Hydroelectric generating set runner plate fault judgment method based on analytic hierarchy process and fuzzy membership degree

Abstract

The invention relates to a hydroelectric generating set mirror plate fault judgment method based on analytic hierarchy process and fuzzy membership, which comprises the following steps: selecting symptom indexes reflecting faults, and systematically analyzing and constructing a hierarchical structure model in a fault type-symptom index association form by using the analytic hierarchy process; a triangular-semi-trapezoidal membership function is used as a basis, a membership function is constructed according to a fault condition level, a symptom index membership matrix is calculated according to a deviation degree between a symptom index real-time value and a health reference value, a fault type fuzzy judgment matrix is further obtained, and a fault type fuzzy judgment matrix is obtained according to a maximum membership principle. And obtaining a diagnosis result whether the hydroelectric generating set has a mirror plate levelness standard exceeding fault. According to the invention, through systematic analysis, objective weight distribution, dynamic reference comparison and a fuzzy determination method, the accuracy of fault determination is significantly improved.

Description

Technical Field

[0001] The invention belongs to the field of hydropower unit fault diagnosis, and in particular relates to a method for determining a hydropower unit mirror plate fault based on hierarchical analysis and fuzzy membership. Background Art

[0002] With the adjustment of my country's energy structure, hydropower has become an indispensable way of power generation. As the key equipment for hydropower energy conversion, the safety and stability of hydropower units have always been a hot topic in the power industry. As the installed capacity of power stations increases, the unit structure becomes more and more complex. As a key component of hydropower units, thrust bearings bear the weight of the rotor of the generator unit and the axial water thrust. The health status of thrust bearings is an important factor affecting the safe and stable operation of power stations.

[0003] As the core component of the thrust bearing system, the mirror plate plays a vital role in the stable operation and high-efficiency energy conversion process of the unit. The main function of the mirror plate is to effectively transmit the axial thrust generated by the turbine to the generator, while maintaining the precise alignment of the turbine and generator axes to ensure synchronous rotation between the two. This process is crucial to achieving efficient conversion of water energy into electrical energy, because any imbalance in axial force may cause unit vibration, affect power generation efficiency, and even cause equipment damage.

[0004] The design and material selection of the mirror plate must be able to withstand huge axial loads, while having good wear resistance and corrosion resistance to adapt to the complex working environment of the hydropower station. Mirror plate surface treatment technologies such as hardening and lubricating coatings have a significant impact on reducing friction coefficients, reducing energy losses and extending equipment life. In addition, the rapid response capability of the mirror plate in emergency shutdown situations is also indispensable to ensure the safe operation of the hydropower station and the safety of the lives of maintenance personnel. The mirror plate not only bears the basic functions of transmitting thrust and maintaining axial positioning in the hydropower unit, but is also the key to ensuring the efficient, stable and safe operation of the entire power generation system.

[0005] Therefore, in order to ensure the safe and stable operation of the unit, improve equipment utilization, and avoid major production safety accidents and casualties, it is necessary to study the fault determination method for the mirror plate of the hydropower unit. Summary of the invention

[0006] The purpose of the present invention is to address the above-mentioned problems and provide a method for determining the fault of a mirror plate of a hydropower unit based on hierarchical analysis and fuzzy membership. By combining the emerging hierarchical analysis algorithm and fuzzy theory in the field of state evaluation, a fault determination model for the excessive horizontality of the mirror plate of a hydropower unit is constructed, and a more effective new method in the field of fault diagnosis of hydropower units is found.

[0007] In order to achieve the above object, the technical solution provided by the present invention is: The method for determining the fault of the mirror plate of a hydropower unit based on hierarchical analysis and fuzzy membership includes the following steps: Step 1: Analyze the fault mechanism and determine the symptom indicators used to characterize the fault of excessive horizontality of the mirror plate of the hydropower unit; Step 2: Construct a hierarchical structure model for the fault of excessive horizontality of the mirror plate of the hydropower unit; Step 3: Based on the hierarchical structure model, a weight matrix of the symptom indicators is calculated; Step 4: Obtain the health benchmark value of each symptom indicator under different working conditions, and calculate the deviation between the real-time value of the symptom indicator and the health benchmark value; Step 5: Classify the fault conditions, and construct a membership function based on the deviation of the symptom indicators obtained in step 4; Step 6: combining the sign indicator deviation and the membership function to calculate the sign indicator membership matrix; Step 7: Based on the weight matrix obtained in step 3 and the symptom indicator membership matrix obtained in step 6, the fault type fuzzy judgment matrix is ​​calculated to diagnose whether the hydropower unit has a mirror plate horizontality exceeding the standard fault.

