A fault positioning test method for a civil aircraft environment control system

Abstract

The application discloses a kind of civil aircraft environmental control system fault positioning test methods of number-real fusion, belong to civil aviation aircraft system fault digitization positioning test field, comprising: S1, real: based on physical test is carried out fault positioning;S2, number: based on the physical test data obtained in S1, execute digital twin driven fault positioning deduction;S3, number-real fusion: based on the physical test result of S1 and the digital twin deduction result of S2, execute fusion analysis and iterative optimization based on Bayes: the deviation of physical test and digital twin simulation result is calculated;If deviation exceeds threshold value, then utilize the parameter of Bayes method optimization digital twin model and adjust physical test scheme, iteratively execute S1-S2 until deviation meets requirements;Output final fusion fault positioning result.The application improves the precision of civil aircraft environmental control system fault positioning, provides reliable technical support for fault positioning and maintenance decision.

Description

Technical Field

[0001] This invention belongs to the field of digital fault location testing for civil aircraft systems, and specifically relates to a fault location test method for civil aircraft environmental control systems that integrates digital and real data. Background Technology

[0002] To ensure the flight safety and operational economy of civil aircraft, the environmental control system, as the core system for maintaining an airworthy cabin environment, directly impacts passenger comfort and the safe operation of critical onboard equipment. With the increasing integration, complexity, and intelligence of civil aircraft environmental control systems, their failure modes exhibit characteristics such as concealed propagation paths, strong coupling of multiple parameters, and cross-subsystem correlations. If a failure cannot be quickly and accurately located, it will directly affect flight safety and flight regularity, and may even trigger a chain reaction of system failures.

[0003] In existing technologies, physical testing methods locate faults by building real hardware environments and can obtain actual response data, but they have limitations such as high testing costs, poor repeatability of fault injection, and high risks of testing under extreme conditions.

[0004] Digital twins simulate system behavior by constructing virtual models, enabling low-cost fault scenario simulation. However, the accuracy of these models is highly dependent on prior knowledge input. Existing digital twin models often employ simplified physical models, which introduce modeling errors when dealing with aerodynamic and thermodynamic nonlinear characteristics and dynamic coupling effects of subsystems. This leads to significant deviations between virtual simulation results and actual physical responses, directly impacting the reliability of fault location.

[0005] In current industry practice, physical testing and digital twins are often used independently, with their data failing to form an effective closed loop. Physical test results are only used to validate models, while model correction lacks real-time physical data feedback, leading to long model iteration cycles and aimless parameter optimization. This fragmented application model cannot fully leverage the synergistic advantages of both methods, hindering further improvements in fault location accuracy. Therefore, there is an urgent need to develop a fault location testing method for civil aircraft environmental control systems that integrates digital twins and physical testing. Summary of the Invention

[0006] To address the aforementioned technical problems, this invention provides a data-real fusion fault location testing method for civil aircraft environmental control systems. Through collaborative iterative optimization of physical experiments and digital twin models, the accuracy of fault location is dynamically improved. This invention is based on the deep fusion of multi-source physical experimental data and high-fidelity digital twin models of the environmental control system. It utilizes Bayesian optimization methods to dynamically correct model parameters and optimize the test plan, constructing a fault location testing mechanism that integrates digital twins and physical experiments, significantly improving the accuracy of fault location. Without altering the physical experimental conditions of the civil aircraft environmental control system, this invention overcomes the limitations of single experimental or simulation methods through interactive verification and iterative optimization of digital twins and physical experiments, providing highly reliable technical support for fault location and maintenance decisions in civil aircraft environmental control systems.

[0007] To achieve the above objectives, the present invention adopts the following technical solution:

[0008] A fault location test method for a civil aircraft environmental control system that integrates data and reality, the method comprising the following steps:

[0009] S1, Real: Fault location based on physical experiments;

[0010] S2, Number: Based on the physical test data obtained from S1, perform fault location simulation driven by digital twin;

[0011] S3, Data-Real Fusion: Based on the physical test results of S1 and the digital twin simulation results of S2, perform Bayesian-based fusion analysis and iterative optimization: calculate the deviation between the physical test and the digital twin simulation results; if the deviation exceeds the threshold, optimize the digital twin model parameters and adjust the physical test scheme using the Bayesian method, iteratively execute S1-S2 until the deviation meets the requirements; output the final fused fault location result.

