Fault comprehensive diagnosis logic generation method considering multiple working conditions
By extending FMEA and the principle of 'detect first, then isolate', a fault detection and isolation logic adapted to multiple operating conditions is generated, which solves the problem of fault misjudgment or missed detection in traditional methods and realizes accurate fault identification and isolation of complex systems under multiple operating conditions.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-12-11
- Publication Date
- 2026-03-31
AI Technical Summary
Traditional fault diagnosis methods based on a single threshold or fixed model cannot effectively address the problem of misjudgment or missed detection of faults in modern complex systems under multiple operating conditions, and are difficult to meet the diagnostic needs of complex systems under multiple operating conditions.
Based on Extended Failure Mode and Effects Analysis (FMEA), and following the principle of 'detect first, then isolate', a unified diagnostic logic framework is established. By extracting the failure modes and condition monitoring parameters of key components, fault detection and isolation logic adapted to different operating conditions is generated.
It enables accurate fault identification and isolation of complex systems under multiple operating conditions, and generates a hierarchical set of diagnostic logic rules adapted to different operating scenarios, thereby improving the accuracy and comprehensiveness of fault diagnosis.
Smart Images
Figure CN121764024A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a method for generating comprehensive fault diagnosis logic that considers multiple operating conditions, belonging to the field of fault diagnosis; it is applicable to the generation of multi-condition fault diagnosis logic for modern complex systems. Background Technology
[0002] With the rapid development of electronic and intelligent technologies, the structural complexity of modern complex systems has significantly increased, and their operating states exhibit diverse and dynamically changing characteristics. In practical applications, the manifestations of faults and their corresponding key monitoring parameters often differ under different operating conditions. Traditional fault diagnosis methods based on single thresholds or fixed models, failing to fully consider these differences in operating conditions, are prone to misdiagnosis or missed detection of faults, making it difficult to meet the diagnostic needs of complex systems operating under multiple conditions.
[0003] In the field of fault diagnosis and health management, rule-based logic diagnostic methods have become a key technical means for fault detection and isolation due to their intuitiveness and interpretability. The core of these methods lies in establishing a precise correlation between fault modes and observable monitoring parameters through in-depth analysis of the system's functional structure and fault mechanisms, and generating reliable diagnostic logic rules accordingly to directly determine the occurrence of faults.
[0004] Therefore, there is an urgent practical need to develop a comprehensive method that can effectively integrate multi-condition information and generate diagnostic logic to achieve accurate identification of faults in complex systems. Summary of the Invention
[0005] This invention provides a method for generating comprehensive fault diagnosis logic considering multiple operating conditions, aiming to solve the problem of insufficient coverage of multi-condition fault diagnosis in existing technologies. Based on the results of Extended Fault Mode and Effects Analysis (FMEA), this method establishes a unified diagnostic logic framework according to the principle of "detect first, then isolate," enabling fault detection and isolation of complex systems under various operating conditions.
[0006] The method includes the following steps:
[0007] S1: For the device S with the fault to be tested, extract the set of fault modes of each key component from historical data and obtain the set of condition monitoring parameters;
[0008] Historical data includes: typical failure cases of objects, historical fault data, reliability information, and domain expert knowledge;
[0009] Extract the main failure modes of each key component of the device under test S to form a failure mode set; each failure mode includes failure number, failure name, failure occurrence probability level (high, medium and low), severity level (I, II and III) and failure evolution characteristics (abrupt / gradual change).
[0010] The status monitoring parameters for each key component include its number, name, type, and description.
[0011] S2: Based on the operating characteristics of the equipment S with the fault to be tested, the operating state is divided into several typical operating condition sets;
[0012] The resulting set of operating conditions is represented as follows:
[0013]
[0014] Indicates the first Each working condition;
[0015] Each operating condition includes: operating condition number, operating condition name, operating condition characterization parameters, and corresponding discrimination conditions.
[0016] Operating condition characterization parameters are parameters that characterize the operating state of equipment S.
