready to enter the next working cycle. Combined with the position data of the actuator, the system terminal can accurately determine the start and end time of each working cycle, as well as the various stages in the cycle, thereby deriving the working cycle of each actuator. This is very important for the monitoring, control and maintenance of the hydraulic system.

[0024] Based on each working cycle, the pressure, flow, temperature and vibration monitoring data are integrated to construct the monitoring data of each working cycle.

[0025] In one embodiment, in a hydraulic system, each working cycle is accompanied by a series of operations and changes, which are captured and recorded by monitoring different parameters. When the load and set parameters remain unchanged, ideally, the monitoring data within each cycle should be consistent. In order to construct the monitoring data of each working cycle, the system terminal performs real-time monitoring and data collection on key parameters such as pressure, flow, temperature and vibration in each cycle. These parameters are important bases for evaluating the performance of the hydraulic system, identifying potential problems and optimizing system operation. Among them, pressure monitoring data is one of the most critical parameters in the hydraulic system. In each working cycle, the change in pressure reflects the process of energy conversion and transmission inside the hydraulic system. By monitoring the pressure monitoring data in real time, the system terminal can understand the working status of the hydraulic system at different stages. Flow monitoring data is also important for evaluating the performance and efficiency of the hydraulic system. In each working cycle, the change in flow reflects the speed and efficiency of fluid transmission inside the hydraulic system. By monitoring the flow monitoring data, the system terminal can understand the fluid distribution of the hydraulic system at different stages. Temperature monitoring data is of great significance. Temperature changes in the hydraulic system may directly affect the performance and stability of the hydraulic system. Too high temperature may cause the hydraulic system to fail or be damaged, while too low temperature may affect the normal operation of the hydraulic system. Therefore, monitoring the temperature monitoring data in each working cycle helps the system terminal to detect temperature anomalies in time. Vibration monitoring data is important data for evaluating the stability and reliability of the hydraulic system. Vibration data can reflect the wear and looseness of the internal components of the hydraulic system. Monitoring vibration monitoring data in each working cycle helps the system terminal to detect potential problems in time and avoid failure or damage to the hydraulic system. Subsequently, the system terminal combines the monitoring data of the above parameters in each working cycle to construct a monitoring data set for each working cycle. These data sets provide the system terminal with rich information for evaluating the performance of the hydraulic system and identifying potential problems. When the load and setting parameters remain unchanged, if the monitoring data of each cycle is consistent, it reflects the good stability and predictability of the hydraulic system. However, if the data is abnormal or fluctuating, it may mean that there is a problem with the hydraulic system or it needs to be adjusted. Therefore, in-depth analysis and processing of these data are of great significance to ensure the long-term stable operation of the hydraulic system.

[0026] The monitoring data of each working cycle are configured with a collaborative relationship according to the workload and the control strategy.

[0027] In one embodiment, the monitoring data of each working cycle not only reflects the operating state of the system within the cycle, but is also closely related to the workload and control strategy. In order to ensure the efficient and stable operation of the hydraulic system, the system terminal configures the collaborative relationship of these monitoring data according to the workload and control strategy. Among them, the workload refers to the task or force that the hydraulic system needs to bear in different working stages. Different workloads will produce different pressure and flow requirements for the hydraulic system, which in turn affects the temperature, vibration and other parameters of the hydraulic system. Therefore, when configuring the collaborative relationship, it is necessary to consider the impact of the workload on the monitoring data to ensure that the hydraulic system can maintain stable performance under different loads. The control strategy is the key to the efficient and stable operation of the hydraulic system. It determines how the hydraulic system allocates resources and adjusts parameters to achieve optimal performance in different working stages. When configuring the collaborative relationship, the system terminal determines the scope of collaboration according to the workload and control strategy, and then clusters all the working cycles according to the determined collaborative scope, and classifies the working cycles under the same control strategy and load into one category. Subsequently, according to the clustering results, the data in the same cluster cluster is regarded as a collaborative relationship. By properly configuring the synergistic relationships, the system terminals can make multiple working cycles of the hydraulic system comparable, which helps to identify the similarities and differences between different working cycles under the same load and control strategy, and thus better understand the performance and behavior of the hydraulic system.

[0028] Furthermore, the present application provides the configuration of the collaborative relationship of the monitoring data of each working cycle according to the workload and the control strategy, including:

[0029] According to the workload and control strategy, a collaborative scope is determined, the collaborative scope includes a working time cycle, and the working time cycle is used to characterize the working time of the same workload and control strategy; the working time is used to cluster the working cycle to obtain multiple collaborative clustering sets; according to the multiple collaborative clustering sets, the monitoring data of each working cycle in a unified collaborative clustering set is configured as a collaborative relationship.

