Sequential diagnosis test point fault location method for electromechanical system based on electrical topology model
By employing a sequential diagnostic method based on electrical topology models and utilizing graph theory and information entropy calculation, the set of faulty devices can be automatically inferred. This solves the problem of low fault location efficiency in complex electromechanical systems using traditional methods, and achieves rapid and accurate fault diagnosis and cost reduction.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ZHEJIANG UNIV
- Filing Date
- 2023-03-06
- Publication Date
- 2026-06-02
AI Technical Summary
Traditional electromechanical system fault diagnosis methods based on human experience and expert knowledge are insufficient in complex and rapidly iterating electromechanical systems, unable to quickly and accurately locate the cause of the fault, and are also costly.
A sequential diagnostic method based on electrical topology models is adopted. By establishing an electrical topology model of the electromechanical system, graph theory and information entropy calculation are used to automatically infer the set of faulty devices. The fault location is determined by measurement point recommendation and testing methods, thereby reducing the number of diagnostic steps and time.
It enables rapid and accurate fault diagnosis, reduces maintenance costs, improves fault identification efficiency, and is applicable to different electromechanical systems.
Smart Images
Figure CN116224163B_ABST
Abstract
Description
Technical Field
[0001] This invention relates to a sequential diagnostic test point fault location method, specifically a sequential diagnostic test point fault location method for electromechanical systems based on an electrical topology model. Background Technology
[0002] With the continuous upgrading of the manufacturing industry, more and more intelligent electromechanical equipment is being applied in fields such as military equipment and industrial production. As the functions of electromechanical systems become increasingly powerful, the number and complexity of internal components are rapidly increasing, posing significant challenges to the maintenance and upkeep of electromechanical equipment. Therefore, how to quickly and accurately locate the causes of failures in complex electromechanical systems and reduce their maintenance and upkeep costs has become a popular research area.
[0003] Sequential diagnostic methods for electromechanical systems are mainly divided into experience-based sequential diagnostic methods and model-based sequential diagnostic methods. Experience-based methods have a long history and are widely used in fault diagnosis. Their basic idea is to guide sequential diagnosis by recording long-term maintenance experience or the human experience of design experts, such as expert systems, fault tree generation methods, or neural network analysis methods. Their advantage is that if the accumulated fault diagnosis knowledge is comprehensive or the maintenance personnel are experienced, then this method has strong practical application value, and previously occurring or frequently occurring faults can be quickly identified. However, their disadvantages are also very obvious. For example, they require long-term fault repair data accumulation or experienced experts, resulting in extremely high time and resource costs. Furthermore, each expert knowledge base can only be applied to a specific electromechanical system; modifications to the electromechanical system or changes in the environment will require updating the entire expert knowledge base, and it cannot diagnose unpredictable or newly occurring faults.
[0004] Faced with the increasingly complex and rapidly evolving nature of electromechanical equipment, traditional methods based on human experience and expert knowledge are no longer effective. Summary of the Invention
[0005] To address the problems existing in the background technology, this invention provides a method for fault location of electromechanical system sequential diagnostic test points based on an electrical topology model. This model can automatically infer a set of potentially faulty devices based on the input measurement values of the test points in sequential diagnostic testing. Then, by integrating the fault probabilities of each device, it calculates the information entropy of each preliminary test point. The effectiveness of the test points is evaluated based on the weighted calculation results of the information entropy of each device and the corresponding test cost, and the test points with the highest effectiveness are recommended. Measurements are then performed based on the recommended test points, and the measurement results are input into the model, repeating the previous steps until the faulty device or fault fuzzy group is finally identified. This method features fast computation speed, reasonable recommended test points, fewer fault diagnosis steps, and high fault identification efficiency, which can greatly improve the efficiency of sequential diagnostic testing personnel in diagnosing faults in electromechanical systems.
[0006] The sequential diagnostic method based on testable models does not rely on a large amount of historical data or prior knowledge. Instead, it models testable information such as device functions, system coupling relationships, and system failure modes, and then uses the established testable model to perform fault reasoning. The model not only ensures rapid and accurate fault diagnosis, but also allows for easy reuse in different electromechanical systems.
[0007] The technical solution adopted in this invention is:
[0008] The electromechanical system sequential diagnostic test point fault location method of the present invention includes the following steps:
[0009] Step S1: Use graph theory to establish the electrical topology model of the electromechanical system under test. Each device in the electromechanical system under test includes energy source devices, intermediate devices, and actuators. Each actuator includes external actuators and internal actuators. Perform sequential diagnosis on each device in the electromechanical system under test. Input the preset number of each faulty external actuator and the fault mode into the electrical topology model. The electrical topology model outputs the initial fault fuzzy group.
