A generalized method and platform for verifying and evaluating fault prediction and health management
By constructing an evaluation index system and a fault simulation mathematical model, designing test cases, and conducting dual-dimensional verification, the complexity and high cost of the PHM system were solved, and the reliability and effectiveness of the PHM equipment were achieved.
Patent Information
- Application Number
- CN202610287104.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-03-10
- Publication Date
- 2026-05-26
AI Technical Summary
Existing fault prediction and health management (PHM) systems suffer from complexity, diversity, high cost, and lack of universality and integration in functional verification and performance evaluation, making it difficult to meet the needs of different application scenarios.
An evaluation index system was constructed, test cases were designed, and verification data streams were generated through fault simulation mathematical models and data-driven models to conduct dual-dimensional verification, including algorithm function verification and principle prototype verification, and a comprehensive evaluation report was generated.
It has achieved standardization and reliability in the functional verification and performance evaluation of PHM equipment, supported the development and optimization of PHM equipment, and ensured its effectiveness in practical applications.
Smart Images
Figure CN122089158A_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of fault prediction and health management technology, and specifically relates to a universal verification and evaluation method and platform for fault prediction and health management. Background Technology
[0002] With the increasing demands for reliability and safety in modern complex systems, Prognostics and Health Management (PHM) technology has been widely applied in fields such as aviation, aerospace, rail transportation, and industrial automation. PHM systems effectively reduce maintenance costs and improve system availability by monitoring equipment status in real time, predicting potential failures, and performing health management. However, the development and verification of PHM algorithms and prototypes face numerous challenges. On the one hand, the complexity and diversity of PHM algorithms make functional verification and performance evaluation difficult, requiring a scientific and systematic evaluation index system to quantify their performance. On the other hand, the actual operation verification of prototypes requires a large amount of high-quality data support, and the verification process is complex and costly. Furthermore, existing verification methods often lack universality and integration, making it difficult to meet the needs of PHM system development and verification in different application scenarios. Summary of the Invention
[0003] The purpose of the embodiments in this specification is to provide a generalized verification and evaluation method and platform for fault prediction and health management.
[0004] To solve the above-mentioned technical problems, the embodiments of this application are implemented in the following ways: Firstly, this application provides a generalized verification and evaluation method for fault prediction and health management, the method comprising: S1. Construct an evaluation index system: Determine the evaluation elements based on the needs analysis and industry standards, break down each evaluation element into specific evaluation indicators, select matching evaluation methods based on the evaluation indicators, and integrate the evaluation elements, evaluation indicators, and evaluation methods into an evaluation index system. S2. Constructing a fault simulation mathematical model: By analyzing the physical structure and working principle of the equipment, the fault type and mechanism are determined, and a fault mechanism model is established; a data-driven model is trained based on multi-source data; the fault mechanism model and the data-driven model are integrated to generate a fault simulation mathematical model. S3. Determine the original data: Collect normal working condition data; generate feature data containing fault information based on the fault simulation mathematical model; simulate actual working conditions to obtain actual working condition data; the normal working condition data, the feature data containing fault information, and the actual working condition data constitute the original dataset. S4. Design test cases: Based on the evaluation index system, design the test case framework and logic, generate test case templates, and adapt the test cases to the actual needs of the equipment. S5. Allocate and verify the data stream: According to the requirements of the test cases, retrieve the test data from the original dataset, allocate the normal working status data, feature data containing fault information, and actual working condition data, process the test data, and generate the verification data stream. S6. Dual-dimensional verification: Input the verification data stream into the fault prediction and health management algorithm and the prototype respectively to verify the algorithm function and the prototype of the device. S7. Comprehensive Evaluation: Based on the evaluation index system, conduct a comprehensive analysis of the verification process and results of the two-dimensional verification, calculate the comprehensive score, and generate an evaluation report.
[0005] In one embodiment, S4, designing test cases: Based on the evaluation index system, design the framework and logic of test cases, generate test case templates, and adapt the test cases to the actual needs of the equipment, including: Based on the various evaluation indicators in the evaluation indicator system, standardized test case templates for different testing objectives are generated. Receive verification requirements for specific devices, select a matching template from the test case templates, and generate test cases that match the actual requirements of the specific devices by modifying at least one of the operating parameters, fault types, and performance thresholds in the template.
[0006] In one embodiment, the test data is processed, including filtering, combining, and format conversion of the test data.
[0007] In one embodiment, the two-dimensional verification includes algorithm function verification, which includes: The verification data stream is input into the fault prediction and health management algorithm. The operating status of the equipment is simulated through a comprehensive simulation environment. The fault prediction and health management algorithm is run to detect, diagnose and predict faults in the equipment, and the output results are recorded.