[0008] Preferably, in step 1, based on the structural characteristics and operation mechanism of the mirror plate of the hydropower unit, the fault mechanism of the excessive horizontality of the mirror plate of the hydropower unit is analyzed, and in combination with historical monitoring data information and literature, symptom indicators reflecting the fault characteristics are selected, and the symptom indicators specifically include: Z-direction vibration amplitude of the upper and lower frames; Z-direction vibration frequency of the upper and lower frames; X- and Y-direction amplitudes of the upper guide, lower guide, and water-guide bearings; time-frequency indicators of the upper guide, lower guide, water-guide bearings and the upper and lower frames; oil quality; thrust bearing force; cooling water temperature and flow rate; impeller upper cavity pressure; axial displacement; rotation horizontality and thrust bearing temperature difference.

[0009] Preferably, in step 2, the mirror plate horizontality exceeding standard fault is hierarchically decomposed according to the principle from fault type to symptom indicator by using the hierarchical analysis method to form a hierarchical model in the form of fault type-symptom indicator association, and the hierarchical model includes a fault type layer and a symptom indicator layer.

[0010] Preferably, in step 3, the symptom index membership matrix is ​​calculated using the CRITIC weight method, and the CRITIC weight method specifically includes: (1) Normalization of symptom indicators; For Monitoring indicators Perform normalization calculation, remove the dimension of the symptom indicator, and obtain the normalized symptom indicator , n Indicates the number of symptom indicators; (2) Calculate the information content of the symptom indicators; The information content of the symptom indicators includes the difference of the symptom indicators and the conflict of the symptom indicators. The difference is expressed by the standard deviation, which is recorded as ; Set Signs and symptoms With Signs and symptoms The Pearson correlation coefficient is , then Signs and symptoms The conflict is: ; In the formula Indicates Signs and symptoms Conflict with other symptoms and indicators; No. The amount of information in each symptom indicator The calculation formula is: ; (3) Calculate the objective weights of symptom indicators; Normalize the information amount of the symptom indicators to obtain the objective weight of each symptom indicator: ; In the formula Indicates i The weight of each symptom indicator; ; In the formula is the objective weight matrix of symptom indicators.

[0011] Furthermore, in step 4, the deviation between the real-time value of the symptom indicator and the health benchmark value is calculated to obtain the symptom indicators and operating parameters of the historical health data of the hydropower unit shaft system under different operating conditions, and the operating parameters include guide vane opening, head, blade opening and active power; a nonlinear mapping relationship between the operating parameters and the health symptom indicators is constructed based on the AI ​​algorithm to obtain the health benchmark values ​​of the symptom indicators under different operating conditions; based on the health benchmark values ​​of the symptom indicators, the deviation of the real-time value of the symptom indicator is calculated.

[0012] Preferably, for the smaller the better type of symptom indicator, the calculation formula of the deviation is: ; (1) In the formula, is the deviation of the symptom indicator, is the measured value of the symptom indicator, is the upper limit of the symptom index, is the health benchmark value of the symptom indicator, and .

[0013] For the larger the better type of symptom indicators, the calculation formula for the deviation is: ; (2) In the formula, is the lower limit of the symptom indicator. is the health benchmark value of the symptom indicator, and .

[0014] Furthermore, the step 5 specifically includes the following sub-steps: Step 5.1: Classify the fault level of the hydropower unit mirror plate exceeding the standard, and the fault level includes no fault, suspected fault and fault; Step 5.2: Based on the triangle-semi-trapezoidal membership function, a membership function is constructed according to the fault condition level: ; Where f is the membership function and g is the deviation of the symptom indicator.

[0015] Further, in step 6, the calculation obtains the symptom indicator membership matrix, takes the deviation of each symptom indicator as the input of the membership function F, determines the membership of the symptom indicator to different fault condition levels according to the membership function F, and obtains the symptom indicator membership matrix : ; in n is the number of characteristic indicators, For the n The membership degree of each characteristic indicator belonging to different fault condition levels.

