Aircraft hydraulic pipeline system fault prediction method and system based on digital twinning

Abstract

The invention relates to the technical field of data processing, and discloses an aircraft hydraulic pipeline system fault prediction method and system based on digital twinning. The method comprises the steps that flight parameters are acquired through an aircraft data bus for stage identification, and four flight stage identifiers are obtained; carrying out mapping processing on the load spectrum according to the identifier, and constructing a staged digital twin pipeline model; performing weight distribution on the multi-sensor data according to identifiers, and extracting fault feature vectors according to a twin model; and performing prediction processing through the BP neural network in stages to obtain a fault prediction result. The method solves the problem that the existing digital twinborn model cannot adapt to the full flight envelope load difference by adopting fixed parameters, so that the sensor data weight distribution lacks flight stage adaptability, and the fault prediction algorithm cannot accurately identify the differentiated fault features of different flight stages.

Description

Technical Field

[0001] The present application relates to the technical field of data processing, and in particular, to a method and system for fault prediction of an aircraft hydraulic pipeline system based on digital twin. Background Art

[0002] Existing fault prediction methods for aircraft hydraulic pipeline systems mainly rely on traditional prognostics and health management (PHM) technologies. Data is collected by installing pressure, temperature, and flow sensors at key nodes of the hydraulic pipeline, and a single machine learning model is used to diagnose and predict hydraulic pipeline faults. As an emerging intelligent maintenance method, digital twin technology has been widely applied in the industrial field. By constructing a virtual model that maps to the physical entity, it realizes real-time data interaction between the physical space and the virtual space, providing technical support for on-line condition monitoring, fault diagnosis, and predictive maintenance of equipment. Traditional digital twin hydraulic pipeline fault prediction methods usually establish a unified digital twin model based on ground test data or single working condition conditions, and use a virtual model with fixed parameters for fault analysis.

[0003] However, during different flight phases (takeoff, cruise, maneuvering flight, landing) of an aircraft, there are significant differences in the load spectra and working conditions that the hydraulic pipeline system bears. During the takeoff phase, it needs to withstand the transient high-pressure impact caused by the retraction and extension of the landing gear. During the cruise phase, it faces the change in the viscosity of hydraulic oil in the high-altitude low-temperature environment. During the maneuvering flight phase, it bears the stress concentration under large overload conditions. During the landing phase, it experiences frequent hydraulic system operations. The existing single digital twin model cannot adapt to these differentiated load characteristics, and it is difficult for the virtual model with fixed physical parameters to accurately reflect the true working state of the hydraulic pipeline under different flight phases, resulting in insufficient fault prediction accuracy. At the same time, the existing methods lack specialized fault feature extraction and prediction algorithms for different flight phases and cannot effectively identify the unique fault modes in each phase.

[0004] Further analysis reveals that there are deeper technical problems in the existing technology: the weight allocation strategy of sensor data lacks adaptability to flight phases and cannot be dynamically adjusted according to the monitoring key points in different phases, resulting in the masking of key fault features; the physical parameter configuration of the digital twin model lacks an adaptive switching mechanism and cannot update the model parameters in real time according to the change of flight state; the fault prediction algorithm lacks a phased dedicated design, and a single BP neural network model is difficult to accurately learn and identify the differentiated fault features in different flight phases. Especially when dealing with faults in a multi-field multi-scale and strongly nonlinear pipeline system, the prediction accuracy and real-time performance cannot meet the fault prediction requirements of the full flight envelope. Summary of the Invention

[0005] The present application provides a method and system for fault prediction of an aircraft hydraulic pipeline system based on digital twin, which is used to solve the problem that the existing digital twin model adopts fixed parameters and cannot adapt to the load differences in the full flight envelope, resulting in the lack of flight phase adaptability in the weight allocation of sensor data, and further making the fault prediction algorithm unable to accurately identify the differential fault characteristics in different flight phases.

[0006] In the first aspect, the present application provides a method for fault prediction of an aircraft hydraulic pipeline system based on digital twin. The method for fault prediction of an aircraft hydraulic pipeline system based on digital twin includes: obtaining flight parameters through the aircraft data bus, performing stage identification processing on the flight envelope according to the flight parameters to obtain four flight stage identifiers of takeoff, cruise, maneuver, and landing; performing staged mapping processing on the hydraulic pipeline load spectrum according to the flight stage identifiers to obtain a set of differential physical parameters for each stage, and constructing a staged digital twin pipeline model through the set of differential physical parameters for each stage; performing weight allocation processing on the multi-sensor data according to the flight stage identifiers to obtain differential monitoring data, and performing fault feature extraction processing on the differential monitoring data according to the staged digital twin pipeline model to obtain a fault feature vector; and performing prediction processing on the fault feature vector through a staged BP neural network to obtain a fault prediction result.

[0007] In the second aspect, the present application provides a system for fault prediction of an aircraft hydraulic pipeline system based on digital twin. The system for fault prediction of an aircraft hydraulic pipeline system based on digital twin includes: an identification module, configured to obtain flight parameters through the aircraft data bus, perform stage identification processing on the flight envelope according to the flight parameters to obtain four flight stage identifiers of takeoff, cruise, maneuver, and landing; a mapping module, configured to perform staged mapping processing on the hydraulic pipeline load spectrum according to the flight stage identifiers to obtain a set of differential physical parameters for each stage, and construct a staged digital twin pipeline model through the set of differential physical parameters for each stage; an allocation module, configured to perform weight allocation processing on the multi-sensor data according to the flight stage identifiers to obtain differential monitoring data, and perform fault feature extraction processing on the differential monitoring data according to the staged digital twin pipeline model to obtain a fault feature vector; a prediction module, configured to perform prediction processing on the fault feature vector through a staged BP neural network to obtain a fault prediction result.

[0008] In a third aspect, there is provided a fault prediction device for an aircraft hydraulic pipeline system based on digital twin, including: a memory and at least one processor, wherein instructions are stored in the memory; the at least one processor calls the instructions in the memory so that the fault prediction device for the aircraft hydraulic pipeline system based on digital twin executes the above-mentioned fault prediction method for the aircraft hydraulic pipeline system based on digital twin.

[0009] In a fourth aspect, there is provided a computer-readable storage medium, in which instructions are stored, and when it runs on a computer, it causes the computer to execute the above-mentioned fault prediction method for the aircraft hydraulic pipeline system based on digital twin.