[0008] In one embodiment, the two-dimensional verification further includes prototype verification, which includes: The verification data stream is input into the prototype, and the actual operating environment is simulated through the signal simulation unit and interface adapter. The prototype is then run, and the output data and operating status are recorded.
[0009] Secondly, this application provides a generalized verification and evaluation platform for fault prediction and health management, including: The evaluation indicator system construction module is used to determine evaluation elements based on needs analysis and industry standards, break down each evaluation element into specific evaluation indicators, select matching evaluation methods based on the evaluation indicators, and integrate evaluation elements, evaluation indicators, and evaluation methods into an evaluation indicator system. The fault simulation mathematical model construction module is used to determine the fault type and mechanism by analyzing the physical structure and working principle of the equipment, and to establish a fault mechanism model; to train a data-driven model based on multi-source data; and to merge the fault mechanism model and the data-driven model to generate a fault simulation mathematical model. The raw data determination module is used to collect normal working status data; generate feature data containing fault information based on the fault simulation mathematical model; simulate actual working conditions to obtain actual working condition data; and the normal working status data, feature data containing fault information, and actual working condition data constitute the raw dataset. The test case design module is used to design test case frameworks and logic based on the evaluation index system, generate test case templates, and adapt test cases to the actual needs of the equipment. The verification data stream allocation module is used to retrieve test data from the original dataset according to the requirements of the test cases, allocate normal working status data, feature data containing fault information, and actual working condition data, and process the test data to generate a verification data stream. The dual-dimensional verification module is used to input the verification data stream into the fault prediction and health management algorithm and the prototype respectively, so as to verify the algorithm function and the prototype of the device. The comprehensive evaluation module is used to comprehensively analyze the verification process and results of the two-dimensional verification based on the evaluation index system, calculate the comprehensive score, and generate an evaluation report.
[0010] Thirdly, this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the verification and evaluation method for generalized fault prediction and health management as described in the first aspect.
[0011] Fourthly, this application provides a readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the verification and evaluation method for generalized fault prediction and health management as described in the first aspect.
[0012] As can be seen from the technical solutions provided in the embodiments of this specification above, the solution is: By constructing a standardized evaluation index system, and based on the standardized verification and evaluation process of the evaluation index system, processing and processing multi-source heterogeneous data, building a unified intelligent integrated operating environment for testing algorithms and prototypes, designing customized test cases, and realizing the functional verification and performance evaluation of prototypes and PHM algorithms, we can provide support for the development, optimization, and verification of PHM devices and ensure the reliability and effectiveness of PHM devices in practical applications. Attached Figure Description
[0013] To more clearly illustrate the technical solutions in the embodiments or prior art of this specification, the drawings used in the description of the embodiments or prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this specification. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.
[0014] Figure 1 A flowchart illustrating the verification and evaluation method for the generalized fault prediction and health management provided in this application; Figure 2 A schematic diagram of the structure of the verification and evaluation platform for the generalized fault prediction and health management provided in this application; Figure 3 A schematic diagram of the structure of the electronic device provided in this application. Detailed Implementation
[0015] To enable those skilled in the art to better understand the technical solutions in this specification, the technical solutions in the embodiments of this specification will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this specification, and not all embodiments. Based on the embodiments in this specification, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this specification.
[0016] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.
[0017] Various modifications and variations can be made to the specific embodiments described in this application without departing from the scope or spirit of this application, as will be apparent to those skilled in the art. Other embodiments derived from this application will be obvious to those skilled in the art. This application specification and embodiments are merely exemplary.
[0018] The terms “include,” “including,” “have,” “contain,” etc., used in this article are all open-ended terms, meaning that they include but are not limited to.
[0019] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments.
[0020] Reference Figure 1 It illustrates a flowchart of a verification and evaluation method applicable to the generalized fault prediction and health management provided in the embodiments of this application.
[0021] like Figure 1 As shown, the validation and evaluation method for generalized fault prediction and health management may include: S1. Construct an evaluation index system: Determine the evaluation elements based on the needs analysis and industry standards, break down each evaluation element into specific evaluation indicators, select matching evaluation methods based on the evaluation indicators, and integrate the evaluation elements, evaluation indicators, and evaluation methods into an evaluation index system.
[0022] Specifically, we need to obtain specific requirements from platform users (such as PHM algorithm developers and prototype design units), such as "need to evaluate the fault detection rate of the algorithm" or "need to test the stability of the prototype in a vibration environment." These requirements point to different aspects of the system.