[0012] A computing device includes: at least one processor and a memory storing program instructions; when the program instructions are read and executed by the processor, the computing device performs the aforementioned data-real fusion civil aircraft environmental control system fault location test method.

[0013] Beneficial effects:

[0014] This invention integrates physical testing with digital twin technology to construct a fault location testing method for civil aircraft environmental control systems. Physical testing injects faults into a real hardware environment and collects multi-source data, providing a highly reliable calibration benchmark for the digital twin model. This effectively overcomes the inference errors caused by the simplified physical mechanisms in existing digital twin models. On the other hand, the digital twin environmental control system fault inference model can simulate extreme operating conditions and sporadic fault scenarios at low cost, compensating for the limitations of poor repeatability and high testing risk of physical testing. By dynamically adjusting model parameters and physical testing schemes using Bayesian optimization methods, this invention achieves bidirectional feedback between real physical data and virtual inference results. The final output of the fused fault location result retains the empirical reliability of physical testing while leveraging the scenario expansion capabilities of digital twins. This provides high-precision and high-reliability technical support for fault location in civil aircraft environmental control systems, helping to reduce maintenance decision-making risks and improve flight operation safety. Attached Figure Description

[0015] Figure 1 This is a flowchart of a fault location test method for a civil aircraft environmental control system that integrates data and reality, provided as an embodiment of the present invention. Detailed Implementation

[0016] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0017] like Figure 1 As shown, the present invention provides a fault location test method for a civil aircraft environmental control system based on data-real fusion, comprising the following steps:

[0018] S1, Reality: Fault location based on physical experiments, specifically including:

[0019] S1.1, Correlation Analysis within the Fault Subsystem: Based on the working principle of the functional components of the fault subsystem, analyze its internal operating logic and determine the correlation between the state parameters of each functional component;

[0020] S1.2, Analysis of the correlation between fault-related subsystems: Based on historical fault data of the environmental control system, analyze the operation process, constraints and key parameter interaction relationships of each related subsystem;

[0021] S1.3, Extraction and Reorganization of Strongly Correlated Parameters: Combining the analysis results of S1.1 and S1.2, state parameters that are strongly correlated with the target fault mode are screened out, and these state parameters are logically grouped into the same virtual detection unit to narrow the fault search range and reduce interference from redundant parameters;

[0022] S1.4, Physical Test Design and Execution: Based on historical failure modes, build a high-precision ground physical test environment, configure sensor networks and data acquisition systems, and formulate reproducible fault injection and condition monitoring schemes; execute physical tests and collect multi-source data under failure conditions;

[0023] S1.5, Physical test result storage: The system response data, fault-related parameter changes, fault propagation characterization information and preliminary location results obtained from the physical test are structured and stored in the database.

[0024] S2, Data: Based on the physical test data obtained from S1, perform digital twin-driven fault location simulation, specifically including:

[0025] S2.1, Construction of Fault Inference Model for Digital Twin Environmental Control System: Based on the composition of the civil aircraft environmental control system and the analysis in S1.1-S1.3, a high-fidelity digital twin environmental control system fault inference model is constructed, which includes the geometric model, physical model, behavioral model, rule model, control model and fault reasoning model of the civil aircraft environmental control system. It is used to infer the working state, control logic and fault characteristics of the environmental control system under fault.

[0026] S2.2, Fault Injection: In the fault simulation model of the digital twin environmental control system, the same fault conditions as the physical test in S1.4 are simulated and injected.