[0010] Step S2: If the initial fault fuzzy group in step S1 contains only one device or several devices that cannot be classified, the initial fault fuzzy group is directly used as the diagnostic result. That is, one or all devices in the initial fault fuzzy group are located as faulty devices, and the diagnosis ends. If the initial fault fuzzy group contains several devices that can be classified, a recommended measurement point is obtained by applying the measurement point recommendation method to the initial fault fuzzy group. The actual measurement result of the recommended measurement point is obtained by applying the measurement point testing method to the recommended measurement point and input into the electrical topology model. The electrical topology model outputs the current fault fuzzy group. The same operation as the initial fault fuzzy group in step S2 is repeated for the current fault fuzzy group until the current output fault fuzzy group contains only one device or several devices that cannot be classified. The current fault fuzzy group is then used as the diagnostic result, and the diagnosis ends, realizing sequential diagnostic measurement point fault location.
[0011] A fault fuzzy group is divisible if some of the devices in the fault fuzzy group can be detected by the measurement point test method, while a fault fuzzy group is indivisible if some of the devices in the fault fuzzy group cannot be detected, such as devices that cannot be detected within an electromechanical system.
[0012] In step S1, the electrical topology model mainly consists of several system energy transfer links and device description information for each component. The energy transfer links describe the connection relationships between various components in the system, as well as the logical sequence of energy propagation from the energy source to the actuator. The specific details of each system energy transfer link in the electrical topology model are as follows:
[0013]
[0014] Among them, ETL1, ETL2...ETL M These represent the 1st, 2nd...Mth system energy transfer links in the electrical topology model, respectively. These represent the connection relationships of the first, second, ..., Nth devices in the energy propagation process of the first system energy transfer link in the electrical topology model; These represent the connection relationships of the 1st, 2nd...Nth devices in the energy propagation process of the second system energy transfer link in the electrical topology model; These represent the connections of the 1st, 2nd...Nth devices in the Mth system energy transfer link of the electrical topology model during energy propagation; sou1, sou2...sou M These represent the energy source devices in the first, second...Mth system energy transfer links of the electrical topology model; These represent the 1st, 2nd...nth intermediate devices in the first system energy transfer link of the electrical topology model; These represent the 1st, 2nd...nth intermediate devices in the second system energy transfer link of the electrical topology model, respectively. These represent the 1st, 2nd...nth intermediate devices in the Mth system energy transfer link of the electrical topology model; des1, des2...des M These represent the actuators in the first, second...Mth system energy transfer links of the electrical topology model.
[0015] For each system energy transfer link in the electrical topology model, the energy source device in the system energy transfer link transmits energy to the actuator through various intermediate devices.
[0016] The device description information of each device in the electrical topology model includes the input-output correspondence, function and operating mode, fault mode and preset fault probability of each device.
[0017] The input-output relationship of each device is as follows: the output voltage and input voltage are proportionally amplified; the function and operating mode of each device are as follows: the output power of the device is different when the operating range is different; the fault modes of each device are as follows: the circuit breaker may have different fault modes such as "stuck in short circuit state", "stuck in open circuit state", "internal fuse", "poor contact"; the device description information can be used to determine whether the device is faulty and to deduce the input and output when the device is working normally.
[0018] In step S1, each actuator includes external actuators and internal actuators. External actuators are visible actuators located outside the electromechanical system under test, such as lights, motors, mechanical actuators, hydraulic rods, etc. Internal actuators are invisible actuators located inside the electromechanical system under test, such as relays or coils installed inside the cabinet. In the energy transfer circuit model of its input circuit, the relay is the actuator.
[0019] The fault state of an external actuator is similar to that of an open circuit in a light bulb. If the light bulb is turned on by the power source device switch but does not light up, it can be preliminarily determined that the external actuator, the light bulb, is in an open circuit state.
[0020] In step S2, the specific fault fuzzy group output by the electrical topology model for the current time is as follows:
[0021] FaultGroup j = <COMPS j ,MODE j PROB j >
[0022]
[0023]
[0024]
[0025] Among them, FaultGroup j FaultGroup0 represents the fault fuzzy group output by the electrical topology model at the j-th iteration, where j ≥ 0. When j = 0, FaultGroup0 represents the initial fault fuzzy group. j MODE represents the set of faulty devices in the fault fuzzy group of the j-th output of the electrical topology model; j PROB represents the set of fault modes for each faulty device in the fault fuzzy group output by the j-th iteration of the electrical topology model. j This represents the normalized fault probability of each faulty device in its respective fault mode in the fault fuzzy group output by the j-th time of the electrical topology model. These represent the 1st, 2nd...Ath faulty devices in the fault fuzzy group output by the j-th time of the electrical topology model; These represent the 1st, 2nd...Bth fault modes in the fault fuzzy group output by the j-th time of the electrical topology model; Let represent the 1st, 2nd...Bth normalized fault probabilities in the fault fuzzy group output by the j-th time of the electrical topology model.