[0016] Preferably, the step 7 specifically includes: (1) According to the symptom index membership matrix R and the weight matrix w, the fault type fuzzy judgment matrix is ​​calculated V : ; (2) Adopt the maximum membership principle and fuzzy judgment matrix according to the fault type V , and obtain the final diagnosis result of the mirror plate horizontality exceeding the standard fault.

[0017] Compared with the prior art, the beneficial effects of the present invention include: 1) The present invention analyzes the evolution mechanism of the fault of the mirror plate level exceeding the standard of the hydropower unit, summarizes the fault manifestation form, selects the symptom indicators reflecting the fault, and uses the hierarchical analysis method to systematically analyze and construct a hierarchical structure model of the fault type-symptom indicator association form; based on the triangle-semi-trapezoidal membership function, the membership function is constructed according to the fault situation level, and the symptom indicator membership matrix is ​​calculated according to the deviation between the real-time value of the symptom indicator and the healthy benchmark value, and then the fault type fuzzy judgment matrix is ​​obtained. According to the maximum membership principle, the diagnosis result of whether the hydropower unit has the fault of the mirror plate level exceeding the standard is obtained. The present invention significantly improves the accuracy of fault judgment through systematic analysis and objective weight allocation, as well as dynamic benchmark comparison and fuzzy judgment methods. The fault judgment method of the present invention can detect potential faults in advance, avoid greater losses caused by fault deterioration, and thus reduce maintenance costs. The method provides technical support for the intelligent management of hydropower units, helps to realize fault prediction and preventive maintenance, and improves the operating reliability and safety of hydropower units.

[0018] 2) The present invention obtains the healthy baseline value under different working conditions and calculates the deviation between the real-time value and the baseline value, taking into account the impact of working condition changes on fault judgment, can adapt to the fault judgment requirements under different working conditions, and improves the accuracy of judgment.

[0019] 3) The present invention obtains the symptom indicator membership matrix through fuzzy membership function calculation, obtains the fault fuzzy judgment matrix through weighted summation, further analyzes and identifies whether the fault of excessive horizontality of the mirror plate of the hydropower unit has occurred and the confidence level, thereby achieving hydropower unit fault identification with higher credibility and more accurate results. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0021] Figure 1 The present invention is a flowchart of a method for determining a mirror plate fault of a hydropower unit according to an embodiment of the present invention.

[0022] Figure 2 It is a schematic diagram of a hierarchical structure model of fault-symptom indicators of excessive horizontality of mirror plates of a hydropower unit according to an embodiment of the present invention.

[0023] Figure 3 Schematic diagram of a triangle-semi-trapezoidal membership function according to an embodiment of the present invention. DETAILED DESCRIPTION

[0024] like Figure 1 As shown in the figure, the fault determination method of the hydropower unit mirror plate based on hierarchical analysis and fuzzy membership is implemented in the following specific steps: Step 1: Analyze the fault mechanism and extract the symptom indicators that characterize the fault of excessive horizontality of the mirror plate of the hydropower unit; In this embodiment, based on the structural characteristics of the mirror plates of the hydropower unit and the fault evolution mechanism, the fault mechanism of the mirror plate horizontality exceeding the standard of the hydropower unit is analyzed, and combined with historical monitoring data information and literature, symptom indicators reflecting the fault characteristics are selected to construct a knowledge graph of the mirror plate horizontality exceeding the standard fault of the hydropower unit.

[0025] In this embodiment, the reasons why the mirror plate horizontality exceeds the standard mainly include: 1) The upper and lower planes of the thrust head and mirror plate are not parallel due to installation and manufacturing not meeting the requirements; 2) Due to aging, the mirror plate roughness does not meet the requirements; the relationship between the mirror plate roughness and the oil film thickness must meet the film thickness ratio requirements. The film thickness ratio refers to the ratio of the minimum oil film thickness to the roughness. Under normal circumstances, its value should be in the range of 2 to 5. Once it is lower than this range, it will affect the formation of the oil film thickness, causing dry friction between the mirror plate and the bearing bush, resulting in a bush burning accident; 3) The mirror plate is too thin, resulting in insufficient rigidity. The wave degree of the mirror plate is generally required to be less than or equal to 0.05mm. During the rotation of the unit, wave creep may occur, and a gap will be generated between the mirror plate and the inference head during creep. The gap will be closed and generated continuously, resulting in cavitation and pitting on the surface of the mirror plate. 4) Thermoelastic deformation of the thrust head and mirror plate due to the non-isothermal and non-isothermal oil film.