[0010] In the technical solution provided by this application, by establishing a phased digital twin hydraulic pipeline fault prediction system for the entire flight envelope, the key technical problem that the prior art cannot adapt to the load differences in different flight phases is effectively solved. The technical feature of obtaining flight parameters through the aircraft data bus and performing stage identification processing on the flight envelope according to the flight parameters to obtain four flight stage identifiers breaks through the limitation of traditional fault prediction methods that ignore flight state differences, and establishes an accurate identification mechanism for four typical flight stages: takeoff, cruise, maneuver, and landing, laying a foundation for subsequent differential analysis. The technical feature of performing phased mapping processing on the hydraulic pipeline load spectrum according to the flight stage identifier to obtain a set of differential physical parameters for each stage and constructing a phased digital twin pipeline model innovatively establishes a dynamic mapping relationship between the load spectrum and the flight stage, enabling the digital twin model to adaptively adjust physical parameters according to the actual flight state, significantly improving the synchronization accuracy between the virtual model and the physical entity, and greatly enhancing the prediction accuracy in different flight stages compared with the traditional fixed-parameter model. The technical feature of performing weight assignment processing on multi-sensor data according to the flight stage identifier to obtain differential monitoring data solves the problem that the key fault features are masked due to the fixed weight of sensor data in the traditional method. Through the dynamic weight assignment mechanism, the monitoring focus of each flight stage is highlighted. In the takeoff stage, the focus is on pressure and flow changes. In the cruise stage, the focus is on temperature changes. In the maneuver stage, the focus is on vibration and stress, realizing targeted fault feature extraction. The technical feature of performing fault feature extraction processing on the differential monitoring data according to the phased digital twin pipeline model to obtain a fault feature vector establishes a fault feature extraction mechanism based on the digital twin model. By comparing the virtual and real data, abnormal patterns can be accurately identified, which has stronger real-time performance and accuracy compared with the traditional feature extraction method based on historical data. The technical feature of performing prediction processing on the fault feature vector through a phased BP neural network to obtain a fault prediction result reflects the important contribution of artificial intelligence algorithms in specific application fields. The design of the phased BP neural network fully considers the fault feature differences of the aircraft hydraulic pipeline system in different flight phases. Each dedicated network is specially trained for the physical characteristics and fault modes of a specific stage. The network structure and parameter configuration are optimized. The takeoff dedicated network focuses on learning transient pressure shock patterns. The cruise dedicated network focuses on learning temperature gradient change patterns. The maneuver dedicated network focuses on learning stress concentration and vibration abnormal patterns. This phased artificial intelligence algorithm design has significant improvements in key performance indicators such as fault identification accuracy, false alarm rate control, and prediction response time compared with the traditional single BP network model. Especially when dealing with the characteristics of complex mechanism, multi-source excitation, multi-field multi-scale, and strong nonlinearity of the hydraulic pipeline system, the phased BP neural network can better learn and express complex fault evolution laws, and the generalization ability and robustness of the algorithm are significantly enhanced.Overall, the present invention realizes the technological leap from passive maintenance to predictive maintenance by constructing a phased adaptive digital twin fault prediction system, provides technical support for the whole-life cycle health management of the aircraft hydraulic pipeline system, effectively reduces the maintenance cost, and improves flight safety. Description of the Drawings

[0011] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0012] Figure 1 It is a schematic diagram of an embodiment of the method for predicting faults in an aircraft hydraulic pipeline system based on digital twins in the embodiments of the present application; Figure 2 It is a schematic diagram of an embodiment of the system for predicting faults in an aircraft hydraulic pipeline system based on digital twins in the embodiments of the present application; Figure 3 It is a structural schematic block diagram of the device for predicting faults in an aircraft hydraulic pipeline system based on digital twins in the embodiments of the present invention. Detailed Embodiments

[0013] The embodiments of the present application provide a method and system for predicting faults in an aircraft hydraulic pipeline system based on digital twins. The terms "first", "second", "third", "fourth", etc. (if any) in the specification, claims and drawings of the present application are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "comprising" or "having" and any variation thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0014] For ease of understanding, the specific processes of the embodiments of the present application will be described below. Please refer to Figure 1 , an embodiment of the method for predicting faults in an aircraft hydraulic pipeline system based on digital twins in the embodiments of the present application includes: Step S101: Obtain flight parameters through the aircraft data bus, perform stage identification processing on the flight envelope according to the flight parameters, and obtain four flight stage identifiers: takeoff, cruise, maneuver, and landing; Step S102: Perform a phased mapping process on the hydraulic pipeline load spectrum according to the flight phase identifier to obtain a set of differential physical parameters for each phase, and construct a phased digital twin pipeline model through the set of differential physical parameters for each phase; Step S103: Perform a weight assignment process on the multi-sensor data according to the flight phase identifier to obtain differential monitoring data, and perform a fault feature extraction process on the differential monitoring data according to the phased digital twin pipeline model to obtain a fault feature vector; Step S104: Perform a prediction process on the fault feature vector through a phased BP neural network to obtain a fault prediction result.

[0015] It can be understood that the execution subject of this application can be a fault prediction system for an aircraft hydraulic pipeline system based on digital twins, or it can also be a terminal or a server, and specific limitations are not made here. In this embodiment of the application, the server is used as the execution subject for illustration.

[0016] Specifically, the aircraft data bus collects key flight parameters such as flight altitude, airspeed, angle of attack, overload factor, and rudder surface deflection angle in real time. These parameters are normalized to form a flight state feature vector. The flight phase discrimination matrix is a pre-set threshold judgment table. The flight state feature vector is compared with this matrix. For example, when the flight altitude is lower than 500 meters and the airspeed is greater than 80 meters per second, it is determined as the takeoff phase. When the flight altitude exceeds 3000 meters and the rate of speed change is less than 0.5 meters per second squared, it is determined as the cruise phase. The state machine algorithm generates a state transition sequence based on the discrimination result. This sequence records the change process of the flight phase and finally assigns the corresponding flight phase identifier. Secondly, the load spectrum parameter library stores the physical property data of the hydraulic pipeline in each flight phase. According to the flight phase identifier, the corresponding parameters are queried. In the takeoff phase, the transient pressure peak parameter is extracted to handle the sudden high pressure during the retraction and extension of the landing gear. In the cruise phase, the viscosity-temperature coefficient parameter of the hydraulic oil is extracted to cope with the high-altitude low-temperature environment. In the maneuvering phase, the pipeline stress concentration coefficient parameter is extracted to cope with large overload conditions. These parameters are configured into the pipeline geometric model after being aggregated. For example, the transient pressure peak parameter in the takeoff phase is matched and calculated with the pipeline wall thickness data, and the corresponding wall thickness configuration data is obtained through the pressure bearing capacity formula. The temperature coefficient parameter in the cruise phase corrects the temperature compensation of the pipeline inner diameter. The virtual mapping construction process integrates these physical parameters into the three-dimensional digital model to form a phased digital twin pipeline model that can automatically switch parameters according to the flight phase. The multi-sensor data weight distribution processing is adjusted according to the monitoring focus in different flight phases. In the takeoff phase, the weight of the pressure sensor is set to 0.4 and the weight of the flow sensor is 0.4. In the cruise phase, the weight of the temperature sensor is 0.4 and the weight of the pressure sensor is 0.3. In the maneuvering phase, the weight of the vibration sensor is 0.4 and the weight of the pressure sensor is 0.4. The weighted fusion calculation sums the product of each sensor data according to the corresponding weight to obtain the differential monitoring data reflecting the key monitoring characteristics of the current phase. The fault feature extraction processing compares the differential monitoring data with the theoretical data of the phased digital twin pipeline model point by point, calculates the data deviation value, and determines it as an abnormal state when the deviation value exceeds the set threshold. Through feature vectorization processing, a fault feature vector containing features such as pressure abnormality, temperature abnormality, flow abnormality, and vibration abnormality is generated.

[0017] The phased BP neural network selects the corresponding dedicated network model according to the flight phase identifier. For example, in the takeoff phase, the network trained for transient pressure changes is selected, and in the cruise phase, the network trained for temperature gradient changes is selected. The BP neural network inputs the fault feature vector into the hidden layer through forward propagation. After weight matrix calculation and activation function processing, the output layer generates the probability distribution of various faults, including the occurrence probabilities of pipeline rupture, seal failure, blockage, loosening, corrosion, and fatigue cracks. Finally, the maximum probability value is selected as the fault prediction result.

[0018] In a specific embodiment, the process of performing step S101 may specifically include the following steps: Collect data on flight altitude, airspeed, angle of attack, overload factor, and control surface deflection angle through the aircraft data bus, and perform normalization processing on the data of flight altitude, airspeed, angle of attack, overload factor, and control surface deflection angle to obtain a flight state feature vector; Based on the flight state feature vector, establish a flight phase discrimination matrix, and perform threshold comparison processing on the flight state feature vector to obtain a flight phase discrimination result, where the discrimination condition for the takeoff phase is a low altitude and high speed state, and the discrimination condition for the cruise phase is a high altitude and stable speed state; Input the flight phase discrimination result into the state machine algorithm for logical determination processing to obtain a state transition sequence of the current flight phase, and generate a corresponding flight phase code according to the state transition sequence; Based on the flight phase code, perform identifier allocation processing to obtain four flight phase identifiers for takeoff, cruise, maneuver, and landing, where the takeoff phase identifier is the takeoff identifier, the cruise phase identifier is the cruise identifier, the maneuver phase identifier is the maneuver identifier, and the landing phase identifier is the landing identifier.