[0026] The electrical topology model determines whether a measurement point is functioning correctly by comparing the actual measurement value with the inferred measurement value. This allows the model to infer the set of possible faulty devices and their corresponding probability sets, and to determine the next possible measurement point.
[0027] In step S2, each faulty device in the fault fuzzy group includes one or more measurement points, such as the two endpoints of the connecting line of the intermediate device serving as two measurement points of the connecting line; after using the measurement point recommendation method, a recommended measurement point is obtained, as follows:
[0028] S2.1 For each measurement point of each faulty device in each system energy transfer link in the current fault fuzzy group output by the electrical topology model, calculate the weighted comprehensive information entropy of the measurement point based on its fault probability and its possible test results for the segmentation of the fault fuzzy group.
[0029] S2.2 For each test point, the validity value of the test point is obtained by weighting the weighted comprehensive information entropy and the test cost of the test point.
[0030] S2.3 Sort the validity values of each measurement point and recommend the measurement point with the smallest validity value as the recommended measurement point for this time.
[0031] After establishing the electrical topology model, faulty devices can be inferred based on the measured values of the measuring points. For example, if a test point is selected in the middle of the energy transfer link, if the actual measured value of the measuring point matches the inferred value in the model, it proves that the device between the energy source and the measuring point is working normally, and the faulty device is in another part of the link; if they do not match, it indicates that the faulty device is in the device between the energy source and the measuring point.
[0032] In step S2.1, for each measurement point of each faulty device in each system energy transfer link in the current fault fuzzy group output by the electrical topology model, the weighted comprehensive information entropy of the measurement point is calculated, as follows:
[0033] For each system energy transfer link in the electrical topology model, a set U is formed by s faulty devices between the self-test point and the energy source device, and a set D is formed by r faulty devices between the self-test point and the actuator. First, the information entropy of sets U and D is calculated respectively, as follows:
[0034]
[0035]
[0036] Where InformU and InformD represent the information entropy of set U and set D, respectively; P Ui and P Di Let U and D represent the normalized failure probabilities of the i-th faulty device in sets U and D, respectively.
[0037] Then calculate the probability P that the measurement result at the measuring point is normal. pass The probability P of abnormal measurement results fail The details are as follows:
[0038]
[0039]
[0040] The probability of a normal measurement result at a test point considers not only the probability that the device at that test point is functioning correctly, but also the probability that all s faulty devices along the link from the test point to the energy source device are functioning correctly. Similarly, the probability of an abnormal measurement result requires considering the probability that all s faulty devices along the link from the test point to the actuator are functioning correctly. A correct test result indicates that the connection between the test point and the energy source is normal, and the fault lies between the test point and the actuator; an incorrect test result indicates that the fault lies between the test point and the energy source, but not between the test point and the actuator.
[0041] Finally, based on the information entropy of set U and set D, and the probability P that the measurement results at the measuring point are normal, passThe probability P of abnormal measurement results fail The weighted comprehensive information entropy (Inform) of the measurement points is calculated as follows:
[0042] Inform = P pass *InformD+P fail *InformU
[0043] Inform represents the weighted comprehensive information entropy of the measurement points.
[0044] The smaller the weighted comprehensive information entropy (Inform), the smaller the average uncertainty of the system after measurement at that point, and the higher the effectiveness of the measurement point.
[0045] 8. The method for locating faults at measurement points in a sequential diagnostic process for electromechanical systems based on an electrical topology model, as described in claim 6, is characterized in that:
[0046] In step S2.2, for each test point, the weighted comprehensive information entropy and test cost of the test point are weighted and averaged to obtain the validity value of the test point, as follows:
[0047] Rank = a * Inform + b * Co
[0048] Where Rank represents the validity value of the measurement point; Inform represents the weighted comprehensive information entropy of the measurement point; Co represents the test cost of the measurement point; and a and b represent the weighted average coefficients of the weighted comprehensive information entropy and test cost of the measurement point, respectively.
[0049] The testing cost described in this invention can typically be measured in terms of measurement time. The testing cost is related to the difficulty of the test; the greater the difficulty, the greater the testing cost.