[0023] Industry standards can refer to general standards, specifications, or guidelines in the PHM field or related equipment fields. For example, the ARP (Aerospace Recommended Practice) standard in the aerospace field and the requirements for equipment testing and evaluation in relevant IEEE (Institute of Electrical and Electronics Engineers) standards.
[0024] The acquired requirements analysis and industry standards are summarized into several core, high-level evaluation dimensions, resulting in assessment elements. These assessment elements include functional completeness and performance stability. Functional completeness refers to whether the equipment possesses all designed functions; for example, functional completeness is used to measure whether the PHM (Prognostics and Health Management) has functions such as fault detection, diagnosis, prediction, and health management. Performance stability refers to the equipment's performance under long-term operation or high load; for example, performance stability is used to measure the consistency of the PHM's performance under long-term operation, high load, and high-load conditions.
[0025] Each evaluation element is hierarchically broken down to form specific, quantifiable or qualitative evaluation indicators, and weights are assigned to these indicators. Understandably, weights can be determined using methods such as the Analytic Hierarchy Process (AHP) or the Delphi method. For example, evaluation indicators corresponding to functional integrity include: fault detection coverage, diagnostic accuracy, prediction lead time, and completeness of health management functions. For example, evaluation indicators corresponding to performance stability include: algorithm response latency, false alarm rate under high load, long-term data processing accuracy, and performance fluctuation amplitude under operating condition disturbances.
[0026] Evaluation methods include quantitative evaluation, qualitative evaluation, and comparative evaluation. Quantitative evaluation uses quantifiable indicators to measure equipment performance through specific numerical metrics. For example, fault detection coverage is calculated using the formula: (Number of detected fault types / Total number of fault types) × 100%. Algorithm response latency is statistically analyzed using a timer. Qualitative evaluation uses non-quantifiable indicators to measure equipment performance by describing the evaluation of the PHM equipment. For example, the completeness of health management functions is graded (Excellent / Good / Average / Poor) through expert review and function-by-function comparison. Comparative evaluation is used for horizontal performance verification. It compares the equipment with other similar equipment or standards to identify strengths and weaknesses. For example, it compares the diagnostic accuracy and response latency of the PHM algorithm to be validated with industry benchmark algorithms to analyze performance differences.
[0027] The evaluation index system is used to clarify the relationship between different levels, forming a complete framework of evaluation elements, evaluation indicators, weights, and evaluation methods, and is equipped with indicator calculation rules, data collection specifications, and result judgment standards.
[0028] S2. Constructing a fault simulation mathematical model: By analyzing the physical structure and working principle of the equipment, the fault type and mechanism are determined, and a fault mechanism model is established; a data-driven model is trained based on multi-source data, and the fault mechanism model and the data-driven model are integrated to generate a fault simulation mathematical model.
[0029] Specifically, the failure mechanism model analyzes equipment failure mechanisms based on design data and historical failure statistics. The data-driven model generates failure feature data through machine learning, based on historical failure statistics, simulation data, and experimental data.
[0030] The construction of the fault mechanism model specifically involves: analyzing the physical structure and working principle of the equipment to determine the possible fault types and mechanisms of each component; establishing a fault mechanism model to clarify the laws governing the occurrence, development, and propagation of faults, as well as the relationship between fault characteristic parameters and equipment operating parameters. In other words, based on the analysis of the equipment's physical structure and working principle, the potential fault types and failure mechanisms of each component are determined. The construction process follows a technical path of "mechanism analysis and variable extraction → principle determination and model selection → mathematical equation establishment → parameter calibration and verification." First, the failure process of a specific component is analyzed to extract key "operating state variables" and "fault characteristic variables." Then, based on the failure mechanism, Paris's law, Arcard's wear formula, etc., are selected as modeling principles. Subsequently, the extracted variables are substituted into the principle formulas to establish mathematical equations describing the quantitative relationships between variables. Finally, specific constants in the equations are calibrated using equipment design data and historical statistical information, and the rationality of the model is verified through data. Through these steps, a computable model with clear physical meaning—the fault mechanism model—is ultimately obtained, with operating parameters as input and quantitative fault characteristics as output.
[0031] Collect multi-source data from the equipment, such as historical fault statistics, normal operation data (and experimental data), and relevant simulation data. Utilize machine learning algorithms to mine and analyze fault characteristics, and train a data-driven model. Through model validation and optimization, ensure that the data-driven model accurately reflects the fault characteristics. Historical fault statistics can be obtained from actual equipment maintenance records and fault reports; simulation data is generated by running simulation software. Use the collected multi-source data as a training set to train the data-driven model. The data-driven model can use deep learning models such as CNN (Convolutional Neural Network) and LSTM (Long Short-Term Memory).