[0017] The comparison between the characterization parameters of each working condition and their respective set thresholds, combined with logical relationships, constitutes the discrimination conditions for each working condition.
[0018] S3: For each typical operating condition, set the fault criteria for each operating condition based on the failure modes of key components and the associated status monitoring parameters under each condition.
[0019] The specific steps are as follows:
[0020] For the current typical operating conditions, identify the failure mode set of each key component under these conditions. and the set of condition monitoring parameters involved ; The number of failure modes; The number of condition monitoring parameters;
[0021] No. One fault mode The fault criterion consists of its associated / related state monitoring parameters, the fault detection thresholds for each parameter, and logical expressions, used to formally describe the fault triggering conditions; its form is:
[0022]
[0023] in, For logical comparison, For logical operators, including AND ( ),or( ),No( Operators such as ) Indicates the first Each condition monitoring parameter corresponds to a fault detection threshold; the fault detection threshold for each condition monitoring parameter is not unique. Indicates the fault mode The number of associated / related status monitoring parameters.
[0024] S4: Utilize the failure modes of key components under various operating conditions, associated condition monitoring parameters, and failure criteria to form a set of test units.
[0025] The test unit set includes test number, test name, status monitoring parameters, and fault detection logic;
[0026] For the Status monitoring parameters , and the corresponding fault detection threshold Perform logical comparisons separately This forms a fault detection logic; the result of each fault detection logic, combined with the test number, test name, and status monitoring parameters, forms a test unit; therefore, the same status monitoring parameter may correspond to multiple test units.
[0027] The fault detection threshold is used to identify deviation characteristics of monitoring parameters for fault modes under different operating conditions, and to establish detection rules accordingly. The selection of the threshold includes two categories:
[0028] Operating condition threshold: The normal fluctuation range of the same parameter is different under different operating conditions, so different thresholds need to be set to ensure accurate detection.
[0029] Multi-level thresholds: Under the same operating conditions, different faults cause different degrees of deviation from parameters, requiring multi-level thresholds to achieve fine differentiation.
[0030] S5: Utilize the test numbers of the test unit set and combine them with fault criteria to generate comprehensive fault diagnosis logic.
[0031] Specifically:
[0032] First, based on the fault criteria and test units under each working condition, the fault detection logic is sorted out to obtain a fault isolation logic table composed of each test number;
[0033] When a test unit triggers an alarm, it indicates that the equipment has malfunctioned, thus realizing fault detection. Then, by combining the mapping relationship between the test unit and the fault criteria, the judgment rules corresponding to various faults under different operating conditions are constructed by sorting out the fault detection logic, that is, the fault isolation logic. The fault isolation logic corresponding to all test units under each operating condition is formed into a fault isolation logic table.
[0034] Secondly, the alarm information of all test units is comprehensively analyzed and matched with the characteristic patterns of each fault mode to determine the specific fault that occurred and achieve fault isolation.
[0035] Specifically, the process involves finding the corresponding fault number and fault name for each test number, integrating the fault isolation logic tables for different operating conditions, and summarizing the results of multiple operating conditions under the same fault number to form a comprehensive fault diagnosis logic.
[0036] The advantages and positive effects of this invention are as follows:
[0037] First, traditional FMEA (Failure Mode and Effects Analysis) data is expanded. Through in-depth failure mode characteristic analysis and testability design considerations, the monitoring parameters and quantitative detection criteria corresponding to each failure mode under multiple operating conditions are clarified, laying a data foundation for the generation of diagnostic logic. Second, following the principle of "detect first, then isolate," a unified diagnostic logic framework is constructed. This framework utilizes the multi-condition-monitoring parameter-criteria correlation information provided by the expanded FMEA to generate a hierarchical set of diagnostic logic rules adapted to different operating scenarios. Attached Figure Description
[0038] Figure 1 This is an overall flowchart of a fault comprehensive diagnosis logic generation method that considers multiple operating conditions according to the present invention. Detailed Implementation
[0039] The specific implementation method of the present invention will be further described in detail below with reference to the accompanying drawings.