[0019] Specifically, the aircraft data bus, as a high-speed digital communication network inside the aircraft, collects key flight parameters required for hydraulic pipeline fault prediction in real time. During the data collection process, the flight altitude sensor measures the vertical height of the aircraft from sea level, the airspeed sensor detects the movement speed of the aircraft relative to the surrounding air, the angle of attack sensor monitors the angle between the longitudinal axis of the aircraft and the relative airflow direction, the overload factor sensor records the ratio of the acceleration borne by the aircraft to the gravitational acceleration, and the control surface deflection angle sensor obtains the deflection angles of each control surface. The normalization process uniformly converts these raw data with different dimensions to the standard interval from zero to one. The specific operation is to subtract the minimum value of each parameter from the parameter value and then divide by the difference between the maximum value and the minimum value of the parameter, thereby eliminating the dimensional differences and numerical range differences between different parameters and forming a flight state feature vector. The flight phase discrimination matrix is a pre-established two-dimensional data table, where the rows represent different flight phases and the columns represent the discrimination thresholds of each flight parameter. The threshold comparison process compares each element in the flight state feature vector with the corresponding threshold in the discrimination matrix one by one. The discrimination condition for the takeoff phase is set as the flight altitude being less than the preset low altitude threshold and the airspeed being greater than the preset high speed threshold, and the discrimination condition for the cruise phase is set as the flight altitude being greater than the preset high altitude threshold and the airspeed change rate being less than the preset stability threshold. When the flight state feature vector simultaneously satisfies all the discrimination conditions of a certain phase, that phase is marked as the current flight state, forming a flight phase discrimination result.

[0020] The state machine algorithm is a logical decision-making method based on the theory of finite state automata, which processes the flight phase discrimination result as an input signal. The logical decision-making process first checks the consistency between the current discrimination result and the flight phase at the previous moment. If the same phase is determined at consecutive time points, the state transition is confirmed. If the discrimination result changes frequently, the previous state is maintained to avoid misjudgment. The state transition sequence records the change trajectory of the flight phase, such as the complete process of transitioning from the takeoff identifier to the cruise identifier and then to the maneuver identifier. The flight phase encoding generates corresponding digital encodings according to the state transition sequence. Encoding one is assigned to the takeoff phase, encoding two to the cruise phase, encoding three to the maneuver phase, and encoding four to the landing phase.

[0021] The identifier assignment process converts the digital encoding into a character identifier that is convenient for the hydraulic pipeline fault prediction system to recognize. The identifier for the takeoff phase is the takeoff identifier, which is used to trigger the monitoring mode of the hydraulic pipeline system for the transient high pressure during the retraction and extension of the landing gear. The identifier for the cruise phase is the cruise identifier, which is used to activate the analysis mode of the viscosity change of hydraulic oil in the high-altitude low-temperature environment. The identifier for the maneuver phase is the maneuver identifier, which is used to start the detection mode of pipeline stress concentration under large overload conditions. The identifier for the landing phase is the landing identifier, which is used to turn on the monitoring mode of the frequent actions of the hydraulic system during landing.

[0022] For example, when a civil airliner is performing a normal flight mission, the aircraft data bus collects the original data of a flight altitude of 8,500 meters, an airspeed of 250 meters per second, an angle of attack of 3 degrees, an overload factor of 1.02, and a control surface deflection angle of 0 degrees. The normalization process divides the flight altitude of 8,500 meters by the maximum flight altitude of 12,000 meters to get 0.708, divides the airspeed of 250 meters per second by the maximum airspeed of 300 meters per second to get 0.833, and forms a flight state feature vector after processing other parameters in the same way. The discrimination condition for the cruise phase in the flight phase discrimination matrix is that the flight altitude is greater than 0.5 and the airspeed change rate is less than 0.01. The current flight state feature vector meets these conditions, so the flight phase discrimination result is the cruise state. The state machine algorithm detects that the cruise state has been determined at the previous three time points, confirms the state transition to the cruise phase, generates a state transition sequence of the transition from takeoff to cruise, and the corresponding flight phase encoding is two. The identifier assignment process converts encoding two into the cruise identifier. After receiving the cruise identifier, the hydraulic pipeline fault prediction system automatically switches to the monitoring mode of the cruise phase, focusing on the impact of the viscosity change of hydraulic oil in the high-altitude low-temperature environment on the pipeline pressure distribution.

[0023] In a specific embodiment, the process of executing step S102 may specifically include the following steps: Query the preset load spectrum parameter library based on the flight phase identifier, and perform staged extraction processing on the hydraulic pipeline load spectrum to obtain the transient pressure peak parameter in the takeoff phase, the viscosity-temperature coefficient parameter of hydraulic oil in the cruise phase, the pipeline stress concentration coefficient parameter in the maneuver phase, and the flow pulsation frequency parameter in the landing phase; Perform parameter aggregation processing on the transient pressure peak parameter in the takeoff phase, the viscosity-temperature coefficient parameter of hydraulic oil in the cruise phase, the pipeline stress concentration coefficient parameter in the maneuver phase, and the flow pulsation frequency parameter in the landing phase to obtain the differential physical parameter sets for each phase; Perform physical property configuration processing on the pipeline geometric model according to the differential physical parameter sets for each phase to obtain a staged pipeline physical model, where the high dynamic response property is configured in the takeoff phase, the temperature sensitivity property is configured in the cruise phase, and the stress analysis property is configured in the maneuver phase; Perform virtual mapping construction processing based on the staged pipeline physical model to obtain a staged digital twin pipeline model, and the staged digital twin pipeline model includes a physical parameter adaptive switching mechanism corresponding to each flight phase.

[0024] Specifically, the load spectrum parameter library is a pre-constructed multi-dimensional data storage structure that stores the physical characteristic parameters of the hydraulic pipeline under different working conditions classified by flight phase. The query processing based on the flight phase identifier locates the corresponding data record in the parameter library through the identifier matching algorithm. The takeoff identifier triggers the access to the takeoff phase data table to extract the transient pressure peak parameter, which reflects the instantaneous high-pressure impact characteristics borne by the hydraulic pipeline during the retraction and extension of the landing gear. The cruise identifier corresponds to the cruise phase data table to extract the viscosity-temperature coefficient parameter of hydraulic oil, which describes the variation law of the flow performance of hydraulic oil in the high-altitude low-temperature environment. The maneuver identifier triggers the access to the maneuver phase data table to extract the pipeline stress concentration coefficient parameter, which quantifies the stress distribution characteristics of the key nodes of the pipeline under large overload conditions. The landing identifier corresponds to the landing phase data table to extract the flow pulsation frequency parameter, which reflects the flow fluctuation characteristics caused by the frequent actions of the hydraulic system during landing.

[0025] The parameter aggregation process structures the single parameters extracted from different flight phases to form a set of differential physical parameters for each phase. This process first establishes a parameter classification index table, classifying the transient pressure peak parameter into the pressure category, the viscosity-temperature coefficient parameter of hydraulic oil into the temperature category, the pipeline stress concentration coefficient parameter into the stress category, and the flow pulsation frequency parameter into the flow category. Subsequently, a parameter correlation analysis is carried out to check the coupling relationship between different parameters, such as the correlation between the transient pressure peak and the pipeline stress concentration coefficient, and the mutual influence between the viscosity-temperature coefficient of hydraulic oil and the flow pulsation frequency. The aggregation process combines the relevant parameters into a parameter matrix, where each row represents a flight phase and each column represents a type of physical parameter, and the matrix elements record the specific parameter values, forming a set of differential physical parameters covering the entire flight envelope.