[0027] S2.3, Digital Twin-Driven Simulation Execution: Run the digital twin environment control system fault simulation model to simulate fault location, collect and record system response data, fault-related parameter changes, simulate fault propagation paths and fault location under the simulation environment;

[0028] S2.4, Storage of Digital Twin Simulation Results: Store the simulation results of the fault simulation model of the digital twin environmental control system in the database.

[0029] S3, Data-Real Fusion: Based on the physical test results of S1 and the digital twin simulation results of S2, perform Bayesian-based fusion analysis and iterative optimization: calculate the deviation between the physical test and digital twin simulation results; if the deviation exceeds a threshold, optimize the digital twin model parameters and adjust the physical test scheme using Bayesian methods, iteratively executing S1-S2 until the deviation meets the requirements; output the final fused fault location results, specifically including:

[0030] S3.1, Comprehensive Deviation Calculation: Based on the physical test results stored in S1.5 and the digital twin simulation results stored in S2.4, calculate the differences between the two in terms of key parameter time-series changes, fault propagation path matching degree and final fault location results, and integrate them to form a comprehensive deviation evaluation index.

[0031] S3.2, Deviation threshold judgment: Determine whether the comprehensive deviation evaluation index is less than or equal to the preset threshold; if it is satisfied, proceed to S3.5; if it is not satisfied, proceed to S3.3.

[0032] S3.3, Optimization of the Fault Inference Model and Adjustment of the Physical Test Scheme for the Digital Twin Environmental Control System: Using the Bayesian optimization method, with the comprehensive deviation evaluation index as the objective function, the key physical model parameters in the fault inference model of the digital twin environmental control system are estimated and optimized; based on the optimization results, the parameters of the fault inference model of the digital twin environmental control system are updated, and the physical test scheme in S1.4 is simultaneously optimized and adjusted to improve the quality and relevance of subsequent iteration data;

[0033] S3.4, Iterative execution: Return to S1.4, execute a new round of physical experiments using the updated test plan; and return to S2.3, execute a new round of fault location simulation under the updated digital twin environmental control system fault simulation model;

[0034] S3.5, Fault Location Result Output: When the deviation judgment in S3.2 meets the requirements, the iteration is terminated, and the current fault location result of the digital twin environment control system and the physical test result are fused using a dynamic weighted average method to obtain the final fault location result. As shown in formula (1):

[0035] (1)

[0036] in, This is the final fault location result; This is the result of fault simulation for the current digital twin environmental control system; These are the results of a physical experiment; , These are the weighting coefficients. The weighting coefficients are derived from the confidence level of the fault inference results of the current digital twin environment control system. Confidence level of physical test results The determination is dynamic, and the calculation is as shown in formula (2):

[0037] (2)

[0038] Among them, the confidence level of the current digital twin environmental control system fault simulation results Determined by historical verification errors; confidence level of physical test results Determined by historical experimental errors; through dynamic weight adjustment, high-confidence results are ensured to dominate in the fusion, thereby improving the robustness of fault location.

[0039] Furthermore, S1.1 includes:

[0040] S1.1.1, Fault Subsystem Functional Component Decomposition: Based on the functional flowchart, the fault subsystem is decomposed hierarchically to clarify the physical connection relationship, signal transmission path and fault mode of each functional component of the compressor, heat exchanger and sensor;

[0041] S1.1.2, Fault Subsystem Parameter Correlation Modeling: For the decomposed functional components, a state-space equation is established to describe the dynamic relationship between parameters such as temperature, pressure, and flow rate, as shown in formula (3):

[0042] (3)

[0043] in, This is the reciprocal vector of the state parameters. A vector of state parameters, including temperature ,pressure ,flow Physical quantities that change over time; For input control quantities, such as valve opening commands and heating power output by the controller; Describe the dynamic correlation of internal state parameters of the faulty subsystem. Describe the effect of input control variables on state parameters.

[0044] S1.1.3, Fault-Parameter Mapping Relationship Construction: Based on historical fault data, establish a mapping relationship between fault modes and key parameters to form a fault feature matrix. As shown in formula (4):

[0045] (4)

[0046] in, This is a marker for associating fault types with status parameters. Indicates the first Class of faults and the first These parameters are strongly correlated. Indicates a weak correlation. Number of fault types This represents the number of state parameters.

