Fault diagnosis method and device, electronic equipment and storage medium

Abstract

The invention discloses a fault diagnosis method and device, electronic equipment and a storage medium, and relates to the technical field of fault diagnosis. The fault diagnosis method comprises the following steps: acquiring vibration data of the steam turbine generator set according to measuring point parameters of the steam turbine generator set; time domain features and frequency domain features of the steam turbine generator set are determined according to the vibration data; according to the time domain features and the frequency domain features, determining the confidence of each fault symptom to which the measuring point belongs; and determining the occurrence probability corresponding to each fault type of the steam turbine generator set according to the confidence coefficient, and determining a target fault according to the occurrence probability. According to the embodiment of the invention, the occurrence probability corresponding to each fault type of the steam turbine generator set is determined by determining the confidence coefficient of each fault symptom, automatic diagnosis of the faults of the steam turbine generator set is realized, and meanwhile, the accuracy of fault diagnosis is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fault diagnosis, and in particular to a fault diagnosis method, device, electronic equipment and storage medium. Background Art

[0002] Vibration-based fault diagnosis has a history of more than 40 years and is currently the most effective method for evaluating the status of steam turbine units. With the development of sensor technology, the number of monitoring points for mechanical equipment is increasing, and fault diagnosis technology is gradually entering the era of digitalization and intelligence.

[0003] The excellent classification performance of existing deep learning diagnostic methods is often highly dependent on a large amount of labeled data, and the models usually lack reasoning capabilities. In the actual operation cycle of a steam turbine unit, multiple homogeneous or heterogeneous sensors need to be deployed on the equipment, and the generated monitoring data is mostly healthy data, while fault data is scarce or even missing. As a result, the existing deep learning diagnostic methods that can only use the quasi-static features of a single sensor and have scarce training samples are unable to perform fault diagnosis tasks in the actual production process, resulting in inaccurate fault diagnosis. Therefore, how to accurately and quickly perform fault diagnosis tasks has become a problem that needs to be solved urgently. Summary of the invention

[0004] The present invention provides a fault diagnosis method, device, electronic equipment and storage medium to solve the above problems.

[0005] According to one aspect of the present invention, a fault diagnosis method is provided, wherein the method comprises:

[0006] Acquiring vibration data of the steam turbine generator set according to measurement point parameters of the steam turbine generator set;

[0007] Determine the time domain characteristics and frequency domain characteristics of the steam turbine generator set according to the vibration data;

[0008] Determine the confidence level that the measurement point belongs to each fault sign according to the time domain characteristics and the frequency domain characteristics;

[0009] The occurrence probability corresponding to each fault type of the steam turbine generator set is determined according to the confidence level, and the target fault is determined according to the occurrence probability.

[0010] According to another aspect of the present invention, a fault diagnosis device is provided, wherein the device comprises:

[0011] A data acquisition module, used for acquiring vibration data of the steam turbine generator set according to the measurement point parameters of the steam turbine generator set;

[0012] A feature determination module, used to determine the time domain features and frequency domain features of the steam turbine generator set according to the vibration data;

[0013] A confidence determination module, used to determine the confidence that the measurement point belongs to each fault sign according to the time domain characteristics and the frequency domain characteristics;

[0014] The fault determination module is used to determine the occurrence probability corresponding to each fault type of the steam turbine generator set according to the confidence level, and determine the target fault according to the occurrence probability.

[0015] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising:

[0016] at least one processor; and

[0017] a memory communicatively connected to the at least one processor; wherein,

[0018] The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can execute the fault diagnosis method described in any embodiment of the present invention.

[0019] According to another aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the fault diagnosis method described in any embodiment of the present invention when executed.

[0020] The technical solution of the embodiment of the present invention obtains vibration data of the steam turbine generator set according to the measuring point parameters of the steam turbine generator set, determines the time domain characteristics and frequency domain characteristics of the steam turbine generator set according to the vibration data, determines the confidence of the measuring point belonging to each fault symptom according to the time domain characteristics and frequency domain characteristics, determines the occurrence probability corresponding to each fault type of the steam turbine generator set according to the confidence, determines the target fault according to the occurrence probability, realizes automatic diagnosis of steam turbine generator set faults, determines the confidence of each fault symptom and then determines the occurrence probability corresponding to each fault type of the steam turbine generator set, improves the accuracy of fault diagnosis, and enhances the user experience.

[0021] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present invention, nor are they intended to limit the scope of the present invention. Other features of the present invention will become easily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0023] Figure 1 is a flowchart of a fault diagnosis method provided according to Embodiment 1 of the present invention;

[0024] Figure 2 is a flowchart of a fault diagnosis method provided according to Embodiment 2 of the present invention;

[0025] Figure 3 is a flowchart of a fault diagnosis method provided according to Embodiment 3 of the present invention;

[0026] Figure 4 is a flowchart of the operation of a symptom judgement unit and a fault judgement unit in a reasoner provided in Embodiment 3 of the present invention;

[0027] Figure 5 is a flowchart of the operation of a conclusion output unit in a reasoner provided in Embodiment 3 of the present invention;

[0028] Figure 6 is a structural schematic diagram of a fault diagnosis device provided according to Embodiment 3 of the present invention;

[0029] Figure 7 It is a structural schematic diagram of an electronic device for implementing a fault diagnosis method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0030] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0031] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0032] Embodiment 1

[0033] Figure 1 1 is a flow chart of a fault diagnosis method provided according to the first embodiment of the present invention. This embodiment can be applied to diagnose faults of steam turbine generator sets. The method can be executed by a fault diagnosis device. The fault diagnosis device can be implemented in the form of hardware and / or software. The fault diagnosis device can be configured in an electronic device. Figure 1 As shown, the method includes:

[0034] S110. Obtain vibration data of the steam turbine generator set according to measurement point parameters of the steam turbine generator set.