[0040] This invention provides a method for generating comprehensive fault diagnosis logic considering multiple operating conditions, which is specifically divided into the following parts: sorting out fault components and status monitoring parameters; dividing operating conditions and determining operating condition characterization parameters; conducting extended fault mode influence analysis; and generating comprehensive fault diagnosis logic for multiple operating conditions.
[0041] This invention adopts the principle of detection before isolation. First, the detection logic is established, then the isolation logic is established, and finally, diagnostic knowledge is constructed for application in fault diagnosis. Specifically, after completing the extended FMEA, the complex fault criteria are further decomposed into several independent test units. Each test unit consists of monitoring parameters and fault detection logic. When a test unit triggers an alarm, it indicates that the equipment has malfunctioned (S4), thus achieving fault detection. Then, the alarm information from all test units is comprehensively analyzed to construct judgment rules corresponding to various faults under different operating conditions. Combining the mapping relationship between test units and fault criteria, and matching them with the characteristic patterns of each fault mode, the specific fault that occurred is determined, i.e., the fault isolation logic. Finally, these are integrated to generate a multi-condition fault comprehensive diagnostic logic (S5).
[0042] like Figure 1 As shown, the method includes the following steps:
[0043] S1: For the device S with the fault to be tested, extract the set of fault modes of each key component from historical data and obtain the set of condition monitoring parameters;
[0044] Historical data includes: typical failure cases of objects, historical fault data, reliability information, and domain expert knowledge;
[0045] Extract the main failure modes of each key component of the device under test S to form a failure mode set; each failure mode includes failure number, failure name, failure occurrence probability level (high, medium and low), severity level (I, II and III) and failure evolution characteristics (abrupt / gradual change).
[0046] The status monitoring parameters for each key component include its number, name, type, and description.
[0047] S2: Based on the operating characteristics of the equipment S with the fault to be tested, the operating state is divided into several typical operating condition sets;
[0048] The resulting set of operating conditions is represented as follows:
[0049]
[0050] Indicates the first Each working condition;
[0051] Each operating condition includes: operating condition number, operating condition name, operating condition characterization parameters, and corresponding discrimination conditions.
[0052] Operating condition characterization parameters are parameters that characterize the operating state of equipment S.
[0053] The comparison between the characterization parameters of each working condition and their respective set thresholds, combined with logical relationships, constitutes the discrimination conditions for each working condition.
[0054] S3: For each typical operating condition, set the fault criteria for each operating condition based on the failure modes of key components and the associated status monitoring parameters under each condition.
[0055] The specific steps are as follows:
[0056] For the current typical operating conditions, identify the failure mode set of each key component under these conditions. and the set of condition monitoring parameters involved ; The number of failure modes; The number of condition monitoring parameters;
[0057] No. One fault mode The fault criterion consists of its associated / related state monitoring parameters, the fault detection thresholds for each parameter, and logical expressions, used to formally describe the fault triggering conditions; its form is:
[0058]
[0059] in, For logical comparison, For logical operators, including AND ( ),or( ),No( Operators such as ) Indicates the first Each condition monitoring parameter corresponds to a fault detection threshold; the fault detection threshold for each condition monitoring parameter is not unique. Indicates the fault mode The number of associated / related status monitoring parameters.
[0060] S4: Utilize the failure modes of key components under various operating conditions, associated condition monitoring parameters, and failure criteria to form a set of test units.
[0061] The test unit set includes test number, test name, status monitoring parameters, and fault detection logic;
[0062] The fault detection threshold associated with each condition monitoring parameter may not be unique; therefore, a condition monitoring parameter may correspond to multiple test units. Status monitoring parameters , and the corresponding fault detection threshold Perform logical comparisons separately This forms a fault detection logic; the result of each fault detection logic, combined with the test number, test name, and status monitoring parameters, forms a test unit.