[0050] Each faulty device in each system energy transfer link has its measurement points forming a measurement point set. Each time, a measurement point is randomly selected from the set for calculation. After each measurement point is calculated, its validity value is recorded, and the point is removed from the set. The calculation ends when the measurement point set is empty. In the device information of the electrical topology model, different measurement points for each device have corresponding test cost Co data. This data range can be normalized to 0-1, with values ranging from very low test cost (easily measurable) to extremely high test cost (almost impossible to measure).
[0051] 9. The method for locating faults at measurement points in a sequential diagnostic process for electromechanical systems based on an electrical topology model, as described in claim 1, is characterized in that:
[0052] In step S2, the electrical testing method specifically involves using one or more of the following methods—electrical testing, hydraulic testing, and mechanical testing—to test the measuring point and obtain the actual measurement result of the measuring point.
[0053] The point-based testing method involves conducting tests based on recommended test points. For example, when a test point transmits electrical signals, electrical testing methods are used to measure its voltage, current, short circuit, and open circuit properties. When a test point transmits hydraulic signals, hydraulic testing methods are used to measure its pressure and flow rate. When a test point transmits mechanical signals, mechanical testing methods are used to measure whether its movement is in place and whether it is jammed. The actual test results are then input into the model. For test points of composite devices, several methods from electrical, hydraulic, and mechanical testing are used.
[0054] The beneficial effects of this invention are:
[0055] 1. Intelligent fault diagnosis reasoning based on electrical topology model greatly reduces the learning and usage costs for personnel maintaining the system.
[0056] 2. By adopting the measurement point evaluation method based on comprehensive information entropy, theoretically each test step can be a locally optimal solution, which greatly reduces the number of tests and test time required to locate faults.
[0057] 3. The measurement point is evaluated by a weighted average of comprehensive information entropy and test cost. This ensures that the measurement point has a good theoretical ability to reduce the fault fuzzy set and that the measurement point is actually measurable.
[0058] In summary, the method of the present invention has the advantages of fast calculation speed, reasonable recommended test points, fewer fault diagnosis steps, and high fault identification efficiency, which can greatly improve the efficiency of sequential diagnostic testing for fault diagnosis of electromechanical systems. Attached Figure Description
[0059] Figure 1 This is a schematic diagram of the measurement point recommendation and fault diagnosis process of the present invention;
[0060] Figure 2 This is a schematic diagram of the calculation process for the entropy of the measurement point information and the effective value of the present invention;
[0061] Figure 3 This is a schematic diagram of an embodiment of the lamp circuit of the present invention and its electrical topology model. Detailed Implementation
[0062] To make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be further described in detail with reference to the accompanying drawings. This description illustrates specific embodiments consistent with the principles of the present invention by way of example rather than limitation. These embodiments are described in sufficient detail to enable those skilled in the art to practice the invention, use other embodiments, and modify and / or substitute the structure of various elements without departing from the scope and spirit of the invention. Therefore, the following detailed description should not be construed as limiting.
[0063] like Figure 1 As shown, the electromechanical system sequential diagnostic test point fault location method of the present invention includes the following steps:
[0064] Step S1: Use graph theory to establish the electrical topology model of the electromechanical system under test. Each device in the electromechanical system under test includes energy source devices, intermediate devices, and actuators. Each actuator includes external actuators and internal actuators. Perform sequential diagnosis on each device in the electromechanical system under test. Input the preset number of each faulty external actuator and the fault mode into the electrical topology model. The electrical topology model outputs the initial fault fuzzy group.
[0065] In step S1, the electrical topology model mainly consists of several system energy transfer links and device description information for each component. The energy transfer links describe the connection relationships between various components in the system, as well as the logical sequence of energy propagation from the energy source to the actuator. The specific system energy transfer links in the electrical topology model are as follows:
[0066]
[0067] Among them, ETL1, ETL2...ETL M These represent the 1st, 2nd...Mth system energy transfer links in the electrical topology model, respectively. These represent the connection relationships of the first, second, ..., Nth devices in the energy propagation process of the first system energy transfer link in the electrical topology model; These represent the connection relationships of the 1st, 2nd...Nth devices in the energy propagation process of the second system energy transfer link in the electrical topology model; These represent the connections of the 1st, 2nd...Nth devices in the Mth system energy transfer link of the electrical topology model during energy propagation; sou1, sou2...sou M These represent the energy source devices in the first, second...Mth system energy transfer links of the electrical topology model; These represent the 1st, 2nd...nth intermediate devices in the first system energy transfer link of the electrical topology model; These represent the 1st, 2nd...nth intermediate devices in the second system energy transfer link of the electrical topology model, respectively. These represent the 1st, 2nd...nth intermediate devices in the Mth system energy transfer link of the electrical topology model; des1, des2...des M These represent the actuators in the first, second...Mth system energy transfer links of the electrical topology model.