[0026] The physical property configuration process assigns property values to the pipeline geometric model according to the set of differential physical parameters for each phase. The pipeline geometric model contains the three-dimensional geometric structure information of the pipeline, including geometric parameters such as pipe diameter, wall thickness, length, and bending radius, as well as physical parameters such as material properties and boundary conditions. The high-dynamic response property configuration in the takeoff phase writes the transient pressure peak parameter into the pressure-bearing property field of the pipeline model and adjusts the pipeline wall thickness parameter to withstand the transient high pressure. The temperature-sensitive property configuration in the cruise phase writes the viscosity-temperature coefficient parameter of hydraulic oil into the fluid property field and modifies the pipeline inner diameter parameter to adapt to the viscosity change. The stress analysis property configuration in the maneuvering phase writes the pipeline stress concentration coefficient parameter into the structural mechanics property field and adjusts the material density parameter of the key nodes to analyze the stress distribution. The configuration process generates a phased pipeline physical model, and each model is specifically optimized for the load characteristics of a specific flight phase.

[0027] The virtual mapping construction process converts the phased pipeline physical model into a virtual model in the digital twin environment. The mapping process first performs geometric data conversion, converting the three-dimensional geometric information of the pipeline physical model into a format compatible with the digital twin platform to establish the geometric structure of the virtual pipeline. Then, physical property mapping is carried out to map the physical property parameters configured for each phase into the corresponding property fields of the virtual model, establishing the association relationship between the virtual model and the physical parameters. The physical parameter adaptive switching mechanism is the key control logic. This mechanism monitors the change of the flight phase identifier and automatically calls the set of physical parameters for the corresponding phase when the identifier changes to update the property configuration of the virtual model. The switching mechanism includes a parameter verification module to ensure the continuity and consistency of the parameters during the switching process and avoid model distortion caused by parameter mutations.

[0028] In a specific embodiment, the process of performing the step of physically configuring the pipeline geometric model according to the set of differential physical parameters for each phase may specifically include the following steps: Match and calculate the transient pressure peak parameters in the take-off stage with the wall thickness data of the pipeline geometric model to obtain the pipeline wall thickness configuration data in the take-off stage, and update the wall thickness attribute value of the pipeline geometric model according to the pipeline wall thickness configuration data in the take-off stage; Perform temperature compensation calculation on the inner diameter size of the pipeline based on the viscosity-temperature coefficient parameter of the hydraulic oil in the cruise stage to obtain the temperature-corrected inner diameter data in the cruise stage, and write the temperature-corrected inner diameter data in the cruise stage into the inner diameter parameter field of the pipeline geometric model; Perform stress weighted calculation on the pipeline node position coordinates according to the pipeline stress concentration coefficient parameter in the maneuvering stage to obtain the node stress weight data in the maneuvering stage, and adjust the material density parameters of each node in the pipeline geometric model based on the node stress weight data in the maneuvering stage; Perform flow resistance calculation on the flow pulsation frequency parameter in the landing stage and the pipeline length data to obtain the pipeline flow resistance coefficient data in the landing stage, and modify the surface roughness parameter of the pipeline geometric model according to the pipeline flow resistance coefficient data in the landing stage; Perform data integration processing on the updated pipeline wall thickness attribute value, inner diameter parameter field, material density parameter and surface roughness parameter to obtain a phased pipeline physical model.

[0029] Specifically, read the initial wall thickness values of each pipe section from the pipeline geometric model, and then calculate the maximum pressure load that each pipe section needs to bear in the take-off stage according to the transient pressure peak parameters. The matching calculation is carried out through the pressure bearing capacity evaluation algorithm, comparing the transient pressure peak with the material strength limit of each pipe section. When the transient pressure peak exceeds the bearing capacity of the existing wall thickness, calculate the wall thickness value that needs to be increased. When the transient pressure peak is lower than the bearing capacity, calculate the wall thickness value that can be reduced, so as to obtain the pipeline wall thickness configuration data in the take-off stage. This configuration data contains the optimized wall thickness values of each pipe section. Subsequently, write these values into the wall thickness attribute field of the pipeline geometric model to replace the original wall thickness values, and complete the update operation of the wall thickness attribute value.

[0030] The temperature compensation calculation process corrects and adjusts the inner diameter of the pipeline based on the viscosity-temperature coefficient parameter of the hydraulic oil during the cruise stage. This calculation process first obtains the ambient temperature data during the cruise stage, and then calculates the actual viscosity value of the hydraulic oil at this temperature through the viscosity-temperature coefficient parameter of the hydraulic oil. The temperature compensation calculation uses the pipe flow theory in fluid mechanics. When the viscosity of the hydraulic oil increases due to temperature reduction, it is necessary to appropriately increase the inner diameter of the pipeline to maintain the same flow rate. When the viscosity of the hydraulic oil decreases due to temperature increase, the inner diameter of the pipeline can be appropriately reduced to save materials. The calculation process uses the Reynolds number correction formula to convert the viscosity change into the inner diameter adjustment amount, and obtains the temperature-corrected inner diameter data during the cruise stage. This data contains the optimal inner diameter values of each pipe segment during the cruise stage. Subsequently, these values are written into the inner diameter parameter field of the pipeline geometric model, overwriting the original inner diameter parameters, and completing the temperature compensation adjustment of the inner diameter size.

[0031] The stress weighted calculation process performs stress analysis and weight distribution on the pipeline node positions according to the pipeline stress concentration coefficient parameter during the maneuvering stage. This processing process first identifies all key nodes in the pipeline geometric model, including positions prone to stress concentration such as elbows, tees, and diameter changes. The stress weighted calculation applies the pipeline stress concentration coefficient parameter during the maneuvering stage to the position coordinates of each node, calculates the stress distribution state of each node under the maneuvering flight overload condition through the finite element analysis method, and assigns different weight coefficients to each node according to the stress magnitude. The higher the stress of a node, the higher its weight coefficient. Based on the calculated node stress weight data during the maneuvering stage, the material density parameters of the pipe segments corresponding to each node in the pipeline geometric model are adjusted. The pipe segments corresponding to high-weight nodes use higher-density materials to withstand greater stress, and the pipe segments corresponding to low-weight nodes use lower-density materials to reduce the overall weight.

[0032] The flow resistance calculation process combines the flow pulsation frequency parameter during the landing stage with the pipeline length data for flow resistance analysis. This calculation process first extracts the length data of each pipe segment from the pipeline geometric model, and then analyzes the flow characteristics of the hydraulic oil in the pipeline according to the flow pulsation frequency parameter during the landing stage. The flow resistance calculation uses a deformation of the Darcy-Weisbach formula to convert the flow pulsation frequency into an equivalent Reynolds number change, and calculates the flow resistance coefficient of each pipe segment through the relationship between the pipeline length and the flow resistance. When the flow pulsation frequency is high, it indicates that the hydraulic system operates frequently, and it is necessary to reduce the flow resistance of the pipeline. When the flow pulsation frequency is low, the flow resistance can be appropriately increased to improve the system stability. According to the calculated flow resistance coefficient data of the pipeline during the landing stage, the surface roughness parameter of the pipeline geometric model is modified, and the flow resistance characteristics are changed by adjusting the roughness of the inner surface of the pipeline.

[0033] The data integration process uniformly integrates the parameters updated in the foregoing steps to form a complete phased pipeline physical model. In the integration process, the compatibility among the updated pipeline wall thickness attribute values, inner diameter parameter fields, material density parameters, and surface roughness parameters is first checked to ensure the physical rationality and consistency of each parameter. The data integration adopts a method of parameter priority sorting. When there are conflicts between different parameters, the final values of the parameters are determined in the order of safety, economy, and performance optimization. The integration process also includes the verification of the correlation between parameters. For example, when the wall thickness increases, the inner diameter size needs to be adjusted accordingly; when the material density changes, the stress distribution needs to be recalculated; and when the surface roughness changes, the flow resistance coefficient needs to be corrected.