[0063] The fault detection threshold is used to identify deviation characteristics of monitoring parameters for fault modes under different operating conditions, and to establish detection rules accordingly. The selection of the threshold includes two categories:
[0064] Operating condition threshold: The normal fluctuation range of the same parameter is different under different operating conditions, so different thresholds need to be set to ensure accurate detection.
[0065] Multi-level thresholds: Under the same operating conditions, different faults cause different degrees of deviation from parameters, requiring multi-level thresholds to achieve fine differentiation.
[0066] S5: Utilize the test numbers of the test unit set and combine them with fault criteria to generate comprehensive fault diagnosis logic.
[0067] Specifically:
[0068] First, based on the fault criteria and test units under each working condition, the fault detection logic is sorted out to obtain a fault isolation logic table composed of each test number;
[0069] When a test unit triggers an alarm, it indicates that the equipment has malfunctioned, thus realizing fault detection. Then, by combining the mapping relationship between the test unit and the fault criteria, the judgment rules corresponding to various faults under different operating conditions are constructed by sorting out the fault detection logic, that is, the fault isolation logic. The fault isolation logic corresponding to all test units under each operating condition is formed into a fault isolation logic table.
[0070] Secondly, the alarm information of all test units is comprehensively analyzed and matched with the characteristic patterns of each fault mode to determine the specific fault that occurred and achieve fault isolation.
[0071] Specifically, the process involves finding the corresponding fault number and fault name for each test number, integrating the fault isolation logic tables for different operating conditions, and summarizing the results of multiple operating conditions under the same fault number to form a comprehensive fault diagnosis logic.
[0072] Example:
[0073] This embodiment uses a certain type of new energy series hybrid vehicle. The power system mainly includes the following components: engine, generator, drive unit, auxiliary subsystem and power battery.
[0074] S1: For each component of the power system of a new energy series hybrid vehicle, extract the set of fault modes of each component from historical data and obtain condition monitoring parameters.
[0075] Based on typical failure cases, historical fault data, reliability information, and domain expert knowledge, the main failure modes of key components of the object system are extracted to form a complete and well-categorized set of failure modes.
[0076] Each fault mode includes a unique fault number, fault name, fault occurrence probability level (high, medium, and low), severity level (I, II, and III), and fault evolution characteristics (abrupt / gradual). A basic knowledge base is established for subsequent fault diagnosis logic, and the information is systematically organized and recorded in Table 1.
[0077] Table 1
[0078]
[0079] The specific implementation steps are as follows:
[0080] Assign a unique number to each fault (such as "F01", "F02", etc.);
[0081] Clearly define the technical name of the fault (such as "abnormal power module output", "short circuit in motor windings", etc.); in this example, the fault names include: engine fault, generator fault, cooling system fault, short circuit in drive circuit windings, drive motor controller fault, single cell fault in power battery, and overheating of power battery, etc.
[0082] By combining historical data or reliability data, assess and record the probability level of failure occurrence ("high", "medium", "low");
[0083] Based on the degree of impact of the fault on system function, the fault severity level is assigned ("I" being the most severe and "IV" being the least severe).
[0084] Based on the fault development process, it can be determined whether it is a sudden (occurring instantaneously) or a slow (gradually deteriorating) fault.
[0085] Key component monitoring parameters are extracted from design documents such as system functional design, reliability analysis, and test design. Each parameter should be clearly identified by its number, name, type, and description to provide reliable data support for the subsequent construction of fault detection criteria. The parameters are systematically organized and recorded in Table 2.
[0086] Table 2
[0087]
[0088] The specific implementation steps are as follows:
[0089] Assign a unique number to each monitoring parameter (such as "P01", "P02", etc.);
[0090] Clearly define the technical names of the monitored parameters (such as "battery voltage," "vibration frequency," etc.), ensuring that each technical name is accurate, concise, and easy to understand and apply. In this example, the parameter names include: engine speed. Engine power Generator output voltage Generator output current Coolant temperature Drive motor current drive motor winding temperature drive motor speed drive motor controller fault position single cell voltage of power battery and power battery temperature wait.