[0030] Preferably, in order to ensure the effective use and in-depth analysis of the monitoring data, the system terminal determines the scope of coordination according to the workload and control strategy. This coordination scope mainly refers to the working time period, which represents the duration of the working cycle with the same workload and control strategy. First, the system terminal identifies which working cycles have the same workload and control strategy, and determines the working time period of these cycles. These working time periods are the basis of clustering because they can characterize the running time under similar working conditions. Subsequently, the system terminal determines a suitable K value according to actual needs, where K represents the number of clusters. After that, the K-means algorithm is used to cluster the working cycles. The algorithm automatically divides the data into K collaborative cluster sets. Each collaborative cluster set contains working cycles running under the same workload and control strategy. Then, the system terminal creates a collaborative set for each cluster set. This set will contain the monitoring data of all working cycles belonging to the cluster, and mark the monitoring data in each collaborative set as having a collaborative relationship. This means that these data come from working cycles under the same workload and control strategy, so they can be directly compared and analyzed. Through such collaborative relationship configuration, the performance of the hydraulic system under different working cycles can be analyzed more accurately and potential problems can be discovered. At the same time, this also provides more reliable data support for fault diagnosis of the hydraulic system.

[0031] Based on the synergistic relationship, relationship fitting is performed on the monitoring data of each working cycle to obtain a monitoring fitting relationship.

[0032] In one embodiment, based on the determined collaborative relationship, the system terminal performs relationship fitting on the monitoring data of each working cycle to obtain a monitoring fitting relationship. First, the system terminal performs time-series alignment on the working cycle monitoring data, and performs similarity analysis on the aligned working cycle monitoring data. Subsequently, based on the similarity analysis results, the system terminal obtains a first monitoring fitting relationship. Afterwards, a certain time node is selected as a new initial value, the working cycle monitoring data is aligned again in time sequence, and the above-mentioned similarity analysis process is repeated to generate a second monitoring fitting relationship. Then, the system terminal repeats this process until the moving termination threshold is met, and multiple monitoring fitting relationships are generated. By performing relationship fitting on the working cycle monitoring data based on the collaborative relationship, the working mode of the hydraulic system can be better understood, the predictability and controllability of the hydraulic system can be improved, and strong support can be provided for the early warning of the hydraulic system.

[0033] Further, the present application provides the method of performing relationship fitting on monitoring data of each working cycle to obtain a monitoring fitting relationship, including:

[0034] Based on the timing relationship of each working cycle, the working cycle monitoring data are aligned, and similarity analysis is performed on the pressure, flow, temperature, and vibration monitoring data to obtain the similarity of each monitoring parameter; using the similarity of each monitoring parameter, a first monitoring fitting relationship is obtained.

[0035] Preferably, the system terminal uses the timing relationship of each working cycle to align the monitoring data in different cycles on the time axis. This is because different working cycles may have different start and end times. In order to compare and analyze the similarities between them, it is necessary to ensure that these data are synchronized in time. Subsequently, the system terminal performs similarity analysis on each parameter in the aligned monitoring data. These parameters include pressure, flow, temperature, vibration, etc., which are important indicators reflecting the operating status of the hydraulic system. The goal of similarity analysis is to find out whether the changing trends and patterns of these parameters in different working cycles are similar. When performing similarity analysis, the system terminal uses Euclidean distance to construct a distance matrix and searches for the minimum distance of this distance matrix to help the system terminal quantify the similarity between the monitoring parameters and obtain a similarity value. Once the similarity values ​​of each monitoring parameter are obtained, the system terminal assigns weights to each monitoring parameter based on historical experience and professional knowledge, and performs weighted average on the calculated similarity values ​​of each monitoring parameter, and integrates the similarities of all monitoring parameters to comprehensively consider the similarity and correlation between all parameters. The integrated similarity not only provides quantitative information about the similarity between different working cycles, but also reflects the overall change trend and pattern of the monitoring parameters. The system terminal will select a suitable data model based on this information and the characteristics of the data to construct the first monitoring fitting relationship. This first monitoring fitting relationship is used to describe the overall similarity and change trend of the monitoring parameters of different working cycles. The choice of the model to construct the first monitoring fitting relationship depends on the characteristics of the data and the needs of the problem, and can be a linear model, a nonlinear model, a time series model, etc. For example, when the integrated similarity shows that the monitoring parameters present a stable linear relationship over time, a linear model is selected. In general, based on the temporal relationship of each working cycle, the monitoring data is aligned and similarity analyzed to obtain the similarity of each monitoring parameter, and the first monitoring fitting relationship is constructed using these similarities. This process helps to deeply understand the operating laws of the hydraulic system and provide an important basis for fault warning of the hydraulic system.