[0032] When fusing a fault mechanism model and a data-driven model, a weighted fusion method can be used, or other fusion methods can be selected according to actual needs to generate a fault simulation mathematical model. The input to the fault simulation mathematical model includes operating parameters and equipment model, while the output is highly realistic feature data containing fault information, such as simulated fault vibration signals, temperature curves, and current waveforms. For example, when fusing the fault mechanism model and the data-driven model, the physical feature parameters output by the fault mechanism model are combined with the data feature parameters output by the data-driven model, and used together as input to train a unified model, generating the fault simulation mathematical model.
[0033] S3. Determine the original data: Collect normal working condition data; generate feature data containing fault information based on the fault simulation mathematical model; simulate actual working conditions to obtain actual working condition data; the normal working condition data, the feature data containing fault information, and the actual working condition data constitute the original dataset.
[0034] Specifically, normal operating status data refers to data generated when equipment is running under fault-free conditions, which can originate from the actual installation environment and the product's digital physical model. For example, under normal operating conditions, the equipment's operating data is collected through a sensor network.
[0035] Feature data containing fault information refers to the data generated when equipment malfunctions, which is fault feature data generated by a fault simulation mathematical model.
[0036] Actual operating condition data refers to data under various complex conditions that equipment may encounter in actual operating environments. Simulating actual operating conditions means simulating the environmental factors and changes in operating conditions that equipment may encounter during actual use. Data processing can be performed by adding noise, outliers, missing values, and introducing data skew, jitter, and disturbances to simulate the characteristics of data under actual operating conditions. The processed data should accurately reflect the operating status and fault characteristics of the equipment in the actual operating environment.
[0037] The raw dataset includes normal operating status data, feature data containing fault information, and actual operating condition data, which are used to provide data on the equipment under normal operation, fault conditions, and actual operating conditions for subsequent testing and verification.
[0038] S4. Design Test Cases: Based on the evaluation index system, design the framework and logic of test cases, generate test case templates, and adapt the test cases to the actual needs of the equipment, including: Based on the various evaluation indicators in the evaluation indicator system, standardized test case templates for different testing objectives are generated. Receive verification requirements for specific devices, select a matching template from the test case templates, and generate test cases that match the actual requirements of the specific devices by modifying at least one of the operating parameters, fault types, and performance thresholds in the template.
[0039] Specifically, for each specific evaluation indicator in the evaluation indicator system, it is mapped to one or more specific test scenarios that need to be verified. For example, for the fault detection rate indicator, the mapped test scenario is to use a raw dataset containing multiple known fault types and severity levels for testing. Then, for each mapped test scenario, a structured and standardized test case template is designed. For example, the test case template should at least include the following key fields to ensure the standardization and repeatability of the tests: Test objective: To clarify which evaluation metrics(s) this test case aims to verify; Associated evaluation indicator identifier: The ID (Identifier) used to associate evaluation indicators in the evaluation indicator system. Required data types and formats for validation: These specify the characteristics of the data that need to be input for this test; Test execution step sequence: Used to define the detailed steps of automated testing; Expected Result Judgment Criteria: Used to define specific quantitative standards for passing the test.
[0040] The completed standardized test case templates are categorized and stored in the test case knowledge base. Simultaneously, standardized test execution processes are stored in the process library to ensure test repeatability and consistency; data format, precision, and other specifications are stored in the data standard library to ensure the accuracy and consistency of test data.
[0041] Receive user input of verification requests for specific devices. These requests must include at least: device model / name, a list of fault modes to be verified, and the device's operating environment range.
[0042] Based on the input device verification requirements, a search and matching process is performed in the test case knowledge base to select a matching general template. This template is then parameterized by replacing abstract parameters with specific parameters for the particular device, thus completing the instantiation. The modified parameters include at least one of the following: operating condition parameters, fault type, and performance threshold.
[0043] When modifying a template, standard steps from the process library and format requirements from the data standard library are invoked to ensure that the generated test cases conform to the specifications. The generated test cases are stored in the test case library and associated with the specific original datasets they require. The test case library stores test cases for easy management and reuse.
[0044] S5. Configure and Verify Data Stream: Based on the requirements of the test cases, retrieve test data from the original dataset, allocate normal operating status data, feature data containing fault information, and actual operating condition data, and process the test data to generate a verification data stream. This processing includes filtering, combining, and format conversion of the test data.