[0026] In this embodiment, the fault phenomena that may be caused by the mirror plate horizontality exceeding the standard include: 1) Wear between the bearing and the bushing occurs; If the unevenness of the mirror plate exceeds the standard, it will affect the formation of the oil film, causing dry friction between the mirror plate and the thrust washer, thereby wearing the surface of the mirror plate and the thrust washer.

[0027] 2) The connecting bolts between the mirror plate and the thrust head become loose or broken; An uneven mirror plate will cause severe vibration, which may lead to loosening or even breaking of the connecting bolts.

[0028] 3) Insulation pad corrosion; There is usually an insulating pad between the thrust head and the mirror plate. When the waviness of the mirror plate does not meet the requirements, cavitation will occur between the insulating pad and the mirror plate, causing corrosion of the insulating pad.

[0029] 4) Oil quality deteriorates; Due to the excessive horizontality of the mirror plate, friction may occur between the mirror plate and the thrust washer. At this time, metal debris may enter the lubricating oil for the alloy washer, causing the quality of the lubricating oil to deteriorate.

[0030] 5) Lubricating oil temperature rises; The unevenness of the mirror plate will cause friction between the mirror plate and the thrust bearing, thereby causing the lubricating oil temperature to rise.

[0031] 6) The lower frame vibrates violently, and the upper guide, lower guide and water guide swing greatly.

[0032] In this embodiment, after the above fault mechanism analysis, the characteristic indicators selected to characterize the fault of excessive horizontality of the mirror plate of the hydropower unit specifically include: the vibration amplitude of the upper and lower frames in the Z direction; the vibration frequency of the upper and lower frames in the Z direction; the amplitude of the upper, lower and water-guide bearings in the X and Y directions; the time-frequency indicators of the upper guide, lower guide, water-guide bearings and the upper and lower frames; oil quality; thrust bearing force; cooling water temperature and flow rate; impeller upper cavity pressure; axial displacement; rotation horizontality and thrust bearing temperature difference.

[0033] In this embodiment, the vibration amplitude of the upper and lower frames in the Z direction, the amplitude of the upper, lower and water-guide bearings in the X and Y directions, the oil quality, the thrust bearing force, the cooling water temperature and flow rate, the pressure in the upper cavity of the impeller, the axial displacement, the rotation horizontality and the thrust bearing temperature difference are directly obtained through the industrial Internet big data platform.

[0034] In this embodiment, the Z-direction vibration frequency of the upper and lower frames, the upper guide, lower guide, water guide bearings and the time-frequency indexes of the upper and lower frames are obtained by time-frequency analysis method and intelligent analysis method. First, the vibration waveform data of the relevant measuring points are obtained, and the time-frequency analysis and intelligent analysis methods are used to extract and identify the sign indicators of the mirror plate horizontality exceeding the standard fault. The time-frequency analysis methods used to obtain time-frequency indicators include short-time Fourier transform (STFT), continuous wavelet transform (CWT), Wigner-Ville distribution (WVD), empirical mode decomposition (EMD) and a series of its improved algorithms and Hilbert-Huang transform (HHT).

[0035] Intelligent analysis methods used to obtain time-frequency indicators include advanced algorithms such as convolutional neural networks (CNNs) and autoencoders (SAEs). The time-frequency indicators of upper guide, lower guide, water guide bearings, and upper and lower frames include sample entropy, approximate entropy, singular value, energy moment, etc. of the runout signal.

[0036] Step 2: Construct a hierarchical structure model for the fault of excessive horizontality of the mirror plate of the hydropower unit; In this embodiment, based on the hierarchical analysis method, the fault of the mirror plate level exceeding the standard is hierarchically decomposed according to the principle of fault type to symptom index, forming a hierarchical structure model in the form of fault type-symptom index association. The hierarchical structure model includes a fault type layer and a symptom index layer. Figure 2 shown.