[0036] A moving step is set, and the timing relationship is adjusted based on the moving step to establish an adjustment alignment relationship, wherein the moving step includes one or more moving steps, and if there are multiple moving steps, they correspond to the monitoring parameters; based on the adjustment alignment relationship, the similarity of each monitoring parameter is calculated to obtain a second monitoring fitting relationship.

[0037] Preferably, due to the influence of equipment aging, fatigue state or other factors, the monitoring parameters may show a different change trend in the working cycle than in normal operation. These changes are not only related to the properties of the parameters themselves, but also closely related to the time of equipment operation and the timing relationship within the cycle. In order to analyze these changes more accurately, the system terminal introduces a moving step length to adjust the timing relationship and establish an adjustment alignment relationship. The moving step length is set according to the actual changes of each monitoring parameter in the hydraulic system. After obtaining the first monitoring fitting relationship, the system terminal aligns these parameters at a new time point through a set moving step length, so as to perform similarity calculation and establish a fitting relationship. The moving step length can be one or more, and the specific number depends on the number and complexity of the monitoring parameters in the hydraulic system. Each moving step length can correspond to one or more monitoring parameters to ensure that each parameter is properly adjusted in timing. Subsequently, based on the set moving step length, the system terminal adjusts the original timing relationship to establish an adjustment alignment relationship. This process aims to align each monitoring parameter at a new time point so that the similarity and change trend between the parameters can be more accurately reflected in subsequent analysis. After the adjustment alignment relationship is established, the system terminal repeats the aforementioned steps of obtaining the first monitoring fitting relationship, performs similarity calculation on the adjustment alignment relationship, and obtains the second monitoring fitting relationship. By setting the moving step size and adjusting the timing relationship, the system terminal can more accurately reflect the changes in each monitoring parameter in the hydraulic system. The similarity calculation and fitting relationship establishment based on the adjustment alignment relationship help the system terminal to have a deeper understanding of the similarities and correlations between the parameters, thereby providing support for more accurate fault warnings.

[0038] The fitting is completed according to the moving step until the moving termination threshold is reached, and the Nth monitoring fitting relationship is obtained, wherein the moving termination threshold corresponds to the time length of the working cycle; the monitoring fitting relationship is obtained according to the first monitoring fitting relationship, the second monitoring fitting relationship, and until the Nth monitoring fitting relationship.

[0039] Preferably, in order to more accurately reflect the changing trend and similarity of parameters in different working cycles, the system terminal uses the moving step to continuously adjust the timing relationship to establish multiple monitoring fitting relationships. In this process, the system terminal sets a moving termination threshold, which is set to be less than half of the working cycle time length. This is because if the moving step is too short, the alignment relationship of the fit will be less, resulting in the lack of sufficient representativeness of the fitting result, and the changing trend and similarity between the parameters cannot be accurately reflected. The system terminal adjusts the timing relationship multiple times based on the set moving step, and uses the same method as above to establish N monitoring fitting relationships respectively. Through the set moving termination threshold, the system terminal can ensure that when performing multiple fittings, the fitting results will not be inaccurate due to the step length being too short, nor will the important changing trend be missed due to the step length being too long. After obtaining multiple monitoring fitting relationships, the system terminal adds these monitoring fitting relationships to the monitoring fitting relationship to form a final monitoring fitting relationship. The monitoring fitting relationship finally obtained is a mathematical model that integrates multiple fitting results, which can more accurately describe the overall similarity and changing trend of the parameters in different working cycles. In summary, by setting an appropriate moving step size and moving termination threshold, and integrating multiple fitting results, a more accurate and reliable monitoring fitting relationship can be obtained, providing strong support for fault warning of hydraulic systems.

[0040] The present application provides the similarity analysis of pressure, flow, temperature and vibration monitoring data to obtain the similarity of each monitoring parameter, including:

[0041] Based on the aligned monitoring data pairs, the Euclidean distance of the data is calculated; and according to the calculated Euclidean distance of the data, a distance matrix is ​​constructed.