[0068] For each system energy transfer link in the electrical topology model, the energy source device in the system energy transfer link transmits energy to the actuator through various intermediate devices.
[0069] The device description information of each component in the electrical topology model includes the input-output correspondence, function and operating mode, fault mode and preset fault probability of each component.
[0070] The input-output relationship of each device is as follows: the output voltage and input voltage are proportionally amplified; the function and operating mode of each device are as follows: the output power of the device is different when the operating range is different; the fault modes of each device are as follows: the circuit breaker may have different fault modes such as "stuck in short circuit state", "stuck in open circuit state", "internal fuse", "poor contact"; the device description information can be used to determine whether the device is faulty and to deduce the input and output when the device is working normally.
[0071] In step S1, each actuator includes external actuators and internal actuators. External actuators are visible actuators located outside the electromechanical system under test, such as lights, motors, mechanical motion devices, hydraulic rods, etc. Internal actuators are invisible actuators located inside the electromechanical system under test, such as relays or coils installed inside the cabinet. In the energy transfer circuit model of its input circuit, the relay is the actuator.
[0072] The fault state of an external actuator is similar to that of an open circuit in a light bulb. If the light bulb is turned on by the power source device switch but does not light up, it can be preliminarily determined that the external actuator, the light bulb, is in an open circuit state.
[0073] Step S2: If the initial fault fuzzy group in step S1 contains only one device or several devices that cannot be classified, the initial fault fuzzy group is directly used as the diagnostic result. That is, one or all devices in the initial fault fuzzy group are located as faulty devices, and the diagnosis ends. If the initial fault fuzzy group contains several devices that can be classified, a recommended measurement point is obtained by applying the measurement point recommendation method to the initial fault fuzzy group. The actual measurement result of the recommended measurement point is obtained by applying the measurement point testing method to the recommended measurement point and input into the electrical topology model. The electrical topology model outputs the current fault fuzzy group. The same operation as the initial fault fuzzy group in step S2 is repeated for the current fault fuzzy group until the current output fault fuzzy group contains only one device or several devices that cannot be classified. The current fault fuzzy group is then used as the diagnostic result, and the diagnosis ends, realizing sequential diagnostic measurement point fault location.
[0074] A fault fuzzy group is divisible if some of the devices in the fault fuzzy group can be detected by the measurement point test method, while a fault fuzzy group is indivisible if some of the devices in the fault fuzzy group cannot be detected, such as devices that cannot be detected within an electromechanical system.
[0075] In step S2, the specific fault fuzzy group output by the electrical topology model for the current time is as follows:
[0076] FaultGroup j = <COMPS j ,MODE j PROB j >
[0077]
[0078]
[0079]
[0080] Among them, FaultGroup j FaultGroup0 represents the fault fuzzy group output by the electrical topology model at the j-th iteration, where j ≥ 0. When j = 0, FaultGroup0 represents the initial fault fuzzy group. j MODE represents the set of faulty devices in the fault fuzzy group of the j-th output of the electrical topology model; j PROB represents the set of fault modes for each faulty device in the fault fuzzy group output by the j-th iteration of the electrical topology model. j This represents the normalized fault probability of each faulty device in its respective fault mode in the fault fuzzy group output by the j-th time of the electrical topology model. These represent the 1st, 2nd...Ath faulty devices in the fault fuzzy group output by the j-th time of the electrical topology model; These represent the 1st, 2nd...Bth fault modes in the fault fuzzy group output by the j-th time of the electrical topology model; Let represent the 1st, 2nd...Bth normalized fault probabilities in the fault fuzzy group output by the j-th time of the electrical topology model.
[0081] The electrical topology model determines whether a measurement point is functioning correctly by comparing the actual measurement value with the inferred measurement value. This allows the model to infer the set of possible faulty devices and their corresponding probability sets, and to determine the next possible measurement point.
[0082] In step S2, each faulty device in the fault fuzzy group includes one or more measurement points, such as the two endpoints of the connecting line of the intermediate device serving as two measurement points of the connecting line; after using the measurement point recommendation method, a recommended measurement point is obtained, such as... Figure 2 As shown, the details are as follows:
[0083] S2.1 For each measurement point of each faulty device in each system energy transfer link in the current fault fuzzy group output by the electrical topology model, calculate the weighted comprehensive information entropy of the measurement point based on its fault probability and its possible test results for the segmentation of the fault fuzzy group.