[0047] Furthermore, the fault-related subsystem correlation analysis in S1.2 includes:

[0048] S1.2.1, the environmental control system is decomposed into various fault-related subsystems according to functional modules, including: air source subsystem, temperature control subsystem, pressure control subsystem, cabin air distribution system, and the interfaces of material flow, energy flow, and information flow between each subsystem, as well as the key input parameters and output parameters of each subsystem;

[0049] S1.2.2, Construct a parameter interaction model between fault-related subsystems: Use a cause-effect graph to describe the parameter association across subsystems, such as formula (5) to define the output temperature of the temperature control subsystem. Input pressure of the gas source subsystem The lag relationship:

[0050] (5)

[0051] in, The temperature is output by the temperature control subsystem. Input pressure to the gas source subsystem. The pressure-temperature correlation coefficient, This is the heat load-temperature correlation coefficient. Due to time lag, This refers to the heat load.

[0052] S1.2.3, Key parameter constraint extraction: Determine the safety threshold of fault-related subsystem parameters based on the Civil Aviation Maintenance Manual;

[0053] Furthermore, the extraction and recombination of strongly correlated state parameters in S1.3 includes:

[0054] S1.3.1, Strong Correlation Analysis: Calculating the Pearson correlation coefficient between parameters based on historical data. As shown in formula (6), filter Strongly correlated parameter pairs;

[0055] (6)

[0056] in, , This represents two state parameters. Describing covariance, , This represents the standard deviation of the parameter.

[0057] S1.3.2, Parameter Clustering and Reorganization: The K-means algorithm is used to divide the strongly correlated parameters into virtual detection units, and the objective function is shown in formula (7):

[0058] (7)

[0059] in, Indicates the number of clusters. Indicates the first The parameter set in each cluster It is a single parameter vector; For the first The mean vector of each cluster.

[0060] S1.3.3, Redundant parameter removal: Principal component analysis is used to reduce dimensionality and retain principal components with a contribution rate >90%.

[0061] Furthermore, the construction of the digital twin environmental control system fault simulation model in S2.1 includes:

[0062] Geometric Model: Based on computer-aided design data of the environmental control system, a geometric model is accurately constructed to characterize the internal structure of the environmental control system, the layout of pipelines, valve positions, sensor installation points and key components, including but not limited to the geometric dimensions, spatial positions and interconnections of compressors, turbines, heat exchangers, regulating valves;

[0063] Physical Model: Based on key physical parameters such as cabin temperature, pressure, and humidity, and combined with physical property constraints and the correlation between S1.1 and S1.2, a hybrid physical equation is established based on the first law of thermodynamics and the governing equations, as shown in formula (8):

[0064] (8)

[0065] in, The current temperature. For heat input, For air quality, For isobaric specific heat capacity, Thermal resistance is used to describe the efficiency of heat exchange between the system and the cabin environment. For cabin ambient temperature, The rate of temperature change;

[0066] Behavioral Model: Construct the dynamic response of the environmental control system under typical failure modes, including temperature abrupt changes, pressure fluctuations, and flow anomalies;

[0067] Rule model: Embedded with standard fault diagnosis logic, emergency response operation rules, system safety operation boundary constraints, and domain expert experience knowledge base as specified in the Civil Aviation Maintenance Manual and Fault Isolation Manual;

[0068] Control Model: The behavior of the subsystem controller in the environmental control system is described by the PID control algorithm, as shown in formula (9), and control commands are generated and output based on sensor feedback signals and set values;

[0069] (9)

[0070] in, Indicates the controller output; To set the deviation between the state parameters and the actual state parameters; The proportional gain controls the response speed of the current deviation. This is the integral gain, used to eliminate static errors; For differential gain, This is the integral term for the error;

[0071] Fault Reasoning Model Construction: A fault reasoning model is constructed based on symbolic directed graphs. From node set and directed edge set Composition, denoted as ,gather The state parameters of the environmental control system are represented by the node states. This is represented as shown in formula (10):

[0072] (10)

[0073] in, Represents node value, , These represent the high threshold and the low threshold, respectively.