[0042] Optionally, after the system terminal obtains the aligned monitoring data pairs, the system terminal uses these pairs to calculate the Euclidean distance between them. Euclidean distance is a distance measurement method that takes into account the differences between data points on each monitoring parameter and quantifies the distance between data points by calculating the square root of the sum of the squares of these differences. First, the system terminal extracts each pair of aligned monitoring data, which contains the values ​​of the same monitoring parameter under different working cycles. Then, for each pair of data pairs, the difference between them on each monitoring parameter is calculated and the square of these differences is squared. Then the square differences of all parameters are added together to obtain the square of the Euclidean distance between the pair of data. Then, the square root of this square value is taken to obtain the final Euclidean distance. After completing the Euclidean distance calculation for all data pairs, the system terminal organizes these distance values ​​into a distance matrix. Each row and column of this matrix corresponds to a monitoring data pair, and each element in the matrix is ​​the Euclidean distance between the two data pairs corresponding to the corresponding row and column. This distance matrix provides an intuitive and quantitative representation of the degree of difference between all monitoring data pairs.

[0043] The accumulation path and the accumulation direction are set according to the time sequence relationship, the distance is accumulated according to the distance matrix, and the minimum distance is searched; the similarity of the monitoring parameters is configured according to the minimum distance to obtain the similarity of the monitoring parameters.

[0044] Optionally, after obtaining the distance matrix, the system terminal performs distance accumulation to search for the minimum distance, and determines the similarity of the monitoring parameters based on the minimum distance. In this process, the accumulation path and the accumulation direction are the constraints of distance accumulation. The accumulation path and the accumulation direction define the way to perform path search in the distance matrix. The accumulation path refers to a set of a series of adjacent cells passed from the starting position to the ending position of the matrix, for example, from the upper left corner to the lower right corner. The accumulation direction specifies how to move from one element to the next element, for example, from left to right. When performing path search, the system terminal follows the constraints of the accumulation path and the accumulation direction from the starting position of the distance matrix, that is, the first data pair, and accumulates in the accumulation direction. Each accumulation adds the value of the current cell to the previously accumulated value, and this process continues until the ending position. During the accumulation process, the system terminal records the total distance of each path, that is, the accumulation result. Since there are multiple different accumulation paths, the system terminal sorts all possible paths searched in ascending order and finds the minimum distance. Once the minimum distance is found, the system terminal performs similarity configuration of the monitoring parameters based on this minimum distance. Similarity is a quantitative indicator used to describe the degree of similarity between different monitoring parameters. The system terminal uses the minimum distance obtained as a measure of dissimilarity between two monitoring parameter sequences. Subsequently, the system terminal uses a certain inverse proportional function to convert the minimum distance into a similarity score, so that the smaller the distance, the higher the similarity score. In summary, according to the accumulation path and accumulation direction set by the temporal relationship, accumulating distances in the distance matrix to search for the minimum distance, and configuring the similarity of monitoring parameters based on the minimum distance are important steps in monitoring data analysis. This process not only takes into account the temporal relationship between monitoring data, but also uses the information provided by the distance matrix to quantify the similarity between different monitoring parameters, providing strong support for subsequent data analysis.

[0045] Anomalies are identified according to the monitoring fitting relationship, and fault warning information is generated based on the anomaly identification result.

[0046] In one embodiment, the monitoring fitting relationship analyzes the monitoring data of each parameter in the real-time monitoring data, and combines the abnormal threshold of the corresponding parameter to determine whether there is a peak value of sudden change. If there is a peak value of sudden change, the system terminal generates fault warning information of the parameter based on the peak value of these changes. These warning information include detailed information such as the sudden change parameter, the type of fault, and the time of occurrence, so that the operator can respond quickly and take measures.

[0047] The present application provides an abnormality identification according to the monitoring fitting relationship, and generates fault warning information based on the abnormality identification result, including:

[0048] According to the operation history data, a fault sample set is obtained; according to the fault sample set, an abnormal relationship threshold analysis of the pressure, flow, temperature, and vibration monitoring data is performed to determine the abnormal threshold of each parameter; the abnormal threshold of each parameter is used to traverse the fitting relationship in the first monitoring fitting relationship, and when the traversal result exists, the fault warning information of the parameter is generated.