[0084] S2.2 For each test point, the validity value of the test point is obtained by weighting the weighted comprehensive information entropy and the test cost of the test point.
[0085] S2.3 Sort the validity values of each measurement point and recommend the measurement point with the smallest validity value as the recommended measurement point for this time.
[0086] After establishing the electrical topology model, faulty devices can be inferred based on the measured values of the measuring points. For example, if a test point is selected in the middle of the energy transfer link, if the actual measured value of the measuring point matches the inferred value in the model, it proves that the device between the energy source and the measuring point is working normally, and the faulty device is in another part of the link; if they do not match, it indicates that the faulty device is in the device between the energy source and the measuring point.
[0087] In step S2.1, for each measurement point of each faulty device in each system energy transfer link in the current fault fuzzy group output by the electrical topology model, the weighted comprehensive information entropy of the measurement point is calculated, as follows:
[0088] For each system energy transfer link in the electrical topology model, a set U is formed by s faulty devices between the self-test point and the energy source device, and a set D is formed by r faulty devices between the self-test point and the actuator. First, the information entropy of sets U and D is calculated respectively, as follows:
[0089]
[0090]
[0091] Where InformU and InformD represent the information entropy of set U and set D, respectively; P Ui and P Di Let U and D represent the normalized failure probabilities of the i-th faulty device in sets U and D, respectively.
[0092] Then calculate the probability P that the measurement result at the measuring point is normal. pass The probability P of abnormal measurement results fail The details are as follows:
[0093]
[0094]
[0095] The probability of a normal measurement result at a test point considers not only the probability that the device at that test point is functioning correctly, but also the probability that all s faulty devices along the link from the test point to the energy source device are functioning correctly. Similarly, the probability of an abnormal measurement result requires considering the probability that all s faulty devices along the link from the test point to the actuator are functioning correctly. A correct test result indicates that the connection between the test point and the energy source is normal, and the fault lies between the test point and the actuator; an incorrect test result indicates that the fault lies between the test point and the energy source, but not between the test point and the actuator.
[0096] Finally, based on the information entropy of set U and set D, and the probability P that the measurement results at the measuring point are normal, pass The probability P of abnormal measurement results fail The weighted comprehensive information entropy (Inform) of the measurement points is calculated as follows:
[0097] Inform = P pass *InformD+P fail *InformU
[0098] Inform represents the weighted comprehensive information entropy of the measurement points.
[0099] The smaller the weighted comprehensive information entropy (Inform), the smaller the average uncertainty of the system after measurement at that point, and the higher the effectiveness of the measurement point.
[0100] In step S2.2, for each test point, the weighted comprehensive information entropy and test cost of the test point are weighted and averaged to obtain the validity value of the test point, as follows:
[0101] Rank = a * Inform + b * Co
[0102] Where Rank represents the validity value of the measurement point; Inform represents the weighted comprehensive information entropy of the measurement point; Co represents the test cost of the measurement point; and a and b represent the weighted average coefficients of the weighted comprehensive information entropy and test cost of the measurement point, respectively.
[0103] The testing cost described in this invention can typically be measured in terms of measurement time. The testing cost is related to the difficulty of the test; the greater the difficulty, the greater the testing cost.
[0104] Each faulty device in each system energy transfer link has its measurement points forming a measurement point set. Each time, a measurement point is randomly selected from the set for calculation. After each measurement point is calculated, its validity value is recorded, and the point is removed from the set. The calculation ends when the measurement point set is empty. In the device information of the electrical topology model, different measurement points for each device have corresponding test cost Co data. This data range can be normalized to 0-1, with values ranging from very low test cost (easily measurable) to extremely high test cost (almost impossible to measure).
[0105] In step S2, the electrical testing method specifically involves using one or more of the following methods—electrical testing, hydraulic testing, and mechanical testing—to test the measuring point and obtain the actual measurement result of the measuring point.
[0106] The point-based testing method involves conducting tests based on recommended test points. For example, when a test point transmits electrical signals, electrical testing methods are used to measure its voltage, current, short circuit, and open circuit properties. When a test point transmits hydraulic signals, hydraulic testing methods are used to measure its pressure and flow rate. When a test point transmits mechanical signals, mechanical testing methods are used to measure whether its movement is in place and whether it is jammed. The actual test results are then input into the model. For test points of composite devices, several methods from electrical, hydraulic, and mechanical testing are used.