[0074] Furthermore, the Bayesian optimization method in S3.3 includes:

[0075] S3.3.1, Prior Distribution Setting: Based on expert experience and historical data, the parameters to be optimized are set... Assume a Gaussian prior distribution. ;in Based on experience points. For variance, Indicates a Gaussian distribution;

[0076] S3.3.2, Likelihood Function Construction: Based on the bias data D between the physical experiment and the digital twin, the likelihood function is defined. As in formula (11), it is assumed that the deviation follows the principle of minimizing the mean square error:

[0077] (11)

[0078] in, This dataset represents the discrepancy between the results of physical experiments and fault simulations of digital twin environmental control systems. Indicates the first Measurements from the physical experiment This indicates that the digital twin environmental control system model has parameters The predicted value below, To measure the noise variance.

[0079] S3.3.3, Solving for the posterior distribution: The posterior distribution is obtained by sampling using the Metropolis-Hastings algorithm in the Markov chain Monte Carlo method. Update the optimal estimate of the parameters, as shown in formula (12):

[0080] (12)

[0081] in, The optimal estimate of the parameter. Posterior probability distribution.

[0082] S3.3.4, Model Uncertainty Assessment: Calculate the 95% confidence interval of the posterior distribution to verify the robustness of parameter optimization.

[0083] S3.3.5 Iterative correction strategy: If the deviation still exceeds the threshold after optimization, take the current posterior distribution as the new prior distribution and repeat S3.3.3-S3.3.4 until the deviation meets the accuracy requirements for civil aircraft fault location.

[0084] Furthermore, taking cabin temperature anomaly fault location as an example, in the physical test stage: First, the faulty subsystem is analyzed. For the temperature control subsystem, six functional components are decomposed: compressor, heat exchanger, temperature sensor, expansion valve, controller, and cooling fan. The state space equation of temperature-pressure-flow is established according to formula (3), and the fault feature matrix is ​​constructed according to formula (4). "Heat exchanger blockage" and "compressor efficiency reduction" are identified as key candidate faults. Second, the correlation analysis of fault-related subsystems is carried out. Through formula (5), it is found that there is a lag correlation between the output temperature of the temperature control subsystem and the pressure of the air source subsystem. The following 12 strongly correlated parameters are extracted: heat exchanger inlet temperature Heat exchanger outlet temperature Compressor discharge pressure evaporator surface temperature condenser cooling air volume Expansion valve opening Mixing chamber temperature Cabin temperature setpoint Controller PID output signal Ambient temperature air mass flow rate Heater power Parameter reorganization and experimental design: The parameters are divided into 3 virtual detection units by formula (7), and typical data curves of heat exchanger blockage are collected after the fault injection test is performed. Digital twin simulation stage: After simulating the injection of "heat exchanger blockage" fault, the temperature propagation path is deduced: fault source (heat exchanger blockage) → outlet temperature rise ( - → Temperature sensor alarm ( → The controller opens the cooling valve ( → Compressor load increases ( → Cabin temperature fluctuations ( ). Integration and optimization stage: The temperature peak deviation between the physical test and the digital twin is calculated. The thermal resistance parameters of the fault inference model of the digital twin environmental control system are adjusted by Bayesian optimization (formula (11)-(12)). After three iterations, the deviation requirement is met. Based on formula (1) and (2), the final fault location accuracy reaches 91.6%.

[0085] The present invention also provides a computing device, comprising: at least one processor and a memory storing program instructions; when the program instructions are read and executed by the processor, the computing device performs the aforementioned data-real fusion civil aircraft environmental control system fault location test method.