[0049] Optionally, the system terminal collects the operation history data of the hydraulic system and selects samples containing fault states from them, which constitute a fault sample set. The fault sample set contains monitoring data of key parameters such as pressure, flow, temperature, vibration, etc. of the hydraulic system under different fault conditions. Subsequently, the system terminal uses these fault sample sets to analyze the abnormal relationship thresholds of each monitoring parameter. This involves statistically analyzing the parameter data under the fault state to determine which parameter values ​​or parameter combinations are abnormal, and setting corresponding thresholds in combination with professional knowledge. These thresholds will serve as the benchmark for subsequent fault warnings. After obtaining these abnormal thresholds, the system terminal uses the abnormal thresholds of each parameter to traverse the generated first monitoring fitting relationship. During the traversal process, it is checked whether the monitoring data of each parameter exceeds the abnormal threshold of the corresponding parameter, and the exceeded detection parameters are output. If any parameter in the traversal result exceeds its abnormal threshold, this means that the hydraulic system may have a fault or abnormal situation. At this time, the system terminal will immediately generate corresponding fault warning information and transmit this information to the operator so that measures can be taken quickly to solve the problem.

[0050] The present application provides an abnormality identification according to the monitoring fitting relationship, and generates fault warning information based on the abnormality identification result, including:

[0051] Perform a timing anomaly feature analysis based on the fault sample set to construct a timing anomaly feature, where the timing anomaly feature is used to characterize the abnormal relationship of the timing changes of the monitoring data; determine the timing fitting feature based on the second monitoring fitting relationship to the Nth monitoring fitting relationship; use the timing anomaly feature to identify anomalies of the timing fitting feature, and generate fault warning information of the timing trend when the timing anomaly feature is met.

[0052] Optionally, the accurate identification of the abnormal relationship of timing changes by the system terminal is a key step in predicting potential faults. First, based on the obtained fault sample set, the system terminal performs a timing abnormality feature analysis to verify whether the timestamp of each data in the fault sample set is accurate. If the timestamp is wrong or missing, it needs to be corrected or supplemented according to the actual situation. For example, when there is a missing timestamp, the system terminal determines the approximate time range of the missing timestamp based on the timestamps of adjacent data records. If the timestamps of adjacent data are very close, the missing timestamp is estimated using interpolation. If the time interval is long or there is not enough information for interpolation, a reasonable value is estimated based on historical experience and the normal mode of equipment operation. Subsequently, according to the timestamp field, the data in the fault sample set are arranged in chronological order to generate a timing abnormality feature. The sorted data will present a continuous timeline, which is convenient for subsequent timing abnormality feature analysis. Afterwards, the system terminal determines the timing fitting feature based on the obtained second monitoring fitting relationship to the Nth monitoring fitting relationship. These fitting features describe the timing dependency between different parameters. Once the timing fitting feature is obtained, the system terminal uses the determined timing abnormality feature for abnormality identification. Abnormal identification compares the timing fitting features with the timing abnormal features to determine whether the current hydraulic system state is abnormal. When the conditions of the timing abnormal features are met, the system terminal will generate fault warning information of the timing trend. These warning information not only includes the type and location of the fault, but also points out the possible time and trend of the fault, providing operators with more comprehensive and timely fault information. In this way, operators can find problems earlier and take measures to avoid greater damage to the hydraulic system caused by the fault.

[0053] This application provides for adjusting the timing anomaly features, including:

[0054] Obtain fitting parameters of the abnormal relationship of time series changes; and calculate t statistics of the fitting parameters.

[0055] Optionally, the system terminal extracts fitting parameters from the abnormal relationship of the time series change, that is, the parameters involved in the fitting process of the time series feature fitting based on the monitoring fitting relationship. Subsequently, the system terminal calculates the t statistic for each fitting parameter. The t statistic is calculated based on the fitting value of the parameter, the standard error of the fitting value, and the parameter value under the null hypothesis, and is used to test whether the parameter is stable to ensure the reliability of the fault warning.

[0056] Obtain degrees of freedom and a preset confidence level, and search for a t-distribution critical value corresponding to the degrees of freedom according to the preset confidence level; use the t-distribution critical value to perform deviation calculation on the t-statistic to determine a stable evaluation value; generate a correction compensation coefficient according to the stable evaluation value, and use the correction compensation coefficient to adjust the abnormal time series feature.