[0034] For example, when a wide-body airliner performs an intercontinental flight, the hydraulic pipeline system needs to adapt to the stage changes from takeoff to cruise. The transient pressure peak parameters in the takeoff stage show that the pressure impact on the pipeline during the retraction and extension of the main landing gear far exceeds the normal working pressure in the cruise stage. The matching calculation process compares this pressure peak with the existing wall thickness of the landing gear hydraulic pipeline and finds that the existing wall thickness has insufficient safety margin when withstanding the transient impact. It is calculated that the wall thickness of the main pipeline of the landing gear needs to be increased from the original thickness to a thicker size. When the aircraft enters the cruise stage, the low temperature environment at the cruise altitude significantly increases the viscosity of the hydraulic oil. The temperature compensation calculation process calculates the specific increase in viscosity according to the viscosity-temperature coefficient parameter of the hydraulic oil. To maintain normal flow supply, the inner diameter of the main hydraulic pipeline needs to be enlarged from the original size to a larger size. During the maneuvering flight stage, the lateral overload generated when the aircraft performs a turning maneuver causes additional stress on the elbow position of the hydraulic pipeline. The stress weighting calculation process identifies that the stress concentration degree at the elbow node is the highest, assigns the largest weight coefficient to this node, and correspondingly adjusts the pipeline material density at the elbow to a high-strength alloy material. During the landing stage, the hydraulic system frequently operates components such as flaps and landing gears, and the flow pulsation frequency is significantly higher than that in the cruise stage. The flow resistance calculation process calculates that it is necessary to reduce the flow resistance of the pipeline and reduces the flow loss by reducing the inner surface roughness of the pipeline. The data integration process uniformly integrates all these parameter adjustment results to form a phased pipeline physical model optimized for the entire flight stage of this flight.

[0035] In a specific embodiment, the process of performing step S103 may specifically include the following steps: Perform weight coefficient assignment processing on the pressure sensor data, temperature sensor data, flow sensor data, and vibration sensor data of the key nodes of the hydraulic pipeline according to the flight stage identifier to obtain a phased sensor weight coefficient table, where the pressure weight coefficient in the takeoff stage is 0.4, and the temperature weight coefficient in the cruise stage is 0.4; Based on the stage-based sensor weight coefficient table, weighted fusion calculation and processing are performed on the data of the pressure sensor, temperature sensor, flow sensor, and vibration sensor to obtain differential monitoring data, which includes the numerical values of key monitoring parameters for each flight stage. The differential monitoring data is input into the stage-based digital twin pipeline model for data comparison and analysis processing to obtain the deviation value between the monitoring data and the model data, and the abnormal degree value of each monitoring point is calculated based on the deviation value. Feature vectorization processing is performed on the abnormal degree values to obtain a fault feature vector, which includes the numerical combination of pressure anomaly features, temperature anomaly features, flow anomaly features, and vibration anomaly features.

[0036] Specifically, a correspondence table between sensor types and flight stages is established, and weight distribution is performed on the pressure sensor, temperature sensor, flow sensor, and vibration sensor according to the monitoring focus of different flight stages. During the takeoff stage, since the landing gear retraction and extension actions are frequent and the hydraulic system is subjected to transient high-pressure shocks, the pressure sensor is assigned the highest weight coefficient of 0.4, and the weight coefficient of the flow sensor is also set to 0.4 to monitor the rapid flow of hydraulic oil, the weight coefficient of the temperature sensor is set to 0.1, and the weight coefficient of the vibration sensor is set to 0.1. During the cruise stage, due to the significant influence of the high-altitude low-temperature environment on the viscosity of hydraulic oil, the temperature sensor is assigned the highest weight coefficient of 0.4, the weight coefficient of the pressure sensor is set to 0.3 to monitor the steady-state pressure, the weight coefficient of the flow sensor is set to 0.2, and the weight coefficient of the vibration sensor is set to 0.1. During the maneuvering stage, due to the increased pipeline vibration under large overload conditions, the weight coefficient of the vibration sensor is set to 0.4, the weight coefficient of the pressure sensor is set to 0.4, and the weight coefficients of the temperature and flow sensors are each 0.1. During the landing stage, the hydraulic system operates frequently, the weight coefficient of the flow sensor is 0.4, the weight coefficient of the pressure sensor is 0.3, and the weight coefficients of the temperature and vibration sensors are each 0.15, forming a complete stage-based sensor weight coefficient table. The weighted fusion calculation and processing perform numerical fusion operations on various sensor data based on the stage-based sensor weight coefficient table. This calculation process first reads the real-time data values of each sensor, and then queries the corresponding weight coefficients according to the current flight stage identifier. The fusion calculation uses the linear weighted summation method, multiplies the data value of each sensor by the corresponding weight coefficient, and then adds all the weighted numerical values to obtain the fusion result. For example, during the takeoff stage, the data of the pressure sensor is multiplied by the weight coefficient of 0.4, the data of the flow sensor is multiplied by the weight coefficient of 0.4, the data of the temperature sensor is multiplied by the weight coefficient of 0.1, and the data of the vibration sensor is multiplied by the weight coefficient of 0.1. After summing the four weighted numerical values, the differential monitoring data of this monitoring point is obtained. This data reflects the comprehensive operating state of the hydraulic pipeline in the current flight stage, highlights the changes in physical parameters that need to be most concerned about in this stage, and forms a set of differential monitoring data containing the numerical values of key monitoring parameters for each flight stage.

[0037] The data comparison and analysis process compares the differential monitoring data with the theoretical data of the phased digital twin pipeline model point by point. This process first extracts the theoretical data sequence corresponding to the current flight phase from the digital twin pipeline model, and this sequence contains the numerical range that each monitoring point should present under normal working conditions. Through comparative analysis, the deviation value of each monitoring point is obtained by calculating the numerical difference between the actual monitoring data and the theoretical data. A positive deviation value indicates that the actual data is higher than the theoretical data, and a negative deviation value indicates that the actual data is lower than the theoretical data. The calculation of the abnormality degree is quantitatively evaluated based on the magnitude and duration of the deviation value. When the deviation value exceeds the preset normal fluctuation range, the abnormality degree value is calculated according to the exceeding degree. The larger the deviation value, the higher the abnormality degree value, and the longer the duration, the greater the increase in the abnormality degree value. The abnormality degree value adopts a standardized interval from zero to one, where zero represents completely normal, one represents severe abnormality, and intermediate values represent different degrees of abnormal states.

[0038] The feature vectorization process classifies and organizes the abnormality degree values of each monitoring point according to the sensor type to form a multi-dimensional fault feature vector. This process first divides all monitoring points into four groups: pressure monitoring points, temperature monitoring points, flow monitoring points, and vibration monitoring points according to the sensor type, and then calculates the statistical characteristics of the abnormality degree values of the monitoring points within each group, including statistics such as the average value, maximum value, variance, and change trend. The pressure abnormality feature is obtained by statistically analyzing the abnormality degree values of all pressure monitoring points, reflecting the pressure abnormality mode of the hydraulic pipeline. The temperature abnormality feature is obtained by statistically analyzing the data of the temperature monitoring points, reflecting the thermodynamic abnormality mode. The flow abnormality feature and the vibration abnormality feature are obtained by statistically analyzing the data of the corresponding monitoring points respectively. The fault feature vector arranges the four types of abnormality features in a fixed order to form a four-dimensional numerical vector, which contains the numerical combination of the pressure abnormality feature, the temperature abnormality feature, the flow abnormality feature, and the vibration abnormality feature. The numerical range of each dimension is also from zero to one, and the overall pattern of the vector reflects the fault feature distribution of the hydraulic pipeline system.