[0045] Specifically, based on the requirements of the test cases, the data types and data to be allocated are determined. Specifically: according to the requirements of the test cases, the data timing, fault injection points, and operating condition switching instructions defined in the test cases are parsed; corresponding test data is retrieved from the original dataset; and according to the timing, injection points, and operating condition switching instructions, the normal operating state data, the feature data containing fault information, and the actual operating condition data are manually allocated through timing splicing and logical combination; and the test data is processed to generate a verification data stream.
[0046] The test data retrieved from the original dataset includes normal operating condition data, feature data containing fault information, and actual operating condition data. The test data is then filtered, combined, and format-converted according to the requirements of the test cases.
[0047] Filtering test data can include removing duplicates, invalid values, and physical limit values. It can also involve extracting normal operating state data, feature data containing fault information for corresponding fault types, and simulated actual operating conditions from the original dataset according to the data types determined by the test case requirements. Then, based on the execution steps and scenario descriptions of the test cases, the temporal and proportional relationships of the data combinations are clarified. For example, the temporal relationship can be combined in the order of normal operation phase, fault occurrence phase, and operating condition change phase, with the time proportion of each phase set according to the test case requirements.
[0048] Data format conversion ensures that the generated verification data stream can be directly read and processed by subsequent programs. Specific format conversions may include: format conversion and alignment, i.e., converting data from different sources into a unified platform standard format, and performing strict timestamp alignment for multi-channel data. Unit normalization and other operations may also be performed.
[0049] After the test data is processed as described above, a verification data stream that meets the requirements is generated for subsequent use.
[0050] Understandably, verifying the determination of the data flow can also be achieved through the human-computer interaction module, specifically by obtaining the original dataset and manually adjusting the data according to the requirements of the test cases.
[0051] S6. Dual-dimensional verification: Input the verification data stream into the fault prediction and health management algorithm and the prototype respectively to verify the algorithm function and the prototype of the device.
[0052] Specifically, the dual-dimensional verification includes two parallel verification branches: algorithm function verification and prototype verification. Both branches use the same verification data stream to ensure the comparability and consistency of the verification results. Algorithm function verification verifies the functional integrity and performance stability of the device's algorithm, ensuring that the algorithm can correctly handle faulty and normal data and identify faults. Prototype verification verifies the hardware compatibility and actual operational reliability of the device's prototype, ensuring that its functions and performance meet design requirements. This dual-dimensional verification complements software logic verification and hardware physical verification, comprehensively covering the core capability dimensions of the PHM device and avoiding the one-sidedness caused by single-dimensional verification.
[0053] In one embodiment, algorithm function verification includes: inputting the verification data stream into the fault prediction and health management algorithm, simulating the device's operating state through a comprehensive simulation environment, running the fault prediction and health management algorithm to detect, diagnose, and predict faults in the device, and recording the output results.
[0054] Specifically, algorithm function verification is performed in the algorithm verification unit, which provides a controllable and reproducible pure software simulation verification environment.
[0055] Develop a corresponding verification execution plan based on the test cases.
[0056] Setting up the algorithm verification environment includes: starting the integrated simulation environment, loading the PHM algorithm to be verified, configuring simulation parameters, and linking the fault database, model database, and PHM algorithm knowledge base. The fault database stores and manages fault data and fault types for algorithm fault diagnosis and processing verification; the model database stores system models and PHM algorithm models, providing a foundation for verification; and the PHM algorithm knowledge base accumulates design and verification experience for the PHM algorithm, providing a reference for subsequent improvements.
[0057] The verification data stream is injected into the loaded PHM algorithm via software data stream or file reading. The algorithm is then run in a comprehensive simulation environment to simulate its processing of real-time data streams. During algorithm execution, the internal state and output results are recorded, including fault detection time, fault detection results, fault diagnosis results, fault prediction results, and performance parameters. These output results are compared with data in the fault database in real-time or post-processing, and performance indicators such as diagnostic accuracy and prediction error are calculated based on the comparison results.
[0058] All process data, results, and performance metrics are stored in a structured manner and can be used by the PHM algorithm knowledge base to accumulate experience from this verification.
[0059] In one embodiment, prototype verification includes: inputting a verification data stream into the prototype, simulating the actual operating environment through a signal simulation unit and an interface adapter, running the prototype, and recording the output data and operating status.
[0060] Specifically, a prototype is typically a functional board or embedded system that integrates the PHM algorithm. Prototype verification is performed in a prototype verification unit, which constructs a hardware-in-the-loop simulation environment to directly test the physical hardware carrying the PHM functionality, i.e., the prototype. The functional board integrating the PHM algorithm is used to implement fault prediction and health management functions.
[0061] During the prototype verification process, engineering integrated control is used to manage and coordinate the entire prototype verification process, including the formulation and execution of the verification plan.