[0037] Step 3: Obtain the weight matrix of the fault symptom index of the mirror plate horizontality exceeding the standard based on the hierarchical structure model determined in step 2; In this embodiment, the calculation method of determining the objective weight of the symptom index by the CRITIC method is: (1) Normalization of symptom indicators; For Monitoring indicators Perform normalization calculation, remove the dimension of the symptom indicator, and obtain the normalized symptom indicator , n Indicates the number of symptom indicators; (2) Calculate the information content of the symptom indicators; The information content of the symptom indicators includes the difference of the symptom indicators and the conflict of the symptom indicators. The difference is expressed by the standard deviation, which is recorded as ; Set Signs and symptoms With Signs and symptoms The Pearson correlation coefficient is , then Signs and symptoms The conflict is: ; In the formula Indicates Signs and symptoms Conflict with other symptoms and indicators; No. The amount of information in each symptom indicator The calculation formula is: ; (3) Calculate the objective weights of symptom indicators; Normalize the information amount of the symptom indicators to obtain the objective weight of each symptom indicator: ; In the formula Indicates i The weight of each symptom indicator; ; In the formula is the objective weight matrix of symptom indicators.

[0038] Step 4: Obtain the health benchmark values ​​of each symptom indicator under different working conditions, and calculate the deviation of the real-time value of the symptom indicator based on the health benchmark; In this embodiment, the symptom indicators and operating parameters of the historical health data of the shaft system of the hydropower unit under different operating conditions are obtained, and the operating parameters include the guide vane opening, water head, blade opening and active power.

[0039] In this embodiment, a nonlinear mapping relationship between operating condition parameters and health symptom indicators is constructed based on an intelligent algorithm to obtain health reference values ​​of symptom indicators under different operating conditions; and the deviation of the real-time value of the symptom indicator is calculated based on the health reference value.

[0040] In this embodiment, the above intelligent algorithms include machine learning algorithms: such as BP neural network, radial basis function network (RBF Networks), support vector machine (SVM), random forest (Random Forest), etc.; deep learning technology: such as convolutional neural network (CNN), autoencoder (SAE) and long short-term memory network (LSTM), etc. The working condition parameters are used as the input of the intelligent algorithm, and the characteristic indicators under the healthy state are used as the output of the intelligent algorithm. The symptom indicator health model is trained to obtain the health benchmark value of the symptom indicator.

[0041] In this embodiment, the deviation of the real-time value of the symptom index is calculated based on the health benchmark value, and the value range is [0,1]. The deviation calculation method is as follows: For the smaller the better type of symptom index, the deviation calculation formula is: ; In the formula, To judge the deviation of the symptom indicators, is the measured value of the symptom indicator, is the upper limit of the symptom index, is the health benchmark value of the symptom indicator, and .

[0042] For the larger the better type of symptom indicators, the deviation is: ; In the formula, is the lower limit of the symptom indicator. is the health benchmark value of the symptom indicator, and .

[0043] Step 5: Determine the membership function based on the deviation design and classify the fault conditions; In this embodiment, the fault level of the mirror plate of the hydropower unit is divided into three levels: no fault, suspected fault, and fault, and the deviations corresponding to the three levels are respectively The specific description is shown in Table 1: Table 1 Description of fault level

[0044] In this embodiment, based on the triangle-semi-trapezoidal membership function, a membership function is constructed according to the fault condition level. The membership function is as follows Figure 3 shown.

[0045] Step 6: Combining the sign indicator deviation and the above membership function to obtain the sign indicator membership matrix; In this embodiment, the deviation of the fault symptom index of the horizontality of the mirror plate of the hydropower unit is calculated; the deviation is input into the above-mentioned membership function to obtain the membership of the symptom index belonging to three different levels.

[0046] Get the symptom indicator membership matrix: ; in n is the number of characteristic indicators, For the n The membership degree of each characteristic indicator belonging to different fault condition levels.

[0047] Step 7: Calculate the fault type fuzzy judgment matrix based on the fusion weight matrix and the membership matrix to diagnose whether the hydropower unit has a mirror plate horizontality exceeding the standard fault and the possibility of the fault.

[0048] In this embodiment, based on the above-mentioned membership matrix and weight matrix, the fault type fuzzy judgment matrix is ​​calculated and obtained: .

[0049] In this embodiment, the final diagnosis result of the mirror plate horizontality exceeding the standard fault is obtained according to the maximum membership principle.