[0091] The type of monitoring parameter is determined based on its characteristics:
[0092] Continuous types (such as analog quantities like voltage, current, and temperature) can take any value within a certain range;
[0093] Discrete signals (such as switch states, status bits, and other switching or logic signals) will only take on a finite number of discrete values.
[0094] Describe the specific meaning of the monitoring parameters, explain the system state reflected by the parameters, and their possible role in fault diagnosis.
[0095] S2: Based on the operating characteristics of the device under test and the changing patterns of actual operating conditions, the system operating states are divided into several typical operating condition sets:
[0096] After completing the division of operating conditions, it is necessary to characterize each operating condition, clarify the operating condition number, operating condition name, operating condition characterization parameters and corresponding discrimination rules, so as to provide support for accurately identifying the specific operating condition of the system at present, and to systematically organize and record them in the form of Table 3.
[0097] Table 3
[0098]
[0099] The specific implementation steps are as follows:
[0100] Operating condition numbering and naming: Assign a unique number (such as C01, C02) to each operating condition and give it a concise operating condition name.
[0101] Selecting Characterization Parameters: For different operating conditions, select key parameters that can accurately reflect the differences in operating conditions, such as engine start / stop status, motor torque, battery current, vehicle speed, etc.
[0102] Establish discrimination rules: Based on parameter threshold ranges or logical relationships, establish discrimination conditions for each operating condition. For example, "engine speed > 0 and battery current > a certain threshold" can be determined as a certain operating condition.
[0103] This example selects two typical operating conditions: hybrid drive mode and pure electric drive mode.
[0104] In hybrid drive mode, the engine operates and drives the generator to charge the battery, while the generator and battery simultaneously power the drive motor.
[0105] In pure electric drive mode, the engine is off, and the vehicle is driven solely by the battery supplying power to the drive motor. The auxiliary subsystem is responsible for providing functions such as fuel, cooling, and lubrication.
[0106] The characteristic parameters for each operating condition include: unique operating condition number, operating condition name, engine status (running / stopping), and battery current.
[0107] S3: For each typical operating condition, set the fault criteria for each operating condition based on the failure modes of key components and the associated status monitoring parameters under each condition.
[0108] Based on the division of operating conditions, extended failure mode impact analysis is carried out for each typical operating condition in the form of Table 4. The analysis focuses on the performance characteristics of each failure mode on the monitoring parameters and its failure criteria under different operating conditions.
[0109] Table 4
[0110]
[0111] Based on the system's function and structure, the root causes, evolution process, and possible failure paths of faults are analyzed, and the typical impact characteristics of each fault mode on the condition monitoring parameters are identified. After obtaining the potential impact characteristics, parameters highly correlated with the fault modes are selected from the condition monitoring parameters to ensure that the parameters are sufficiently sensitive to faults and can respond promptly when a fault occurs.
[0112] For each failure mode, establish its failure criteria.
[0113] Fault criteria consist of monitoring parameters, thresholds, and logical expressions, used to formally describe the triggering conditions of a fault. Its general form is:
[0114]
[0115] For example, a certain failure mode may manifest as parameter 1 being too large and parameter 2 being too small, or parameter 3 being too large. The criterion for this can be expressed as: ;
[0116] S4: Utilize the failure modes of key components under various operating conditions, associated condition monitoring parameters, and failure criteria to form a set of test units.
[0117] Building upon the extended FMEA analysis, a systematic diagnostic logic is further generated to achieve a complete closed loop from fault logic criteria to multi-condition diagnostic logic. This process includes three stages: fault detection logic organization, fault isolation logic determination, and multi-condition logic fusion.