[0086] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0087] Although the invention has been described with respect to a limited number of embodiments, those skilled in the art will understand from the foregoing description that other embodiments are conceivable within the scope of the invention described herein. Furthermore, it should be noted that the language used in this specification has been chosen primarily for readability and instructional purposes, and not for the purpose of explaining or limiting the subject matter of the invention.

Claims

1. A fault location test method for a civil aircraft environmental control system that integrates data and reality, characterized in that: The method includes the following steps: S1, Real: Fault location based on physical experiments; S2, Number: Combining the physical test data obtained from S1, perform fault location simulation driven by digital twin; S3, Data-Real Fusion: Based on the physical experiment results of S1 and the digital twin simulation results of S2, perform Bayesian-based fusion analysis and iterative optimization: calculate the deviation between the physical experiment and the digital twin simulation results; If the deviation exceeds the threshold, the parameters of the digital twin model are optimized using the Bayesian method and the physical test plan is adjusted. S1-S2 is executed iteratively until the deviation meets the requirements. Output the final fused fault location results; S3 includes: S3.1, Comprehensive Deviation Calculation: Based on the stored physical test results and the stored digital twin simulation results, calculate the differences between the two in terms of key parameter time-series changes, fault propagation path matching degree and final fault location results, and integrate them to form a comprehensive deviation evaluation index; S3.2, Deviation threshold judgment: Determine whether the comprehensive deviation evaluation index is less than or equal to the preset threshold; if it is satisfied, proceed to S3.5; if it is not satisfied, proceed to S3.

3. S3.3, Optimization of the fault inference model and adjustment of the physical test scheme for the digital twin environmental control system: Using the Bayesian optimization method, with the comprehensive deviation evaluation index as the objective function, the key physical model parameters in the fault inference model of the digital twin environmental control system are estimated and optimized; based on the optimization results, the parameters of the fault inference model of the digital twin environmental control system are updated, and the physical test scheme is simultaneously optimized and adjusted to improve the quality and relevance of subsequent iteration data; S3.4, Iterative Execution: Execute a new round of physical experiments using the updated test plan; execute a new round of fault location simulation under the updated digital twin environmental control system fault simulation model; S3.5, Fault Location Result Output: When the judgment deviation in S3.2 meets the requirements, the iteration is exited, and the current fault simulation result of the digital twin environment control system and the physical test result are fused using the dynamic weighted average method to obtain the final fault location result, as shown in formula (1): （1） in, This is the final fault location result; This is the result of fault simulation for the current digital twin environmental control system; These are the results of a physical experiment; , These are the weighting coefficients. The weighting coefficients are based on the confidence level of the fault inference results of the current digital twin environment control system. Confidence level of physical test results The determination is dynamic, and the calculation is as shown in formula (2): （2） Among them, the confidence level of the current digital twin environmental control system fault simulation results Determined by historical verification errors; confidence level of physical test results Determined by historical experimental errors; through dynamic weight adjustment, high-confidence results are ensured to dominate in the fusion, thereby improving the robustness of fault location.

2. The test method for fault location of civil aircraft environmental control system based on data-real fusion according to claim 1, characterized in that: S1 includes: S1.1, Correlation Analysis within the Fault Subsystem: Based on the working principle of the functional components of the fault subsystem, analyze its internal operating logic and determine the correlation between the state parameters of each functional component; S1.2, Analysis of the correlation between fault-related subsystems: Based on historical fault data of the environmental control system, analyze the operation process, constraints and key parameter interaction relationships of each related subsystem; S1.3, Extraction and Reorganization of Strongly Correlated Parameters: Combining the analysis results of S1.1 and S1.2, state parameters that are strongly correlated with the target fault mode are screened out, and these state parameters are logically grouped into the same virtual detection unit to narrow the fault search range and reduce interference from redundant parameters; S1.4, Physical Test Design and Execution: Based on historical failure modes, build a high-precision ground physical test environment, configure sensor networks and data acquisition systems, and formulate reproducible fault injection and condition monitoring schemes; execute physical tests and collect multi-source data under failure conditions; S1.5, Physical test result storage: The system response data, fault-related parameter changes, fault propagation characterization information and preliminary location results obtained from the physical test are structured and stored in the database.