[0057] Optionally, when performing anomaly detection and analysis of time series data, the system terminal calculates the difference between the number of parameters in the time series anomaly feature and the number of parameters involved in the fitting to obtain the degree of freedom. The number of degrees of freedom directly affects the accuracy of statistical inference. Subsequently, in order to evaluate the reliability and significance of the fitting parameters, the system terminal sets a preset confidence level. The setting range of the preset confidence level is 90% to 95%, depending on the specific requirements for accuracy. A higher preset confidence level represents a higher requirement for the accuracy of statistical inference. Once the degrees of freedom and the preset confidence level are determined, the system terminal searches for the t-distribution critical value corresponding to these values. The t-distribution critical value is also called the t-value, which is the t-value of the point on the t-distribution corresponding to the confidence level. The search for the t-distribution critical value is achieved by consulting the t-distribution table in the online resource. In the t-distribution table, the corresponding t-value can be found according to the degrees of freedom and confidence level. Afterwards, the system terminal uses the found t-distribution critical value to perform deviation calculation on the t-statistic of the fitting parameter. By calculating the difference between the t statistic and the critical value of the t distribution, a stability evaluation value can be obtained. The higher the stability evaluation value, the more stable the time series data is; otherwise, there is an abnormal situation. After obtaining the stability evaluation value, the system terminal sets a set of rules based on the stability evaluation value to generate a correction compensation coefficient. This rule is formulated according to the actual application scenario and the experience of the data analyst. For example, a threshold is set. When the stability evaluation value is lower than this threshold, a non-zero correction compensation coefficient is generated based on the set linear relationship, that is, the stability evaluation value is multiplied by a constant. This constant is set according to the specific situation of the data and the experience of the analyst, and determines the size and sensitivity of the correction compensation coefficient. When the stability evaluation value is higher than the threshold, a correction compensation coefficient close to zero or zero is generated, indicating that no major adjustment is required. Then, according to the set rules, the correction compensation coefficient is calculated using the stability evaluation value. Once the correction compensation coefficient is calculated, the system terminal applies it to the abnormal characteristics of the time series to adjust the time series data. The specific adjustment method depends on the type of time series data and the manifestation of the abnormal characteristics. For example, if the pressure value is abnormal, the system terminal will add the correction compensation coefficient to the abnormal pressure value to obtain a pressure value closer to the actual value. Through this series of steps, the abnormal features in the time series data can be effectively identified and adjusted through the correction compensation coefficient, thereby improving the accuracy and reliability of the time series data analysis.

[0058] In the above, refer to Figure 1 and Figure 2 The hydraulic system fault warning method based on digital twin according to an embodiment of the present invention is described in detail. Figure 3 A hydraulic system fault warning device based on digital twin according to an embodiment of the present invention is described.

[0059] The hydraulic system fault warning device based on digital twin according to the embodiment of the present invention is used to solve the technical problem that the existing warning methods are often unable to accurately predict and timely respond to potential faults inside the hydraulic system, resulting in untimely fault handling and affecting the stability and operating efficiency of the hydraulic system. It achieves the effect of fitting monitoring data based on collaborative relationships, ensuring the accuracy and timeliness of fault warnings, and improving the stability and operating efficiency of the hydraulic system. The hydraulic system fault warning device based on digital twin includes: a model building unit 11, a data fitting unit 12, and a module activation unit 13.

[0060] Model building unit 11: The model building unit 11 is used to build a digital twin model based on the physical structure and operation history data of the hydraulic system;

[0061] Data fitting unit 12: The data fitting unit 12 is used to deploy sensors to collect monitoring data of the hydraulic system in real time, including pressure, flow, temperature, vibration, and actuator position, and fit the monitoring data into the digital twin model;

[0062] Module activation unit 13: The module activation unit 13 is used to activate the fault warning module in the digital twin model, including:

[0063] Determining each working cycle according to the position of the actuator;

[0064] Based on each working cycle, integrate the pressure, flow, temperature and vibration monitoring data to build the monitoring data of each working cycle;

[0065] According to the workload and control strategy, the monitoring data of each work cycle are configured in a coordinated manner;

[0066] Based on the collaborative relationship, performing relationship fitting on the monitoring data of each working cycle to obtain a monitoring fitting relationship;

[0067] Anomalies are identified according to the monitoring fitting relationship, and fault warning information is generated based on the anomaly identification result.

[0068] Furthermore, the module activation unit 13 further includes:

[0069] Determine a coordination scope according to the workload and the control strategy, wherein the coordination scope includes a working time period, and the working time period is used to characterize the working time of the same workload and the control strategy;

[0070] Clustering the working cycles using the working time to obtain a plurality of collaborative clustering sets;

[0071] According to the multiple collaborative cluster sets, each work cycle monitoring data in a unified collaborative cluster set is configured to be in a collaborative relationship.