[0050] The embodiment of the present invention further provides a system for determining a fault of a hydropower unit mirror plate based on hierarchical analysis and fuzzy membership, comprising: A symptom indicator module is used to obtain symptom indicators of the unit mirror plate horizontality exceeding the standard fault, and to maintain and manage the symptom indicators; A weight calculation module is used to calculate a weight matrix of symptom indicators of a fault in which the horizontality of the unit's mirror plate exceeds the standard; The deviation acquisition module obtains the health benchmark value of each symptom indicator under different working conditions based on the historical health samples, and calculates the deviation between the real-time value of the symptom indicator and the health benchmark value; The membership function construction module is used to construct the membership function of the fault of the mirror plate horizontality exceeding the standard of the hydropower unit, and input the deviation of the symptom index to calculate the membership matrix of the symptom index; The fuzzy judgment module is used to calculate the fault type fuzzy judgment matrix based on the fusion weight matrix and the membership function judgment matrix, and judge whether the hydropower unit has a mirror plate horizontality exceeding the standard fault and the confidence level of the mirror plate horizontality exceeding the standard fault.

[0051] In another embodiment of the present invention, step 3 determines the subjective weight matrix according to the 1-9 scaling method hierarchical analysis method; determines the objective weight matrix according to the CRITIC objective weight determination method; calculates the contribution value of each symptom indicator leading to the mirror plate horizontality exceeding the standard fault, and obtains the subjective and objective fusion weight matrix of the symptom indicator corresponding to the mirror plate horizontality exceeding the standard fault , combined with the symptom index membership matrix R, the fault type fuzzy judgment matrix is ​​calculated , according to the maximum membership principle, the final diagnosis result of the mirror plate horizontality exceeding the standard fault is obtained.

Claims

1. A method for determining the fault of a hydropower unit mirror plate based on hierarchical analysis and fuzzy membership, characterized in that: The following steps are involved: Step 1: Analyze the fault mechanism and determine the symptom indicators used to characterize the fault of excessive horizontality of the mirror plate of the hydropower unit; Step 2: Construct a hierarchical structure model for the fault of excessive horizontality of the mirror plate of the hydropower unit; Step 3: Based on the hierarchical structure model, a weight matrix of the symptom indicators is calculated; Step 4: Obtain the health benchmark value of each symptom indicator under different working conditions, and calculate the deviation between the real-time value of the symptom indicator and the health benchmark value; Step 5: Classify the fault conditions, and construct a membership function based on the deviation of the symptom indicators obtained in step 4; Step 6: combining the sign indicator deviation and the membership function to calculate the sign indicator membership matrix; Step 7: Based on the weight matrix obtained in step 3 and the symptom indicator membership matrix obtained in step 6, the fault type fuzzy judgment matrix is ​​calculated to diagnose whether the hydropower unit has a mirror plate horizontality exceeding the standard fault.

2. The method for determining the fault of a hydropower unit mirror plate based on hierarchical analysis and fuzzy membership according to claim 1 is characterized in that: In the step 1, based on the structural characteristics and operation mechanism of the mirror plate of the hydropower unit, the fault mechanism of the excessive horizontality of the mirror plate of the hydropower unit is analyzed, and in combination with historical monitoring data information and literature, symptom indicators reflecting the fault characteristics are selected, and the symptom indicators specifically include: Z-direction vibration amplitude of the upper and lower frames; Z-direction vibration frequency of the upper and lower frames; X- and Y-direction amplitudes of the upper guide, lower guide, and water-guide bearings; time-frequency indicators of the upper guide, lower guide, water-guide bearings and the upper and lower frames; oil quality; thrust bearing force; cooling water temperature and flow rate; impeller upper cavity pressure; axial displacement; rotation horizontality and thrust bearing temperature difference.

3. The method for determining the fault of a hydropower unit mirror plate based on hierarchical analysis and fuzzy membership according to claim 3 is characterized in that: In step 2, the mirror plate horizontality exceeding standard fault is hierarchically decomposed according to the principle from fault type to symptom index by using the hierarchical analysis method to form a hierarchical structure model in the form of fault type-symptom index association, and the hierarchical structure model includes a fault type layer and a symptom index layer.

4. The method for determining the fault of a hydropower unit mirror plate based on hierarchical analysis and fuzzy membership according to claim 3 is characterized in that: In step 3, the symptom index membership matrix is ​​calculated using the CRITIC weight method, and the CRITIC weight method specifically includes: (1) Normalization of symptom indicators; For Monitoring indicators Perform normalization calculation, remove the dimension of the symptom indicator, and obtain the normalized symptom indicator , n Indicates the number of symptom indicators; (2) Calculate the information content of the symptom indicators; The information content of the symptom indicators includes the difference of the symptom indicators and the conflict of the symptom indicators. The difference is expressed by the standard deviation, which is recorded as ; Set Signs and symptoms With Signs and symptoms The Pearson correlation coefficient is , then Signs and symptoms The conflict is: ； In the formula Indicates Signs and symptoms Conflict with other symptoms and indicators; No. The amount of information in each symptom indicator The calculation formula is: ； (3) Calculate the objective weights of symptom indicators; Normalize the information amount of the symptom indicators to obtain the objective weight of each symptom indicator: ； In the formula Indicates i The weight of each symptom indicator; ； In the formula is the objective weight matrix of symptom indicators.