[0062] The signal simulation unit consists of a CNC power supply module, a CNC logic signal module, a fiber optic module, a 1394 module, and a load simulation module. It is used to convert digitized verification data streams into real physical signals. Specifically, the CNC power supply module provides controllable power signals to simulate power changes during actual operation; the CNC logic signal module generates and controls logic signals, simulating control signals in the equipment; the fiber optic module is used for simulating and testing fiber optic communication; the 1394 module supports communication and data transmission via the 1394 interface; and the load simulation module simulates actual load conditions to test the prototype's load-bearing capacity and performance. The interface adapter provides compatibility functionality for different interfaces to ensure the compatibility of the prototype's signal simulation unit output with external devices.
[0063] The verification data stream is input into the prototype according to the test case design logic and timing. Through signal simulation units and interface adapters, the actual operating environment of the equipment is simulated. Using integrated engineering control, the prototype is run to observe its functionality and performance under different operating conditions, and the process is recorded in the process control module. The output data and operating status of the prototype at different stages are recorded, including fault alarm signals, diagnostic results, and predicted parameters. The verification results of the prototype are stored in the data recording module, including scores for various evaluation indicators, performance curves, and comparative analysis of fault detection and diagnostic results. All data generated throughout the process is recorded in the integrated data management module. Process control controls each stage of the verification process to ensure its smooth progress. The integrated data management module manages and stores various types of data generated during the verification process, including test data and verification results.
[0064] Understandably, the aforementioned algorithm functionality verification and prototype verification can be executed in parallel or sequentially. By comparing the output of the pure algorithm model and the output of the prototype under the same verification data stream, we can assess the performance loss and consistency of the algorithm after porting it from the software model to the hardware environment, and also discover the actual impact of hardware resource constraints on algorithm performance; at the same time, we can also verify the stability and reliability of the prototype in complex environments. Moreover, the dual-dimensional parallel verification verifies both the logical correctness and robustness of the PHM algorithm and confirms the hardware responsiveness and compatibility of the prototype. The results of the two types of verification corroborate each other, providing a comprehensive and reliable data foundation for subsequent comprehensive evaluation.
[0065] S7. Comprehensive Evaluation: Based on the evaluation index system, conduct a comprehensive analysis of the verification process and results of the two-dimensional verification, calculate the comprehensive score, and generate an evaluation report.
[0066] Specifically, based on the evaluation index system, the entire verification process is analyzed, including data processing, test case design, data allocation, and simulation testing. The rationality, effectiveness, and reliability of the verification process are evaluated, and potential problems and areas for improvement are analyzed.
[0067] A comprehensive analysis of the verification results of the PHM algorithm and the prototype is conducted, and a comprehensive score is calculated based on the evaluation index system. An evaluation report is generated, presenting the evaluation results in the form of charts, tables, etc., including scores for each evaluation index, performance curves, and comparative analysis of fault detection and diagnosis results. The evaluation results are analyzed to determine the performance advantages and disadvantages of the PHM algorithm and the prototype. The evaluation results can be presented visually.
[0068] The comprehensive evaluation process involves evaluating the overall process of project control management and coordination, including the development and execution of the evaluation plan.
[0069] After comprehensive analysis, the data from the evaluation process can be replayed to facilitate analysis and reproduction.
[0070] The generalized fault prediction and health management verification and evaluation method provided in this application embodiment constructs a standardized evaluation index system, processes and manipulates multi-source heterogeneous data based on the standardized verification and evaluation process of the evaluation index system, builds an intelligent integrated operating environment for unified algorithm and prototype testing, designs customized test cases, realizes functional verification and performance evaluation of prototype and PHM algorithm, provides support for the development, optimization and verification of PHM equipment, and ensures the reliability and effectiveness of PHM equipment in practical applications.
[0071] Reference Figure 2 The diagram illustrates the structure of a generalized fault prediction and health management verification and evaluation platform described according to an embodiment of this application.