[0118] The core of fault detection logic lies in threshold setting and logic construction, that is, identifying the deviation characteristics of fault mode monitoring parameters under different operating conditions, and establishing detection rules accordingly. There are two main strategies for selecting thresholds: operating condition thresholds and multi-level thresholds.
[0119] After completing the analysis of fault thresholds and logic, it is necessary to further decompose the complex fault logic into independent test units, and record them in Table 5:
[0120] Table 5
[0121] Each test unit consists of monitoring parameters, their corresponding thresholds, and logical expressions. It can be triggered independently and invoked by the diagnostic system. Different thresholds often correspond to different tests; therefore, the testification process is not merely a formal transformation, but also a refinement and expansion of the detection logic.
[0122] Once a detection criterion under a certain operating condition is triggered, it indicates a potential fault. Next, the system needs to isolate the fault based on logical combinations. When the actual detected criterion combination matches the characteristic pattern of a specific fault mode, that fault mode can be identified.
[0123] In this process, the results of extended failure mode impact analysis and test units are combined to construct a failure logic combination, clarify the criterion triggering mode of each failure mode under different operating conditions, and generate Table 6 for the failure isolation logic to organize and record.
[0124] Table 6
[0125]
[0126] S5: Utilize the test numbers of the test unit set and combine them with fault criteria to generate comprehensive fault diagnosis logic.
[0127] The purpose of multi-condition diagnostic logic fusion is to effectively integrate diagnostic criteria and logic under different operating conditions, ensuring that the diagnostic system can adaptively perform fault detection and isolation under multiple operating conditions. Finally, a comprehensive diagnostic logic covering multiple operating conditions is generated and recorded in Table 7.
[0128] Table 7
[0129]
[0130] First, generate fault isolation logic for each operating condition. For example, the fault criterion for engine failure under mixed drive conditions is " >200rpm∨ >20%∨ <500V. (Based on test unit settings) >200rpm corresponds to test unit T1, >20% corresponds to test unit T2. If the voltage is less than 500V and corresponds to test unit T4, then the fault isolation logic for engine faults is “T1∨T2∨T4”.
[0131] Secondly, by integrating the fault isolation logic of various operating conditions, a comprehensive fault diagnosis logic is formed.
[0132] For example, the fault isolation logic for a cooling system fault under hybrid driving condition (C1) is "T3∨T4∨T7∨T8", and the fault isolation logic for a cooling system fault under pure electric driving condition (C2) is "T7∨T8". Therefore, the comprehensive fault diagnosis logic is (C1∧(T3∨T4∨T7∨T8)) ∨(C2∧(T7∨T8)).
[0133] The fault isolation logic for overheating of the power battery under hybrid driving condition (C1) is "T8", and the fault isolation logic for overheating of the power battery under pure electric driving condition (C2) is "T9". The comprehensive fault diagnosis logic is (C1∧T8) ∨(C2∧T9).
[0134] The main failure modes of the key components of the power system in this example are shown in Table 8.
[0135] Table 8
[0136] The monitoring parameters of the power system are shown in Table 9:
[0137] Table 9
[0138]
[0139] The powertrain system in this example is divided into two typical operating conditions: hybrid drive mode and pure electric drive mode. In hybrid drive mode, the engine operates and drives the generator to charge the battery; the generator and battery simultaneously supply power to the drive motor. In pure electric drive mode, the engine stops, and the battery supplies power to the drive motor alone to drive the vehicle. The auxiliary subsystems are responsible for providing functions such as fuel, cooling, and lubrication. The specific operating conditions and related characterization parameters and rules of the powertrain system are shown in Table 10.