3. The test method for fault location of civil aircraft environmental control system based on data-real fusion according to claim 2, characterized in that: S2 includes: S2.1, Construction of Fault Inference Model for Digital Twin Environmental Control System: Based on the composition of the civil aircraft environmental control system and the analysis in S1.1-S1.3, a high-fidelity digital twin environmental control system fault inference model is constructed, which includes the geometric model, physical model, behavioral model, rule model, control model and fault reasoning model of the civil aircraft environmental control system. It is used to infer the working state, control logic and fault characteristics of the environmental control system under fault. S2.2, Fault Injection: In the fault simulation model of the digital twin environmental control system, the same fault conditions as the physical test in S1.4 are simulated and injected. S2.3, Digital Twin-Driven Simulation Execution: Run the digital twin environment control system fault simulation model to simulate fault location, collect and record system response data, fault-related parameter changes, simulate fault propagation paths and fault location under the simulation environment; S2.4, Storage of Digital Twin Simulation Results: Store the simulation results of the fault simulation model of the digital twin environmental control system in the database.

4. The test method for fault location of civil aircraft environmental control system based on data-real fusion according to claim 3, characterized in that: S1.1 includes: S1.1.1, Fault Subsystem Functional Component Decomposition: Based on the functional flowchart, the fault subsystem is decomposed hierarchically to clarify the physical connection relationship, signal transmission path and fault mode of each functional component of the compressor, heat exchanger and sensor; S1.1.2, Fault Subsystem Parameter Correlation Modeling: For the decomposed functional components, a state-space equation is established to describe the dynamic relationship between temperature, pressure, and flow parameters, as shown in formula (3): （3） in, This is the reciprocal vector of the state parameters. A vector of state parameters, including temperature ,pressure ,flow ; For input control quantity; Describe the dynamic correlation of internal state parameters of the faulty subsystem. Describe the effect of the input control quantity on the state parameters; S1.1.3, Fault-Parameter Mapping Relationship Construction: Based on historical fault data, establish a mapping relationship between fault modes and key parameters to form a fault feature matrix. As shown in formula (4): （4） in, This is a marker for associating fault types with status parameters. Indicates the first Class of faults and the first These parameters are strongly correlated. Indicates a weak correlation. Number of fault types This represents the number of state parameters.

5. The test method for fault location of civil aircraft environmental control system based on data-real fusion according to claim 4, characterized in that: S1.2 includes: S1.2.1, the environmental control system is decomposed into various fault-related subsystems according to functional modules, including: air source subsystem, temperature control subsystem, pressure control subsystem, cabin air distribution system, and the interfaces of material flow, energy flow, and information flow between each subsystem, as well as the key input parameters and output parameters of each subsystem; S1.2.2, Construct a parameter interaction model between fault-related subsystems: Use a cause-effect graph to describe the parameter association across subsystems, such as formula (5) to define the output temperature of the temperature control subsystem. Input pressure of the gas source subsystem The lag relationship: （5） in, The temperature is output by the temperature control subsystem. Input pressure to the gas source subsystem. The pressure-temperature correlation coefficient, This is the heat load-temperature correlation coefficient. Due to time lag, For heat load; S1.2.3, Key parameter constraint extraction: Determine the safety threshold of fault-related subsystem parameters based on the Civil Aviation Maintenance Manual.

6. The test method for fault location of civil aircraft environmental control system based on data-real fusion according to claim 5, characterized in that: S1.3 includes: S1.3.1, Strong Correlation Analysis: Calculating the Pearson correlation coefficient between parameters based on historical data. As shown in formula (6), filter Strongly correlated parameter pairs; （6） in, , This represents two state parameters. Describing covariance, , This represents the standard deviation of the parameter; S1.3.2, Parameter Clustering and Reorganization: The K-means algorithm is used to divide the strongly correlated parameters into virtual detection units, and the objective function is shown in formula (7): （7） in, Indicates the number of clusters. Indicates the first The parameter set in each cluster It is a single parameter vector; For the first The mean vector of each cluster; S1.3.3, Redundant parameter removal: Principal component analysis is used to reduce dimensionality and retain principal components with a contribution rate >90%.