5. The method for determining the fault of the hydropower unit mirror plate based on hierarchical analysis and fuzzy membership according to claim 2, 3 or 4, characterized in that: In step 4, the deviation between the real-time value of the symptom indicator and the health reference value is calculated to obtain the symptom indicators and operating parameters of the historical health data of the shaft system of the hydropower unit under different operating conditions, and the operating parameters include the guide vane opening, water head, blade opening and active power; a nonlinear mapping relationship between the operating parameters and the health symptom indicators is constructed based on the AI ​​algorithm to obtain the health reference values ​​of the symptom indicators under different operating conditions; Based on the health baseline value of the symptom indicator, the deviation of the real-time value of the symptom indicator is calculated.

6. The method for determining the fault of a hydropower unit mirror plate based on hierarchical analysis and fuzzy membership according to claim 5 is characterized in that: For the smaller the better type of symptom indicator, the calculation formula of the deviation is: ；（1） In the formula, is the deviation of the symptom indicator, is the measured value of the symptom indicator, is the upper limit of the symptom index, It is the health benchmark value of the symptom indicator; For the larger the better type of symptom indicators, the calculation formula for the deviation is: ；（2） In the formula, is the lower limit of the symptom indicator. It is the health baseline value of the symptom indicator.

7. The method for determining the fault of a hydropower unit mirror plate based on hierarchical analysis and fuzzy membership according to claim 6 is characterized in that: The step 5 specifically includes the following sub-steps: Step 5.1: Classify the fault level of the hydropower unit mirror plate exceeding the standard, and the fault level includes no fault, suspected fault and fault; Step 5.2: Based on the triangle-semi-trapezoidal membership function, a membership function is constructed according to the fault condition level: ； Where f is the membership function and g is the deviation of the symptom indicator.

8. The method for determining the fault of a hydropower unit mirror plate based on hierarchical analysis and fuzzy membership according to claim 7 is characterized in that: In step 6, the calculation obtains the symptom indicator membership matrix, takes the deviation of each symptom indicator as the input of the membership function F, determines the membership of the symptom indicator to different fault condition levels according to the membership function F, and obtains the symptom indicator membership matrix : ； in n is the number of characteristic indicators, For the n The membership degree of each characteristic indicator belonging to different fault condition levels.

9. The method for determining the fault of a hydropower unit mirror plate based on hierarchical analysis and fuzzy membership according to claim 8 is characterized in that: The step 7 specifically includes: (1) According to the symptom index membership matrix R and the weight matrix w, the fault type fuzzy judgment matrix is ​​calculated V : ； (2) Adopt the maximum membership principle and fuzzy judgment matrix according to the fault type V , and obtain the final diagnosis result of the mirror plate horizontality exceeding the standard fault.

10. The system of the method for determining the fault of the hydropower unit mirror plate based on hierarchical analysis and fuzzy membership as claimed in claim 1 or 2 or 3 or 4 or 6 or 7 or 8 or 9, characterized in that: The system comprises: A symptom indicator module is used to obtain symptom indicators of the unit mirror plate horizontality exceeding the standard fault, and to maintain and manage the symptom indicators; A weight calculation module is used to calculate a weight matrix of symptom indicators of a fault in which the horizontality of the unit's mirror plate exceeds the standard; The deviation acquisition module obtains the health benchmark value of each symptom indicator under different working conditions based on the historical health samples, and calculates the deviation between the real-time value of the symptom indicator and the health benchmark value; The membership function construction module is used to construct the membership function of the fault of the mirror plate exceeding the standard of the hydropower unit, and calculate the membership matrix of the symptom index according to the deviation of the symptom index output by the deviation acquisition module; The fuzzy judgment module calculates the fault type fuzzy judgment matrix according to the weight matrix and the symptom index membership matrix; adopts the maximum membership principle and obtains the final diagnosis result of the mirror plate horizontality exceeding standard fault according to the fault type fuzzy judgment matrix.

Images