[0107] Specific embodiments of the present invention are as follows:
[0108] like Figure 3 As shown, this is a simple circuit that uses a switch to control the brightness of a light bulb. The schematic diagram of this simple circuit model is as follows. Figure 3As shown on the left, its main components include a DC power supply SP, two fuses FU, an air switch QF, a lamp HL01, and six wires L connecting the components. Its normal function is that when the power supply SP is powered, if the air switch QF is closed, the lamp HL01 will light up; if the air switch QF is open, the lamp HL01 will turn off. Figure 3 The right side of the diagram shows the electrical topology model of this simple circuit. The energy source device is power supply SP01, the actuator is lamp HL01, and the remaining devices are intermediate devices. These devices form an energy transfer link according to their energy propagation logic sequence. The corresponding electrical topology model is used to illustrate the measurement point recommendation process. Assuming the circuit fault is an open circuit between ports 3 and 4 of the air switch QF1, the lamp will not light up after the switch is closed, indicating a fault in the circuit. The resulting fault fuzzy group, FaultGroup, is shown in Table 1.
[0109] Table 1
[0110]
[0111]
[0112] The device failure probabilities after normalization are shown in Table 2:
[0113] Table 2
[0114]
[0115] The test point set obtained from the fault fuzzy group also includes the corresponding test costs, as shown in Table 3.
[0116] Table 1
[0117]
[0118]
[0119] Next, the information entropy of the measurement points of all devices in the fault fuzzy group is calculated. Taking the CurrentIn measurement point of fuse FU02 as an example, refer to... Figure 3 In the electrical topology model, if the measurement at the measuring point is normal, the fault range is narrowed down to the range from fuse FU02 to lamp HL01; if the measurement at the measuring point is abnormal, the fault range is narrowed down to the range from power supply SP01 to L02.
[0120] Therefore, the information entropy calculation for this measurement point is as follows:
[0121]
[0122]
[0123]
[0124]
[0125] Inform = P pass *InformD+P fail *InformU=1.8263
[0126] When calculating the cost in conjunction with the test cost, the weighted average coefficients are a = 0.8 and b = 0.2.
[0127] Rank=a*Inform+b*Cost=1.5031
[0128] The calculation process for the comprehensive information entropy of the currentIn value of the fuse FU02 measurement point was applied to all measurement points in the measurement point set, and the final validity value Rank results are shown in Table 4.
[0129] Table 2
[0130]
[0131]
[0132] Based on the standard of minimizing the overall information entropy, the CurrentOut measurement point of FU02 is recommended. After that, the tester inputs the measurement value of this measurement point to obtain a new fault fuzzy group, and continues the previous steps until the fault fuzzy group can no longer be divided, thus locating the faulty device.
[0133] Ultimately, the testers determined the fault location to be between ports 3 and 4 of QF01 by measuring the CurrentOut point of FU02, the CurrentIn point of L05, and QF01.3 and QF01.4 in a total of four steps. Only 4 of the 22 points needed to be measured, which greatly simplified the fault diagnosis and location process.
Claims
1. A method for locating faults at measurement points in a sequential diagnostic process for electromechanical systems based on an electrical topology model, characterized in that: The method includes the following steps: Step S1: Use graph theory to establish the electrical topology model of the electromechanical system under test. Each device in the electromechanical system under test includes energy source devices, intermediate devices and actuators. Each actuator includes external actuators and internal actuators. Perform sequential diagnosis on each device in the electromechanical system under test. Input the preset number of each faulty external actuator and the fault mode into the electrical topology model. The electrical topology model outputs the initial fault fuzzy group. Step S2: If there is only one device or several devices in the initial fault fuzzy group in step S1 but they cannot be divided, the initial fault fuzzy group is directly used as the diagnosis result, that is, one or all the devices in the initial fault fuzzy group are used as faulty devices for location, and the diagnosis ends. If there are several devices in the initial fault fuzzy group and they can be divided, then a recommended measurement point is obtained by using the measurement point recommendation method on the initial fault fuzzy group. The actual measurement result of the recommended measurement point is obtained by using the measurement point testing method and input into the electrical topology model. The electrical topology model outputs the current fault fuzzy group. The same operation as the initial fault fuzzy group in step S2 is repeated for the current fault fuzzy group until there is only one device in the current output fault fuzzy group or there are several devices but they cannot be divided. The current fault fuzzy group is used as the diagnosis result, the diagnosis ends, and the sequential diagnosis measurement point fault location is realized. Step S2 is as follows: S2.1 For each measurement point of each faulty device in each system energy transfer link in the current fault fuzzy group output by the electrical topology model, calculate the weighted comprehensive information entropy of the measurement point; S2.2 For each test point, the validity value of the test point is obtained by weighting the weighted comprehensive information entropy and the test cost of the test point; S2.3 Sort the validity values of each measurement point and recommend the measurement point with the smallest validity value as the recommended measurement point for this time; Step S2.1 is as follows: For each system energy transfer link in the electrical topology model, a set U is formed by s faulty devices between the self-test point and the energy source device, and a set D is formed by r faulty devices between the self-test point and the actuator. The information entropy of sets U and D is calculated as follows: in, and Let U and D represent the information entropy of set U and set D, respectively. and Let U and D represent the normalized failure probabilities of the i-th faulty device in set U and set D, respectively. Then calculate the probability that the measurement results at the measuring point are normal. and the probability of abnormal measurement results as follows: Finally, based on the information entropy of set U and set D, and the probability that the measurement results of the measuring point are normal... and the probability of abnormal measurement results The weighted comprehensive information entropy of the measurement points is calculated as follows: In step S2.2, the validity values of the measuring points are as follows: in, Indicates the validity value of the measuring point; Indicates the test cost at the measurement point; and These represent the weighted comprehensive information entropy of the measurement points and the weighted average coefficient of the test cost, respectively.