[0072] Furthermore, the module activation unit 13 further includes:

[0073] Based on the time sequence relationship of each working cycle, the working cycle monitoring data are aligned, and similarity analysis is performed on the pressure, flow, temperature, and vibration monitoring data to obtain the similarity of each monitoring parameter;

[0074] Using the similarities of the monitoring parameters, a first monitoring fitting relationship is obtained;

[0075] Setting a moving step length, adjusting the timing relationship based on the moving step length, and establishing an adjustment alignment relationship, wherein the moving step length includes one or more moving steps, and if there are multiple moving step lengths, they correspond to monitoring parameters;

[0076] Calculating the similarity of each monitoring parameter based on the adjusted alignment relationship to obtain a second monitoring fitting relationship;

[0077] The fitting is completed according to the moving step length until the moving termination threshold is reached, thereby obtaining an Nth monitoring fitting relationship, wherein the moving termination threshold corresponds to the time length of the working cycle;

[0078] The monitoring fitting relationship is obtained according to the first monitoring fitting relationship, the second monitoring fitting relationship, and up to the Nth monitoring fitting relationship.

[0079] Furthermore, the module activation unit 13 further includes:

[0080] Based on the aligned monitoring data pairs, the Euclidean distance of the data is calculated;

[0081] Constructing a distance matrix according to the calculated Euclidean distance of the data;

[0082] Setting the accumulation path and accumulation direction according to the time sequence relationship, accumulating the distance according to the distance matrix, and searching for the minimum distance;

[0083] The similarity of the monitoring parameters is configured according to the minimum distance to obtain the similarity of the monitoring parameters.

[0084] Furthermore, the module activation unit 13 further includes:

[0085] Obtain fault sample sets based on historical operation data;

[0086] Performing abnormal relationship threshold analysis of pressure, flow, temperature, and vibration monitoring data based on the fault sample set to determine abnormal thresholds of each parameter;

[0087] The fitting relationship in the first monitoring fitting relationship is traversed using the abnormal threshold of each parameter, and when the traversal result exists, fault warning information of the parameter is generated.

[0088] Furthermore, the module activation unit 13 further includes:

[0089] Performing time series anomaly feature analysis according to the fault sample set to construct a time series anomaly feature, wherein the time series anomaly feature is used to characterize an abnormal relationship of time series changes in monitoring data;

[0090] Determine a timing fitting feature based on the second monitoring fitting relationship to the Nth monitoring fitting relationship;

[0091] The time series abnormality feature is used to identify the abnormality of the time series fitting feature, and when the time series abnormality feature is met, fault warning information of the time series trend is generated.

[0092] Furthermore, the module activation unit 13 further includes:

[0093] Obtain fitting parameters of abnormal relationships of time series changes;

[0094] Calculating the t-statistic of the fitting parameters;

[0095] Obtaining degrees of freedom and a preset confidence level, and searching for a t distribution critical value corresponding to the degrees of freedom according to the preset confidence level;

[0096] Calculating the deviation of the t statistic using the t distribution critical value to determine a stable evaluation value;

[0097] A correction compensation coefficient is generated according to the stability evaluation value, and the timing abnormality feature is adjusted using the correction compensation coefficient.

[0098] The hydraulic system fault warning device based on digital twin provided in an embodiment of the present invention can execute the hydraulic system fault warning method based on digital twin provided in any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0099] Although the present application makes various references to certain modules in the device according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of the present invention.

[0100] The above specific implementations do not constitute a limitation on the protection scope of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions and improvements made within the spirit and principles of this application should be included in the protection scope of this application.