[0039] For example, when a regional airliner is performing regional flights, the hydraulic pipeline system transfers from the takeoff stage to the cruise stage, and the flight stage identifier correspondingly switches from the takeoff identifier to the cruise identifier. The weight coefficient allocation process queries the weight coefficient table according to the cruise identifier, adjusts the weight of the temperature sensor from 0.1 in the takeoff stage to 0.4 in the cruise stage, adjusts the weight of the pressure sensor from 0.4 to 0.3, adjusts the weight of the flow sensor from 0.4 to 0.2, and keeps the weight of the vibration sensor unchanged at 0.1. The weighted fusion calculation process reads the hydraulic oil temperature data displayed by the main hydraulic pipeline temperature sensor. This data significantly decreases due to the low-temperature environment at cruise altitude. By multiplying by the high weight coefficient of 0.4, the temperature data dominates in the fusion result. At the same time, the pressure sensor data is multiplied by the weight coefficient of 0.3, the flow sensor data is multiplied by the weight coefficient of 0.2, and the vibration sensor data is multiplied by the weight coefficient of 0.1. The sum of the four weighted values gives the differential monitoring data at this monitoring point. The data comparison and analysis process compares this differential monitoring data with the theoretical temperature data of the digital twin pipeline model in the cruise stage, finds that the actual temperature data is lower than the normal range of the theoretical data, calculates a negative deviation value, and calculates the abnormal degree value of this monitoring point as 0.3 according to the deviation degree and duration. The feature vectorization process statistically analyzes the abnormal degree values of all temperature monitoring points to obtain a temperature abnormal feature value of 0.25. At the same time, statistical analysis of other types of monitoring points gives a pressure abnormal feature value of 0.1, a flow abnormal feature value of 0.05, and a vibration abnormal feature value of 0.02. Finally, a fault feature vector is formed as a numerical combination of temperature abnormality, pressure abnormality, flow abnormality, and vibration abnormality. This vector reflects the fault feature mode mainly characterized by temperature abnormality in the hydraulic pipeline system during the cruise stage.

[0040] In a specific embodiment, the process of performing the step of inputting the differential monitoring data into the stage-based digital twin pipeline model for data comparison and analysis processing may specifically include the following steps: Classify and organize the differential monitoring data according to the monitoring point location to obtain the real-time data sequence of each monitoring point, and extract the corresponding theoretical data sequence from the stage-based digital twin pipeline model according to the flight stage identifier; Perform point-by-point difference calculation processing based on the real-time data sequence and the theoretical data sequence to obtain the data deviation value of each monitoring point, and classify and store the data deviation value according to pressure deviation, temperature deviation, flow deviation, and vibration deviation; Perform abnormal degree quantification calculation processing on each monitoring point according to the data deviation value to obtain the abnormal degree score value. Among them, the monitoring point with a deviation value exceeding 10% of the normal range is rated as slightly abnormal, and the monitoring point with a deviation value exceeding 20% of the normal range is rated as moderately abnormal; Normalize the abnormal degree score value to obtain the abnormal degree value of each monitoring point, and the abnormal degree value is a normalized value ranging from zero to one.

[0041] Specifically, the classification and sorting process performs a spatial grouping operation on the differential monitoring data according to the physical installation positions of the sensors in the hydraulic pipeline. This process first establishes a monitoring point position index table, numbers and classifies all the sensors in the pipeline system according to their specific positions in the pipeline network. The position classification uses the pipeline topology structure analysis method to classify the monitoring points according to structural features such as the main pipeline, branch pipeline, elbow position, tee position, and valve position. Each position category contains multiple specific monitoring point numbers. The sorting process combines the differential monitoring data of the same position category and arranges them in chronological order to form a real-time data sequence, which records the continuous data changes of each monitoring point over a period of time. At the same time, query the phased digital twin pipeline model according to the current flight phase identifier, and extract the theoretical data sequence that exactly corresponds to the real-time monitoring point position from the model database. This theoretical sequence contains the standard numerical change pattern that each monitoring point should present under normal working conditions in the current flight phase.

[0042] The point-by-point difference calculation process performs a one-to-one numerical comparison operation on the real-time data sequence and the theoretical data sequence. This calculation process uses the method of comparing each time point, and subtracts the value at each time point in the real-time data sequence from the value at the corresponding time point in the theoretical data sequence. The specific operation of the difference calculation is to subtract the theoretical standard value from the real-time monitoring value to obtain positive and negative deviation values. A positive deviation value indicates that the actual data is higher than the theoretical standard, a negative deviation value indicates that the actual data is lower than the theoretical standard, and a zero deviation value indicates that the actual data is exactly the same as the theoretical standard. After the calculation is completed, the data deviation values of all monitoring points are classified and stored according to the sensor type. The deviation values of the pressure sensors are classified into the pressure deviation category, the deviation values of the temperature sensors are classified into the temperature deviation category, the deviation values of the flow sensors are classified into the flow deviation category, and the deviation values of the vibration sensors are classified into the vibration deviation category, forming four independent deviation data storage areas.

[0043] The abnormal degree quantification calculation process numerically evaluates the severity of abnormalities at each monitoring point based on the magnitude of the data deviation value. This calculation process first establishes an abnormal degree scoring standard and compares and analyzes the deviation value with the preset normal operating range. The normal operating range is a numerical interval obtained based on the design parameters of the hydraulic pipeline system and historical operation data statistics. When the deviation value of the monitoring point is within this interval, it is determined to be in a normal state. The quantification calculation uses a hierarchical scoring method. When the deviation value exceeds the normal range but the magnitude is small, the score is the mild abnormality level. When the deviation value exceeds the normal range and the magnitude is large, the score is the moderate abnormality level. When the deviation value severely exceeds the normal range, the score is the severe abnormality level. The score values use an integer scoring system, with a score of zero for the normal state, a score of one for mild abnormality, a score of two for moderate abnormality, and a score of three for severe abnormality, forming the abnormal degree score values for each monitoring point. The normalization process converts the abnormal degree score values into a standardized numerical range. This processing process uses the maximum value normalization method, dividing the score value by the maximum score value to obtain a standardized numerical value between zero and one. The specific operation of the normalization calculation is to divide the score value of each monitoring point by the maximum value of three in the scoring system to obtain the abnormal degree value between zero and one, where zero represents the completely normal state, 0.33 represents the mild abnormal state, 0.67 represents the moderate abnormal state, and one represents the severe abnormal state. The normalization process eliminates the numerical jumps between different scoring levels and forms a continuously varying abnormal degree value.

[0044] In a specific embodiment, the process of executing step S104 may specifically include the following steps: Select the corresponding BP neural network model according to the flight phase identifier for model matching processing to obtain the current phase-specific BP neural network. Among them, the takeoff-specific BP neural network is selected during the takeoff phase, the cruise-specific BP neural network is selected during the cruise phase, and the maneuver-specific BP neural network is selected during the maneuver phase; Input the fault feature vector into the current phase-specific BP neural network for forward propagation calculation processing to obtain the hidden layer output data, and perform a non-linear transformation calculation on the hidden layer output data through the activation function; Based on the results of the non-linear transformation calculation, perform output layer weight calculation processing to obtain the fault type probability distribution data, which includes the pipeline rupture probability, seal failure probability, blockage probability, loosening probability, corrosion probability, and fatigue crack probability; Perform the maximum probability selection processing on the fault type probability distribution data to obtain the fault prediction result, which includes the predicted fault type and the corresponding confidence value.

[0045] Specifically, the model matching process selects the corresponding dedicated network model from the pre-trained BP neural network model library according to the flight phase identifier. This process first establishes a correspondence table between the flight phases and the BP neural network models, with each flight phase corresponding to a specifically trained BP neural network model. The takeoff dedicated BP neural network is a network model trained for the transient pressure shock of the hydraulic pipeline and the frequent movement characteristics of the landing gear during the takeoff phase. The number of input layer nodes and the hidden layer structure of this network are optimized specifically for the dimension of the fault feature vector during the takeoff phase. The cruise dedicated BP neural network is a network model trained for the viscosity change of hydraulic oil and temperature anomaly characteristics in the high-altitude and low-temperature environment during the cruise phase. This network focuses on learning the fault patterns related to temperature. The maneuver dedicated BP neural network is a network model trained for the pipeline stress concentration and vibration anomaly characteristics under large overload conditions during the maneuver flight phase. This network specifically identifies the fault types related to stress and vibration. The matching process automatically loads the corresponding dedicated BP neural network from the model library by querying the current flight phase identifier and sets it as the dedicated BP neural network for the current phase.