[0035] Among them, a steam turbine generator set refers to a large-scale power generation equipment that converts the thermal energy of steam into electrical energy. The steam turbine generator set can be composed of a steam turbine and a generator. The measuring point parameters can be understood as the physical parameters of the measuring points of the steam turbine generator set. Generally speaking, the measuring point parameters can include the measurement name and the steam turbine generator set parameters. Exemplarily, the measuring point parameters can include but are not limited to the measurement name, rotor speed, sampling frequency, sampled data length, sampling frequency division, sampling time and number of sampling cycles. Vibration data refers to a set of information that describes the vibration characteristics of the steam turbine generator set. In actual operation, the vibration data can be a vibration waveform sequence.

[0036] In an embodiment, the measuring point data of the steam turbine generator set can be determined, and the vibration data of the steam turbine generator set can be collected according to the measuring point data. In the actual operation process, the measuring point name, sampling frequency, sampling length, sampling time and other information contained in the measuring point parameters of the steam turbine generator set can be determined, and the measuring point to be collected can be determined according to the measuring point name. The horizontal displacement amplitude and the vertical displacement amplitude of the measuring point can be collected according to the sampling frequency and sampling time based on the preset sensor, and the data of the collected sampling length constitute the vibration data of the measuring point.

[0037] S120. Determine the time domain characteristics and frequency domain characteristics of the steam turbine generator set according to the vibration data.

[0038] Among them, time domain features refer to the characteristics of the signal in the time domain, which are the characteristics obtained by directly observing the changes of the signal over time. Frequency domain features refer to the characteristics of the signal after it is converted from the time domain to the frequency domain through mathematical transformation (such as Fourier transform). In the actual operation process, time domain features may include but are not limited to the peak-to-peak value, peak value and axis trajectory of vibration data; frequency domain features may include but are not limited to the full-frequency amplitude of vibration data and the amplitude and phase corresponding to at least one frequency.

[0039] In an embodiment, the time domain characteristics and frequency domain characteristics corresponding to the vibration data can be determined, and the corresponding time domain characteristics and frequency domain characteristics can be used as the time domain characteristics and frequency domain characteristics of the steam turbine generator set. In the actual operation process, the vibration data can be subjected to wavelet transformation to obtain a wavelet coefficient matrix, and the axis trajectory corresponding to the vibration data can be determined by the wavelet coefficient matrix. In the actual operation process, the axis trajectory corresponding to the vibration data can be determined by a pre-trained convolutional neural network model. In one embodiment, the axis trajectory can include but is not limited to an eight-shaped, banana-shaped, plum-shaped, and elliptical shape. Then determine the maximum and minimum values ​​in the vibration data corresponding to each measuring point, take the difference between the maximum and minimum values ​​as the peak-to-peak value, take the maximum value as the peak value, and take the peak-to-peak value, the peak value, and the axis trajectory as the time domain characteristics of the steam turbine generator set. At the same time, Fourier transform is performed on the vibration data to determine the full-band amplitude of the vibration data and the amplitude and phase corresponding to at least one frequency, and take the full-band amplitude, the amplitude and phase corresponding to at least one frequency as frequency domain features.

[0040] S130. Determine the confidence level of the measurement point belonging to each fault sign according to the time domain characteristics and the frequency domain characteristics.

[0041] Among them, the fault signs can be understood as various abnormal phenomena exhibited by the steam turbine generator set before or at the early stage of the fault. For example, the fault signs may include but are not limited to abnormal axis trajectory, large vibration frequency, single harmonic waveform in the time domain, and fast vibration increase rate.

[0042] In an embodiment, the actual symptom value for each fault symptom may be determined according to the time domain features and the frequency domain features, and the confidence that the measuring point belongs to each fault symptom may be determined according to the actual symptom value. In the actual operation process, the actual symptom value corresponding to the time domain features and the frequency domain features may be pre-associated and set. For example, the corresponding actual symptom value may be searched through the time domain features and the frequency domain features, or a corresponding weight may be set for each time domain feature and the frequency domain feature, and the corresponding actual symptom value may be calculated according to the weight. Then, according to the confidence of each fault symptom corresponding to the actual value of the symptom, exemplarily, the confidence of each fault symptom can be determined by a preset confidence confirmation rule, exemplarily, the preset confidence upper limit value, preset confidence lower limit value, preset symptom upper limit value and preset symptom lower limit value of each fault symptom can be extracted to determine the difference between the preset confidence upper limit value and the preset confidence lower limit value of each fault symptom, determine the difference between the preset symptom upper limit value and the preset symptom lower limit value of each fault symptom, then determine the ratio of the difference between the preset confidence upper limit value and the preset confidence lower limit value and the difference between the preset symptom upper limit value and the preset symptom lower limit value of each fault symptom as the first ratio, and take the product of the first ratio and the actual value of the symptom as the confidence of the fault symptom.