[0072] like Figure 2 As shown, the universal fault prediction and health management validation and evaluation platform may include: The evaluation index system construction module 210 is used to determine the evaluation elements based on the needs analysis and industry standards, break down each evaluation element into specific evaluation indicators, select matching evaluation methods based on the evaluation indicators, and integrate the evaluation elements, evaluation indicators, and evaluation methods into an evaluation index system. The fault simulation mathematical model construction module 220 is used to determine the fault type and mechanism by analyzing the physical structure and working principle of the equipment, and to establish a fault mechanism model; to train a data-driven model based on multi-source data; and to merge the fault mechanism model and the data-driven model to generate a fault simulation mathematical model. The raw data determination module 230 is used to collect normal working state data; generate feature data containing fault information according to the fault simulation mathematical model; simulate actual working conditions to obtain actual working condition data; and the normal working state data, the feature data containing fault information, and the actual working condition data constitute the raw dataset. The test case design module 240 is used to design test case frameworks and logic based on the evaluation index system, generate test case templates, and adapt test cases to the actual needs of the equipment. The verification data stream allocation module 250 is used to retrieve test data from the original dataset according to the requirements of the test cases, allocate normal working status data, feature data containing fault information, and actual working condition data, and process the test data to generate a verification data stream. The dual-dimensional verification module 260 is used to input the verification data stream into the fault prediction and health management algorithm and the prototype respectively, so as to verify the algorithm function and the prototype of the device. The comprehensive evaluation module 270 is used to comprehensively analyze the verification process and results of the two-dimensional verification based on the evaluation index system, calculate the comprehensive score, and generate an evaluation report.
[0073] Optionally, the test case design module is also used for: Based on the various evaluation indicators in the evaluation indicator system, standardized test case templates for different testing objectives are generated. Receive verification requirements for specific devices, select a matching template from the test case templates, and generate test cases that match the actual requirements of the specific devices by modifying at least one of the operating parameters, fault types, and performance thresholds in the template.
[0074] Optionally, the verification data stream configuration module is also used to: filter, combine, and convert the format of test data.
[0075] Optionally, the two-dimensional verification module is also used for: The verification data stream is input into the fault prediction and health management algorithm. The operating status of the equipment is simulated through a comprehensive simulation environment. The fault prediction and health management algorithm is run to detect, diagnose and predict faults in the equipment, and the output results are recorded.
[0076] Optionally, the two-dimensional verification module is also used for: The verification data stream is input into the prototype, and the actual operating environment is simulated through the signal simulation unit and interface adapter. The prototype is then run, and the output data and operating status are recorded.
[0077] This embodiment provides a generalized fault prediction and health management verification and evaluation platform that can execute the above-described method. Its implementation principle and technical effect are similar, and will not be repeated here.
[0078] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Figure 3 The diagram shows a structural schematic of an electronic device 300 suitable for implementing embodiments of this application.
[0079] like Figure 3 As shown, the electronic device 300 includes a central processing unit (CPU) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage section 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the device 300. The CPU 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.
[0080] The following components are connected to I / O interface 305: an input section 306 including a keyboard, mouse, etc.; an output section 307 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 308 including a hard disk, etc.; and a communication section 309 including a network interface card such as a LAN card, modem, etc. The communication section 309 performs communication processing via a network such as the Internet. A drive 310 is also connected to I / O interface 305 as needed. Removable media 311, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., are installed on drive 310 as needed so that computer programs read from them can be installed into storage section 308 as needed.
[0081] In particular, according to embodiments of this disclosure, the above references Figure 1The described process can be implemented as a computer software program. For example, embodiments of this disclosure include a computer program product comprising a computer program tangibly embodied on a machine-readable medium, the computer program containing program code for performing the aforementioned verification and evaluation method for generalized fault prediction and health management. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311.
[0082] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.
[0083] The units or modules described in the embodiments of this application can be implemented in software or hardware. The described units or modules can also be located in a processor. The names of these units or modules do not necessarily constitute a limitation on the unit or module itself.
[0084] The systems, devices, modules, or units described in the above embodiments can be implemented by computer chips or entities, or by products with certain functions. A typical implementation device is a computer. Specifically, a computer can be, for example, a personal computer, a laptop computer, a mobile phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or any combination of these devices.
[0085] In another aspect, this application also provides a storage medium, which may be the storage medium included in the aforementioned apparatus in the above embodiments; or it may be a standalone storage medium not assembled into the device. The storage medium stores one or more programs, which are used by one or more processors to execute the verification and evaluation method for generalized fault prediction and health management described in this application.
[0086] Storage media, including both permanent and non-permanent, removable and non-removable media, can be used to store information by any method or technology. Information can be computer-readable instructions, data structures, program modules, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient media, such as modulated data signals and carrier waves.
[0087] It should be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element.
[0088] The various embodiments in this specification are described in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on describing the differences from other embodiments. In particular, the system embodiments are basically similar to the method embodiments, so the description is relatively simple; relevant parts can be referred to the descriptions in the method embodiments.