[0046] The forward propagation calculation process uses the fault feature vector as the input data and passes it to the dedicated BP neural network for the current phase for calculation. In this calculation process, each element in the fault feature vector is first input into the input layer nodes of the network, and each input layer node receives one element of the feature vector. The forward propagation calculation uses the method of weighted summation, multiplies the data in the input layer by the connection weights and then sums them to obtain the weighted input values of each node in the hidden layer. The output data of the hidden layer is calculated by substituting the weighted input values into the net input function of the hidden layer. This function numerically combines all the weighted connections from the input layer to the hidden layer. The activation function performs a non-linear transformation calculation on the output data of the hidden layer. The activation function uses the Sigmoid function or the ReLU function to convert the linear combination result of the hidden layer into a non-linear output. The role of the non-linear transformation calculation is to introduce non-linear characteristics, enabling the BP neural network to learn and identify complex fault patterns. The transformed data is passed as the final output of the hidden layer to the output layer.

[0047] The output layer weight calculation process performs the final fault type prediction calculation based on the results of the non-linear transformation. This process multiplies the non-linear transformation results of the hidden layer by the connection weights of the output layer, and then sums up all the weighted connections from the hidden layer to the output layer. The output layer contains six nodes, corresponding to six fault types: pipeline rupture, seal failure, blockage, loosening, corrosion, and fatigue crack. The calculation result of each output node represents the occurrence probability of the corresponding fault type. The weight calculation uses the softmax function to perform probability normalization on the original calculation results of the output layer, ensuring that the sum of the probability values of the six fault types is equal to one, and at the same time restricting all probability values within the range of zero to one. The fault type probability distribution data contains six values, each value representing the occurrence possibility of a specific fault type. The larger the value, the higher the probability of the occurrence of this fault type. The maximum probability selection process conducts a comparative analysis on the fault type probability distribution data to find the fault type with the largest probability value as the prediction result. This process uses a numerical comparison algorithm to compare the magnitudes of the probability values of the six fault types one by one, and identifies the largest probability value and its corresponding fault type. The fault prediction result includes two parts. The first part is the predicted fault type, that is, the name of the fault type with the largest probability value, and the second part is the corresponding confidence value, that is, the probability value of this fault type. The confidence value reflects the reliability of the prediction result. The higher the confidence value, the more credible the prediction result. When the difference between the maximum probability value and the second largest probability value is small, it indicates a mixed situation of multiple fault types and further diagnostic analysis is required.

[0048] The above describes the method for predicting faults in an aircraft hydraulic pipeline system based on digital twin in the embodiments of the present application. Next, the system for predicting faults in an aircraft hydraulic pipeline system based on digital twin in the embodiments of the present application will be described. Please refer to Figure 2 , an embodiment of the system for predicting faults in an aircraft hydraulic pipeline system based on digital twin in the embodiments of the present application includes: An identification module 201, configured to obtain flight parameters through an aircraft data bus, perform stage identification processing on the flight envelope according to the flight parameters, and obtain four flight stage identifiers: takeoff, cruise, maneuver, and landing; A mapping module 202, configured to perform stage-by-stage mapping processing on the hydraulic pipeline load spectrum according to the flight stage identifier, obtain a set of stage-differentiated physical parameters, and construct a stage-by-stage digital twin pipeline model through the set of stage-differentiated physical parameters; An allocation module 203, configured to perform weight allocation processing on the multi-sensor data according to the flight stage identifier to obtain differentiated monitoring data, and perform fault feature extraction processing on the differentiated monitoring data according to the stage-by-stage digital twin pipeline model to obtain a fault feature vector; A prediction module 204, configured to perform prediction processing on the fault feature vector through a phased BP neural network to obtain a fault prediction result.

[0049] above Figure 2 From the perspective of modular functional entities, the fault prediction system of the aircraft hydraulic pipeline system based on digital twin in the embodiments of the present invention is described in detail. Next, the fault prediction device of the aircraft hydraulic pipeline system based on digital twin in the embodiments of the present invention is described in detail from the perspective of hardware processing.

[0050] Referring to Figure 3 , an aircraft hydraulic pipeline system fault prediction device based on digital twin is further provided in the embodiments of the present invention. The aircraft hydraulic pipeline system fault prediction device based on digital twin may be a server, and its internal structure may be as Figure 3 shown. The aircraft hydraulic pipeline system fault prediction device based on digital twin includes a processor, a memory, a display screen, an input device, a network interface, and a database connected through a system bus. Among them, the processor designed by the computer is used to provide computing and control capabilities. The memory of the aircraft hydraulic pipeline system fault prediction device based on digital twin includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the aircraft hydraulic pipeline system fault prediction device based on digital twin is used to store the corresponding data in this embodiment. The network interface of the aircraft hydraulic pipeline system fault prediction device based on digital twin is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the above method is implemented.

[0051] Those skilled in the art can understand that Figure 3 the structure shown in

[0052] is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the aircraft hydraulic pipeline system fault prediction device to which the solution of the present invention is applied.

[0053] Those skilled in the art can clearly understand that for the convenience and conciseness of description, the specific working processes of the above-described systems, systems, and units can refer to the corresponding processes in the foregoing method embodiments, and will not be described herein again.

[0054] When the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a digital twin-based aircraft hydraulic pipeline system fault prediction device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0055] The above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of various embodiments of the present invention.

Claims

1. A fault prediction method for an aircraft hydraulic pipeline system based on digital twin, characterized in that, The method includes: Obtaining flight parameters through an aircraft data bus, performing stage identification processing on the flight envelope according to the flight parameters, and obtaining four flight stage identifiers: takeoff, cruise, maneuver, and landing; Performing stage-by-stage mapping processing on the hydraulic pipeline load spectrum according to the flight stage identifiers, obtaining a set of stage-differentiated physical parameters, and constructing a stage-by-stage digital twin pipeline model through the set of stage-differentiated physical parameters; Performing weight assignment processing on the multi-sensor data according to the flight stage identifiers to obtain differentiated monitoring data, and performing fault feature extraction processing on the differentiated monitoring data according to the stage-by-stage digital twin pipeline model to obtain a fault feature vector; Performing prediction processing on the fault feature vector through a stage-by-stage BP neural network to obtain a fault prediction result.

2. The method for fault prediction of an aircraft hydraulic pipeline system based on digital twin according to claim 1, wherein, The obtaining flight parameters through an aircraft data bus, performing stage identification processing on the flight envelope according to the flight parameters, and obtaining four flight stage identifiers: takeoff, cruise, maneuver, and landing includes: Collecting flight altitude, airspeed, angle of attack, overload coefficient, and rudder surface deflection angle data through an aircraft data bus, performing normalization processing on the flight altitude, airspeed, angle of attack, overload coefficient, and rudder surface deflection angle data to obtain a flight state feature vector; Establishing a flight stage discrimination matrix based on the flight state feature vector, performing threshold comparison processing on the flight state feature vector to obtain a flight stage discrimination result, where the takeoff stage discrimination condition is a low altitude and high speed state, and the cruise stage discrimination condition is a high altitude and stable speed state; Inputting the flight stage discrimination result into a state machine algorithm for logical determination processing to obtain a state transition sequence of the current flight stage, and generating a corresponding flight stage code according to the state transition sequence; Performing identifier assignment processing based on the flight stage code to obtain four flight stage identifiers: takeoff, cruise, maneuver, and landing, where the takeoff stage identifier is a takeoff identifier, the cruise stage identifier is a cruise identifier, the maneuver stage identifier is a maneuver identifier, and the landing stage identifier is a landing identifier.