[0140] Table 10
[0141]
[0142] The fault modes of the two operating conditions are analyzed in conjunction with the condition monitoring parameters, and fault judgment is made based on the set fault thresholds; the details are as follows:
[0143] For hybrid drive mode operating conditions:
[0144] The fault name is engine fault, and the corresponding status monitoring parameters include: engine speed. Engine power and generator output voltage The basis for fault judgment is: any state monitoring parameter exceeds the set fault threshold;
[0145] The fault name is generator fault, and the corresponding status monitoring parameters include: generator output voltage. and generator output current The basis for fault judgment is: any state monitoring parameter exceeds the set fault threshold;
[0146] The fault name is cooling system fault, and the corresponding status monitoring parameters include: coolant temperature. Engine power Generator output voltage and power battery temperature The basis for fault judgment is: any state monitoring parameter exceeds the set fault threshold;
[0147] The fault is named "drive circuit winding short circuit," and the corresponding status monitoring parameters include: drive motor current. and drive motor winding temperature The basis for fault judgment is: any state monitoring parameter exceeds the set fault threshold;
[0148] The fault name is drive motor controller fault, and the corresponding status monitoring parameters include drive motor speed. and drive motor controller fault position The basis for fault judgment is that all status monitoring parameters exceed their respective set fault thresholds.
[0149] The fault name is "Power Battery Cell Fault," and the corresponding status monitoring parameters include the power battery cell voltage. The basis for fault judgment is: the condition monitoring parameter exceeds one of the set fault thresholds;
[0150] The fault name is "Power Battery Overheating," and the corresponding status monitoring parameters include the power battery temperature. The basis for fault judgment is: the condition monitoring parameters exceed the set fault threshold;
[0151] For the pure electric drive mode:
[0152] The fault name is cooling system fault, and the corresponding status monitoring parameters include: coolant temperature. and power battery temperature The basis for fault judgment is: any state monitoring parameter exceeds the set fault threshold;
[0153] The fault is named "drive circuit winding short circuit," and the corresponding status monitoring parameters include: drive motor current. and drive motor winding temperature The basis for fault judgment is: any state monitoring parameter exceeds the set fault threshold;
[0154] The fault name is drive motor controller fault, and the corresponding status monitoring parameters include drive motor speed. and drive motor controller fault position The basis for fault judgment is that all status monitoring parameters exceed their respective set fault thresholds.
[0155] The fault name is "Power Battery Cell Fault," and the corresponding status monitoring parameters include the power battery cell voltage. The basis for fault judgment is: the condition monitoring parameter exceeds one of the set fault thresholds;
[0156] The fault name is "Power Battery Overheating," and the corresponding status monitoring parameters include the power battery temperature. The basis for fault judgment is: the condition monitoring parameters exceed the set fault threshold;
[0157] Extended failure mode effects (EME) analysis was performed on the power system under various operating conditions to identify typical failure modes and their characteristic manifestations, and to determine failure criteria. The analysis results are shown in Tables 11 and 12.
[0158] Table 11
[0159]
[0160] Table 12
[0161]
[0162] Based on the above analysis of the effects of extended fault modes in the power system, the fault detection logic was streamlined and transformed into a test unit. The results are shown in Table 13.
[0163] Table 13
[0164]
[0165] Fault isolation logic is formed based on fault criteria and test units under different operating conditions, as shown in Tables 14 and 15:
[0166] Table 14
[0167]
[0168] Table 15
[0169]
[0170] The fault isolation logic for the two operating conditions is synthesized to form a comprehensive diagnostic logic, as shown in Table 16:
[0171] Table 16
[0172]
[0173] For example, the fault isolation logic for a cooling system fault (F3) under hybrid drive condition (C1) is "T3∨T4∨T7∨T8", and the fault isolation logic under pure electric drive condition (C2) is "T7∨T8". By combining the operating conditions with the fault isolation logic, and connecting different operating conditions with "OR (∨)" logic, the comprehensive diagnostic logic is "(C1∧(T3∨T4∨T7∨T8)) ∨(C2∧(T7∨T8))".