Claims

1. Hydraulic system fault warning method based on digital twin, It is characterized in that The method comprises: Build a digital twin model based on the physical structure and historical operation data of the hydraulic system; Deploy sensors to collect monitoring data of the hydraulic system in real time, including pressure, flow, temperature, vibration, and actuator position, and fit the monitoring data into the digital twin model; Activating the fault warning module in the digital twin model includes: Determining each working cycle according to the position of the actuator; Based on each working cycle, integrate the pressure, flow, temperature and vibration monitoring data to build the monitoring data of each working cycle; According to the workload and control strategy, the monitoring data of each work cycle are configured in a coordinated manner; Based on the collaborative relationship, performing relationship fitting on the monitoring data of each working cycle to obtain a monitoring fitting relationship; Performing abnormality identification according to the monitoring fitting relationship, and generating fault warning information based on the abnormality identification result; The performing relationship fitting on the monitoring data of each working cycle to obtain the monitoring fitting relationship includes: Based on the time sequence relationship of each working cycle, the working cycle monitoring data are aligned, and similarity analysis is performed on the pressure, flow, temperature, and vibration monitoring data to obtain the similarity of each monitoring parameter; Using the similarities of the monitoring parameters, a first monitoring fitting relationship is obtained; Setting a moving step length, adjusting the timing relationship based on the moving step length, and establishing an adjustment alignment relationship, wherein the moving step length includes one or more moving steps, and if there are multiple moving step lengths, they correspond to monitoring parameters; Calculating the similarity of each monitoring parameter based on the adjusted alignment relationship to obtain a second monitoring fitting relationship; The fitting is completed according to the moving step length until the moving termination threshold is reached, thereby obtaining an Nth monitoring fitting relationship, wherein the moving termination threshold corresponds to the time length of the working cycle; Obtaining the monitoring fitting relationship according to the first monitoring fitting relationship, the second monitoring fitting relationship, and up to the Nth monitoring fitting relationship; The step of configuring the coordinated relationship of the monitoring data of each working cycle according to the workload and the control strategy includes: Determine a coordination scope according to the workload and the control strategy, wherein the coordination scope includes a working time period, and the working time period is used to characterize the working time of the same workload and the control strategy; Clustering the working cycles using the working time to obtain a plurality of collaborative clustering sets; According to the plurality of collaborative cluster sets, configuring each work cycle monitoring data in a unified collaborative cluster set to be in a collaborative relationship; Performing abnormality identification according to the monitoring fitting relationship and generating fault warning information based on the abnormality identification result includes: Obtain fault sample sets based on historical operation data; Performing abnormal relationship threshold analysis of pressure, flow, temperature, and vibration monitoring data based on the fault sample set to determine abnormal thresholds of each parameter; Using the abnormal threshold of each parameter to traverse the fitting relationship in the first monitoring fitting relationship, when the traversal result exists, generating fault warning information of the parameter; Performing abnormality identification according to the monitoring fitting relationship and generating fault warning information based on the abnormality identification result also includes: Performing time series anomaly feature analysis according to the fault sample set to construct a time series anomaly feature, wherein the time series anomaly feature is used to characterize an abnormal relationship of time series changes in monitoring data; Determine a timing fitting feature based on the second monitoring fitting relationship to the Nth monitoring fitting relationship; The time series abnormality feature is used to perform abnormality identification on the time series fitting feature, and when the time series abnormality feature is met, fault warning information of the time series trend is generated.

2. The method according to claim 1, It is characterized in that The similarity analysis is performed on the pressure, flow, temperature and vibration monitoring data respectively to obtain the similarity of each monitoring parameter, including: Based on the aligned monitoring data pairs, the Euclidean distance of the data is calculated; Constructing a distance matrix according to the calculated Euclidean distance of the data; Setting the accumulation path and accumulation direction according to the time sequence relationship, accumulating the distance according to the distance matrix, and searching for the minimum distance; The similarity of the monitoring parameters is configured according to the minimum distance to obtain the similarity of the monitoring parameters.

3. The method according to claim 1, It is characterized in that The method further comprises: Obtain fitting parameters of abnormal relationships of time series changes; Calculating the t-statistic of the fitting parameters; Obtaining degrees of freedom and a preset confidence level, and searching for a t distribution critical value corresponding to the degrees of freedom according to the preset confidence level; Calculating the deviation of the t statistic using the t distribution critical value to determine a stable evaluation value; A correction compensation coefficient is generated according to the stability evaluation value, and the timing abnormality feature is adjusted using the correction compensation coefficient.

4. Hydraulic system fault warning device based on digital twin, It is characterized in that The device is used to implement the hydraulic system fault warning method based on digital twin according to any one of claims 1 to 3, and the method includes: Model building unit: builds a digital twin model based on the physical structure and operation history data of the hydraulic system; Data fitting unit: deploy sensors to collect monitoring data of the hydraulic system in real time, including pressure, flow, temperature, vibration, and actuator position, and fit the monitoring data into the digital twin model; Module activation unit: activating the fault warning module in the digital twin model, including: Determining each working cycle according to the position of the actuator; Based on each working cycle, integrate the pressure, flow, temperature and vibration monitoring data to build the monitoring data of each working cycle; According to the workload and control strategy, the monitoring data of each work cycle are configured in a coordinated manner; Based on the collaborative relationship, performing relationship fitting on the monitoring data of each working cycle to obtain a monitoring fitting relationship; Anomalies are identified according to the monitoring fitting relationship, and fault warning information is generated based on the anomaly identification result.

Images