Claims
1. A universal method for verifying and evaluating fault prediction and health management, characterized in that, The method includes: S1. Construct an evaluation index system: Determine evaluation elements based on demand analysis and industry standards, break down each evaluation element into specific evaluation indicators, select matching evaluation methods based on the evaluation indicators, and integrate the evaluation elements, evaluation indicators, and evaluation methods into the evaluation index system. S2. Constructing a fault simulation mathematical model: By analyzing the physical structure and working principle of the equipment, the fault type and mechanism are determined, and a fault mechanism model is established; a data-driven model is trained based on multi-source data; the fault mechanism model and the data-driven model are merged to generate a fault simulation mathematical model. S3. Determine the original data: Collect normal working state data; generate feature data containing fault information according to the fault simulation mathematical model; simulate actual working conditions to obtain actual working condition data; the normal working state data, the feature data containing fault information, and the actual working condition data constitute the original dataset; S4. Design test cases: Based on the evaluation index system, design the test case framework and logic, generate test case templates, and adapt the test cases to the actual needs of the equipment. S5. Allocate verification data stream: According to the requirements of the test case, retrieve test data from the original dataset, allocate the normal working status data, the feature data containing fault information, and the actual working condition data, and process the test data to generate a verification data stream; S6. Dual-dimensional verification: The verification data stream is input into the fault prediction and health management algorithm and the prototype respectively to verify the algorithm function and the prototype of the device. S7. Comprehensive Evaluation: Based on the evaluation index system, conduct a comprehensive analysis of the verification process and results of the two-dimensional verification, calculate the comprehensive score, and generate an evaluation report.
2. The verification and evaluation method for generalized fault prediction and health management according to claim 1, characterized in that, S4, Designing Test Cases: Based on the evaluation index system, design the test case framework and logic, generate test case templates, and adapt the test cases to the actual needs of the equipment, including: Based on the evaluation indicators in the evaluation indicator system, standardized test case templates for different testing objectives are generated. The system receives verification requirements for a specific device, selects a matching template from the test case templates, and generates test cases that match the actual requirements of the specific device by modifying at least one of the operating parameters, fault types, and performance thresholds in the template.
3. The verification and evaluation method for generalized fault prediction and health management according to claim 1, characterized in that, The processing of the test data includes: filtering, combining, and format conversion of the test data.
4. The verification and evaluation method for generalized fault prediction and health management according to claim 1, characterized in that, The dual-dimensional verification includes algorithm function verification, which includes: The verification data stream is input into the fault prediction and health management algorithm. The operating state of the equipment is simulated through a comprehensive simulation environment. The fault prediction and health management algorithm is run to detect, diagnose, and predict faults in the equipment, and the output results are recorded.
5. The verification and evaluation method for generalized fault prediction and health management according to claim 4, characterized in that, The dual-dimensional verification also includes prototype verification, which includes: The verification data stream is input into the prototype, and the actual operating environment is simulated through the signal simulation unit and interface adapter. The prototype is then run, and the output data and operating status are recorded.
6. A universal fault prediction and health management verification and evaluation platform, characterized in that, include: The evaluation index system construction module is used to determine evaluation elements based on demand analysis and industry standards, decompose each evaluation element into specific evaluation indicators, select matching evaluation methods based on the evaluation indicators, and integrate the evaluation elements, evaluation indicators, and evaluation methods into the evaluation index system. The fault simulation mathematical model construction module is used to determine the fault type and mechanism by analyzing the physical structure and working principle of the equipment, and to establish a fault mechanism model. A data-driven model is trained based on multi-source data; the fault mechanism model and the data-driven model are fused to generate a fault simulation mathematical model. The raw data determination module is used to collect normal working status data; and generate feature data containing fault information based on the fault simulation mathematical model. Simulate actual working conditions to obtain actual working condition data; The normal working status data, the feature data containing fault information, and the actual working condition data constitute the original dataset. The test case design module is used to design test case frameworks and logic based on the evaluation index system, generate test case templates, and adapt test cases to the actual needs of the equipment. The verification data stream allocation module is used to retrieve test data from the original dataset according to the requirements of the test cases, allocate the normal working status data, the feature data containing fault information, and the actual working condition data, and process the test data to generate a verification data stream. The dual-dimensional verification module is used to input the verification data stream into the fault prediction and health management algorithm and the prototype respectively, so as to verify the algorithm function and the prototype of the device. The comprehensive evaluation module is used to comprehensively analyze the verification process and results of the two-dimensional verification based on the evaluation index system, calculate the comprehensive score, and generate an evaluation report.
7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the verification and evaluation method for generalized fault prediction and health management as described in any one of claims 1-5.
8. A readable storage medium having a computer program stored thereon, characterized in that, When executed by the processor, the program implements the verification and evaluation method for generalized fault prediction and health management as described in any one of claims 1-5.