Claims
1. A method for generating failure comprehensive diagnosis logic considering multiple working conditions, characterized in that, The method comprises the following steps: S1: Extracting a fault mode set of each key component from historical data for a device S to be tested for failure, and obtaining a state monitoring parameter set; S2: Dividing the operating state into a plurality of typical working condition sets according to the operating characteristics of the device S to be tested for failure; The generated working condition set is expressed as: represents the first condition; Each working condition comprises a working condition number, a working condition name, a working condition characteristic parameter, and corresponding discrimination conditions; The comparison relationship between the characteristic parameter of each working condition and the threshold value set for the working condition, combined with a logical relationship, constitutes the discrimination conditions of each working condition; S3: For each typical working condition, setting a fault criterion for each working condition according to the fault mode of the key component in the working condition and the associated state monitoring parameter; The specific steps are as follows: For the current typical working condition, confirm the failure mode set of each key component under this working condition and the related state monitoring parameter set ; is the number of failure modes; is the number of state monitoring parameters; The first fault mode of the failure criterion is of the form: wherein, is a logical comparison symbol, is a logical operator; represents the fault detection threshold value corresponding to the th state monitoring parameter; the fault detection threshold value corresponding to each state monitoring parameter is not unique; represents the number of state monitoring parameters associated with the fault mode. S4: Forming a test unit set by using the fault mode of the key component in each working condition, the associated state monitoring parameter, and the fault criterion; The first state monitoring parameter is compared with a corresponding first fault detection threshold value to form a first fault detection logic. The second state monitoring parameter is compared with a corresponding second fault detection threshold value to form a second fault detection logic. The third state monitoring parameter is compared with a corresponding third fault detection threshold value to form a third fault detection logic. The first, second and third fault detection logics are combined to form a fault detection logic. The result of the fault detection logic is combined with the test number, the test name and the state monitoring parameter to form a test unit. Therefore, one state monitoring parameter can correspond to multiple test units. The fault detection threshold value comprises two types: Working condition threshold value: The normal fluctuation range of the same parameter is different in different working conditions, so different threshold values need to be set to ensure accurate detection; Multi-level threshold value: In the same working condition, different faults have different degrees of deviation from the parameter, and multi-level threshold values are needed to achieve fine differentiation; S5: Generating a comprehensive fault diagnosis logic by using the test number of the test unit set and combining the fault criterion; Specifically: First, according to the fault criterion and the test unit in each working condition, the fault detection logic is combed to obtain a fault isolation logic table composed of each test number; When a test unit triggers an alarm, it indicates that the device has failed, achieving fault detection; then, by combing the mapping relationship between the test unit and the fault criterion, the corresponding determination rules of each type of fault in different working conditions, i.e., the fault isolation logic, are constructed; all the fault isolation logics of the test units in each working condition form a fault isolation logic table; Second, the alarm information of all test units is comprehensively analyzed and matched with the feature mode of each fault mode, so as to determine the specific fault and achieve fault isolation. Specifically: According to each test number, the corresponding fault number and fault name are found, the fault isolation logic tables of different working conditions are comprehensively combined, the results of multiple working conditions under the same fault number are comprehensively summarized, and the comprehensive fault diagnosis logic is formed.
2. The method of claim 1, wherein the method further comprises: In step S1, the historical data includes object-based typical failure cases, historical fault data, reliability data, and domain expert knowledge; The main fault modes of each key component of the device S to be tested for failure are extracted to form a fault mode set; each fault mode comprises a fault number, a fault name, a fault occurrence probability level, a severity level, and fault evolution characteristics; The state monitoring parameters of each key component include a number, a name, a type, and a description.
3. The method of claim 1, wherein the method further comprises: In step S2, the working condition characteristic parameter is a parameter representing the operating state of the device S.
4. The method of claim 1, wherein, In step S3, the fault criterion of the fault mode is composed of its associated / related state monitoring parameters, fault detection threshold values of each parameter, and logical expressions, which are used to formally describe the triggering conditions of the fault.
5. The method of claim 1, wherein, The step S4, the fault detection threshold is used to identify the deviation characteristics of the monitoring parameters of the fault mode under different working conditions, and the detection rule is established accordingly.