2. The method for locating faults at measurement points in a sequential diagnostic test of an electromechanical system based on an electrical topology model, as described in claim 1, is characterized in that: In step S1, the electrical topology model mainly consists of several system energy transfer links and device description information for each device. The specific system energy transfer links in the electrical topology model are as follows: ⋮ in, , ... These represent the first, second, ..., Mth system energy transfer links in the electrical topology model, respectively. These represent the connection relationships of the first, second, ..., Nth devices in the energy propagation process of the first system energy transfer link in the electrical topology model; These represent the connection relationships of the 1st, 2nd...Nth devices in the energy propagation process of the second system energy transfer link in the electrical topology model; These represent the connection relationships of the 1st, 2nd...Nth devices in the energy propagation process of the Mth system energy transfer link in the electrical topology model; , ... These represent the energy source devices in the first, second, ..., Mth system energy transfer links of the electrical topology model, respectively. These represent the 1st, 2nd...nth intermediate devices in the 1st system energy transfer link of the electrical topology model, respectively. These represent the 1st, 2nd...nth intermediate devices in the second system energy transfer link of the electrical topology model, respectively. These represent the 1st, 2nd...nth intermediate devices in the Mth system energy transfer link of the electrical topology model, respectively. , These represent the actuators in the first, second, ..., Mth system energy transfer links of the electrical topology model, respectively. For each system energy transfer link in the electrical topology model, the energy source device in the system energy transfer link transmits energy to the actuator through various intermediate devices.
3. The method for locating faults at measurement points in a sequential diagnostic process for electromechanical systems based on an electrical topology model, as described in claim 2, is characterized in that: The device description information of each device in the electrical topology model includes the input-output correspondence, function and operating mode, fault mode and preset fault probability of each device.
4. The method for locating faults at measurement points in a sequential diagnostic process for electromechanical systems based on an electrical topology model, as described in claim 1, is characterized in that: In step S1, each actuator includes an external actuator and an internal actuator. The external actuator is a visible actuator located outside the electromechanical system under test, and the internal actuator is an invisible actuator located inside the electromechanical system under test.
5. The method for locating faults at measurement points in a sequential diagnostic process for electromechanical systems based on an electrical topology model, as described in claim 1, is characterized in that: In step S2, the specific fault fuzzy group output by the electrical topology model for the current time is as follows: in, This represents the fault fuzzy group output by the j-th iteration of the electrical topology model, where j ≥ 0. When j = 0, Indicates the initial fault fuzzy group; This represents the set of faulty devices in the fault fuzzy group output by the j-th iteration of the electrical topology model. This represents the set of fault modes for each faulty device in the fault fuzzy group output by the j-th iteration of the electrical topology model. This represents the normalized fault probability of each faulty device in its respective fault mode in the fault fuzzy group output by the j-th time of the electrical topology model. These represent the 1st, 2nd...Ath faulty devices in the fault fuzzy group output by the j-th time of the electrical topology model; These represent the 1st, 2nd...Bth fault modes in the fault fuzzy group output by the j-th time of the electrical topology model; Let represent the 1st, 2nd...Bth normalized fault probabilities in the fault fuzzy group output by the j-th time of the electrical topology model.
6. The method for locating faults at measurement points in a sequential diagnostic process for electromechanical systems based on an electrical topology model, as described in claim 1, is characterized in that: In step S2, the electrical testing method specifically involves using one or more of the following methods—electrical testing, hydraulic testing, and mechanical testing—to test the measuring point and obtain the actual measurement result of the measuring point.