3. The method for predicting faults in an aircraft hydraulic pipeline system based on digital twin according to claim 1, wherein, The performing stage-by-stage mapping processing on the hydraulic pipeline load spectrum according to the flight stage identifiers, obtaining a set of stage-differentiated physical parameters, and constructing a stage-by-stage digital twin pipeline model includes: Querying a preset load spectrum parameter library based on the flight stage identifiers, performing stage-by-stage extraction processing on the hydraulic pipeline load spectrum, and obtaining a transient pressure peak parameter in the takeoff stage, a hydraulic oil viscosity-temperature coefficient parameter in the cruise stage, a pipeline stress concentration coefficient parameter in the maneuver stage, and a flow pulsation frequency parameter in the landing stage; Performing parameter set processing on the transient pressure peak parameter in the takeoff stage, the hydraulic oil viscosity-temperature coefficient parameter in the cruise stage, the pipeline stress concentration coefficient parameter in the maneuver stage, and the flow pulsation frequency parameter in the landing stage to obtain a set of stage-differentiated physical parameters; Perform physical property configuration processing on the pipeline geometric model according to the set of stage-differentiated physical parameters, and obtain a stage-based pipeline physical model, where high dynamic response properties are configured for the takeoff stage, temperature-sensitive properties are configured for the cruise stage, and stress analysis properties are configured for the maneuvering stage; Perform virtual mapping construction processing based on the stage-based pipeline physical model to obtain a stage-based digital twin pipeline model, and the stage-based digital twin pipeline model includes a physical parameter adaptive switching mechanism corresponding to each flight stage.

4. The method for predicting faults in an aircraft hydraulic pipeline system based on digital twins according to claim 3, wherein The performing physical property configuration processing on the pipeline geometric model according to the set of stage-differentiated physical parameters to obtain a stage-based pipeline physical model includes: Perform matching calculation processing on the transient pressure peak parameter in the takeoff stage and the wall thickness data of the pipeline geometric model to obtain the pipeline wall thickness configuration data in the takeoff stage, and update the wall thickness attribute value of the pipeline geometric model according to the pipeline wall thickness configuration data in the takeoff stage; Perform temperature compensation calculation processing on the inner diameter size of the pipeline based on the viscosity-temperature coefficient parameter of the hydraulic oil in the cruise stage to obtain the temperature-corrected inner diameter data in the cruise stage, and write the temperature-corrected inner diameter data in the cruise stage into the inner diameter parameter field of the pipeline geometric model; Perform stress weighted calculation processing on the pipeline node position coordinates according to the pipeline stress concentration coefficient parameter in the maneuvering stage to obtain the node stress weight data in the maneuvering stage, and adjust the material density parameters of each node in the pipeline geometric model based on the node stress weight data in the maneuvering stage; Perform flow resistance calculation processing on the flow pulsation frequency parameter in the landing stage and the pipeline length data to obtain the pipeline flow resistance coefficient data in the landing stage, and modify the surface roughness parameter of the pipeline geometric model according to the pipeline flow resistance coefficient data in the landing stage; Perform data integration processing on the updated pipeline wall thickness attribute value, inner diameter parameter field, material density parameter, and surface roughness parameter to obtain a stage-based pipeline physical model.

5. The fault prediction method for an aircraft hydraulic pipeline system based on digital twin according to claim 1, characterized in that, The performing weight distribution processing on the multi-sensor data according to the flight stage identifier to obtain differentiated monitoring data, and performing fault feature extraction processing on the differentiated monitoring data according to the stage-based digital twin pipeline model to obtain a fault feature vector includes: Perform weight coefficient distribution processing on the pressure sensor data, temperature sensor data, flow sensor data, and vibration sensor data of the key nodes of the hydraulic pipeline according to the flight stage identifier to obtain a stage-based sensor weight coefficient table, where the pressure weight coefficient in the takeoff stage is 0.4, and the temperature weight coefficient in the cruise stage is 0.4; Perform weighted fusion calculation processing on the pressure sensor data, temperature sensor data, flow sensor data, and vibration sensor data based on the stage-based sensor weight coefficient table to obtain differentiated monitoring data, and the differentiated monitoring data includes the numerical values of the key monitoring parameters in each flight stage; Input the differentiated monitoring data into the stage-based digital twin pipeline model for data comparison and analysis processing to obtain the deviation value between the monitoring data and the model data, and calculate the abnormal degree numerical value of each monitoring point according to the deviation value; Perform feature vectorization processing on the abnormal degree value to obtain a fault feature vector, where the fault feature vector contains a numerical combination of pressure abnormality features, temperature abnormality features, flow rate abnormality features, and vibration abnormality features.

6. The method for predicting faults in an aircraft hydraulic pipeline system based on digital twin according to claim 5, wherein, The process of inputting the differential monitoring data into the phased digital twin pipeline model for data comparison and analysis to obtain the deviation value between the monitoring data and the model data, and calculating the abnormal degree value of each monitoring point according to the deviation value includes: Classify and organize the differential monitoring data according to the monitoring point positions to obtain the real-time data sequence of each monitoring point, and extract the corresponding theoretical data sequence from the phased digital twin pipeline model according to the flight phase identifier; Perform point-by-point difference calculation processing based on the real-time data sequence and the theoretical data sequence to obtain the data deviation value of each monitoring point, and classify and store the data deviation value according to pressure deviation, temperature deviation, flow rate deviation, and vibration deviation; Perform abnormal degree quantization calculation processing on each monitoring point according to the data deviation value to obtain the abnormal degree score value. Among them, the monitoring point with a deviation value exceeding 10% of the normal range is rated as mildly abnormal, and the monitoring point with a deviation value exceeding 20% of the normal range is rated as moderately abnormal; Perform normalization processing on the abnormal degree score value to obtain the abnormal degree value of each monitoring point, and the range of the abnormal degree value is a normalized value between zero and one.

7. The method for predicting faults in an aircraft hydraulic pipeline system based on digital twins according to claim 1, wherein, The process of predicting the fault feature vector through a phased BP neural network to obtain a fault prediction result includes: Select the corresponding BP neural network model for model matching processing according to the flight phase identifier to obtain a dedicated BP neural network for the current phase. Among them, the dedicated takeoff BP neural network is selected during the takeoff phase, the dedicated cruise BP neural network is selected during the cruise phase, and the dedicated maneuver BP neural network is selected during the maneuver phase; Input the fault feature vector into the dedicated BP neural network for the current phase for forward propagation calculation processing to obtain the output data of the hidden layer, and perform non-linear transformation calculation on the output data of the hidden layer through an activation function; Perform output layer weight calculation processing based on the result of the non-linear transformation calculation to obtain the fault type probability distribution data, where the fault type probability distribution data includes pipeline rupture probability, seal failure probability, blockage probability, loosening probability, corrosion probability, and fatigue crack probability; Perform maximum probability selection processing on the fault type probability distribution data to obtain the fault prediction result, and the fault prediction result includes the predicted fault type and the corresponding confidence value.

8. A fault prediction system for an aircraft hydraulic pipeline system based on digital twin, characterized in that, For implementing the digital twin-based aircraft hydraulic pipeline system fault prediction method described in any one of claims 1-7, the digital twin-based aircraft hydraulic pipeline system fault prediction system includes: An identification module for obtaining flight parameters through the aircraft data bus, and performing stage identification processing on the flight envelope according to the flight parameters to obtain four flight phase identifiers: takeoff, cruise, maneuver, and landing; A mapping module, configured to perform stage-based mapping processing on the hydraulic pipeline load spectrum according to the flight stage identifier, obtain a set of stage-differentiated physical parameters, and construct a stage-based digital twin pipeline model through the set of stage-differentiated physical parameters; An allocation module, configured to perform weight allocation processing on the multi-sensor data according to the flight stage identifier to obtain differentiated monitoring data, and perform fault feature extraction processing on the differentiated monitoring data according to the stage-based digital twin pipeline model to obtain a fault feature vector; A prediction module, configured to perform prediction processing on the fault feature vector through a stage-based BP neural network to obtain a fault prediction result.

9. An aircraft hydraulic pipeline system fault prediction device based on digital twin, characterized in that, It includes a memory and a processor. The memory stores a computer program that can run on the processor. When the processor executes the computer program, it implements the digital twin-based aircraft hydraulic pipeline system fault prediction method according to any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program runs on the processor, it causes the processor to execute the digital twin-based aircraft hydraulic pipeline system fault prediction method according to any one of claims 1 to 7.

Images