7. The test method for fault location of civil aircraft environmental control system based on data-real fusion according to claim 5, characterized in that: In S2.1, the construction of the fault simulation model for the digital twin environmental control system includes: Geometric Model: Based on computer-aided design data of the environmental control system, a geometric model is accurately constructed to characterize the internal structure of the environmental control system, the layout of pipelines, valve positions, sensor installation points and key components, including but not limited to the geometric dimensions, spatial positions and interconnections of compressors, turbines, heat exchangers, regulating valves; Physical Model: Based on cabin temperature, pressure, and humidity, combined with physical property constraints and the relationship between S1.1 and S1.2, a hybrid physical equation is established based on the first law of thermodynamics and the governing equations, as shown in formula (8): （8） in, The current temperature. For heat input, For air quality, For isobaric specific heat capacity, Thermal resistance is used to describe the efficiency of heat exchange between the system and the cabin environment. For cabin ambient temperature, The rate of temperature change; Behavioral Model: Construct the dynamic response of the environmental control system under typical failure modes, including temperature abrupt changes, pressure fluctuations, and flow anomalies; Rule model: Embedded with standard fault diagnosis logic, emergency response operation rules, system safety operation boundary constraints, and domain expert experience knowledge base as specified in the Civil Aviation Maintenance Manual and Fault Isolation Manual; Control Model: The behavior of the subsystem controller in the environmental control system is described by the PID control algorithm, as shown in formula (9), and control commands are generated and output based on sensor feedback signals and set values; （9） in, Indicates the controller output; To set the deviation between the state parameters and the actual state parameters; The proportional gain controls the response speed of the current deviation. This is the integral gain, used to eliminate static errors; For differential gain, This is the integral term for the error; Fault Reasoning Model Construction: A fault reasoning model is constructed based on symbolic directed graphs. From node set and directed edge set Composition, denoted as ,gather The state parameters of the environmental control system are represented by the node states. This is represented as shown in formula (10): （10） in, Represents node value, , These represent the high threshold and the low threshold, respectively.

8. The test method for fault location of civil aircraft environmental control system based on data-real fusion according to claim 7, characterized in that: The Bayesian optimization method in S3.3 includes: S3.3.1, Prior Distribution Setting: Based on expert experience and historical data, the parameters to be optimized are set... Assume a Gaussian prior distribution. ;in, Based on experience points. For variance, Indicates a Gaussian distribution; S3.3.2, Likelihood Function Construction: Based on the bias data D between the physical experiment and the digital twin, the likelihood function is defined. As shown in formula (11): （11） in, This dataset represents the discrepancy between the results of physical experiments and fault simulations of digital twin environmental control systems. Indicates the first Measurements from the physical experiment This indicates that the digital twin environmental control system model has parameters The predicted value below, To measure the noise variance; S3.3.3, Solving for the posterior distribution: The posterior distribution is obtained by sampling using the Metropolis-Hastings algorithm in the Markov chain Monte Carlo method. Update the optimal estimate of the parameters, as shown in formula (12): （12） in, This is the optimal estimate of the parameter. This is the posterior probability distribution; S3.3.4, Model Uncertainty Assessment: Calculate the 95% confidence interval of the posterior distribution to verify the robustness of parameter optimization; S3.3.5 Iterative correction strategy: If the deviation still exceeds the threshold after optimization, take the current posterior distribution as the new prior distribution and repeat S3.3.3-S3.3.4 until the deviation meets the accuracy requirements for civil aircraft fault location.

9. A computing device, characterized in that, include: At least one processor and a memory storing program instructions; When the program instructions are read and executed by the processor, the computing device performs the fault location test method for the civil aircraft environmental control system based on the fusion of data and reality as described in any one of claims 1-8.

Images