[0043] S140, determining the occurrence probability corresponding to each fault type of the steam turbine generator set according to the confidence level, and determining the target fault according to the occurrence probability.

[0044] In an embodiment, the fault symptom corresponding to each fault type may be determined, and the occurrence probability corresponding to each fault type may be determined according to the confidence of the fault symptom corresponding to each fault type, and the fault type corresponding to the maximum occurrence probability may be used as the target fault; or, the fault type with an occurrence probability greater than a preset occurrence probability may be used as the target fault. In the actual operation process, there may be a preset fault symptom in each fault type, and there may be a corresponding weight for each preset fault symptom. The confidence of the corresponding fault symptom may be extracted according to the preset fault symptom, and the product of the confidence and the corresponding weight may be determined. The sum of the products may be used as the occurrence probability of the fault type, and the fault types may be sorted according to the occurrence probability, and the fault type with an occurrence probability greater than the preset occurrence probability may be used as the target fault.

[0045] The embodiment of the present invention obtains vibration data of the steam turbine generator set according to the measuring point parameters of the steam turbine generator set, determines the time domain characteristics and frequency domain characteristics of the steam turbine generator set according to the vibration data, determines the confidence that the measuring point belongs to each fault symptom according to the time domain characteristics and the frequency domain characteristics, determines the occurrence probability corresponding to each fault type of the steam turbine generator set according to the confidence, determines the target fault according to the occurrence probability, and realizes automatic diagnosis of the steam turbine generator set fault. By determining the confidence of each fault symptom and then determining the occurrence probability corresponding to each fault type of the steam turbine generator set, the accuracy of fault diagnosis is improved, and the user experience is enhanced.

[0046] In one embodiment, after obtaining the vibration data of the steam turbine generator set according to the measurement point parameters of the steam turbine generator set, the method further includes:

[0047] The vibration data during the start and stop process of the steam turbine generator set is eliminated according to the rotor speed of the steam turbine generator set, and the null value data in the vibration data is eliminated.

[0048] In an embodiment, the rotor speed of the steam turbine generator set can be determined, and the vibration data of the steam turbine generator set during the start-up and shutdown process can be determined according to the rotor speed. The vibration data during the start-up and shutdown process of the steam turbine generator set is removed, and the null value data in the vibration data is eliminated. The data is arranged according to the sampling time, and the data with the same sampling time in the X direction and the Y direction is retained to obtain the processed vibration data. In one embodiment, the vibration data with the same sampling time can then be retained to ensure that the data volume of each measuring point is the same. Among them, the rotor speed can be a pre-set working condition. Exemplarily, the rotor speed can be 3000 revolutions per minute.

[0049] In one embodiment, after determining the occurrence probability corresponding to each fault type of the steam turbine generator set according to the confidence level and determining the target fault according to the occurrence probability, the method further includes:

[0050] Determine a fault symptom corresponding to the target fault, and determine a fault symptom with a confidence level greater than zero as the target fault symptom;

[0051] Extract preset maintenance suggestions corresponding to target faults, and output target fault symptoms and preset maintenance suggestions.

[0052] In an embodiment, the fault symptoms corresponding to the target fault can be extracted, and the confidence of each fault symptom can be determined. The fault symptoms with a confidence greater than zero are used as target fault symptoms, and the preset maintenance suggestions corresponding to the target fault are extracted. The target fault symptoms and the preset maintenance suggestions are output to facilitate business personnel to repair the fault.

[0053] Embodiment 2

[0054] Figure 21 is a flowchart of a fault diagnosis method provided according to Embodiment 2 of the present invention. This embodiment is further optimized and expanded based on the above embodiment and can be combined with various optional technical solutions in the above embodiment. Figure 2 As shown, the method includes:

[0055] S201. Determine the measuring point name, sampling frequency, sampling length and sampling time included in the measuring point parameters of the steam turbine generator set.

[0056] Among them, the measuring point name can be understood as the name information of the measuring point, which is the information of the measuring point used for identification; the sampling frequency refers to the number of times vibration data is collected per second; the sampling length refers to the number of arrays collected during one collection process; the sampling time refers to the duration of one sampling.

[0057] In the embodiment, the measuring point name, sampling frequency, sampling length and sampling time included in the measuring point parameters of the steam turbine generator set can be extracted. In the actual operation process, the sampling frequency can be 2 times the rotor speed of the steam turbine generator set. v times, v can be a positive integer greater than or equal to 6; the sampling length L is the length of the vibration data sequence of a single sampling, L = Div × Cycle = 2 n , where n is a positive integer greater than or equal to 9, Div is the sampling frequency division, which is the number of sampling points in a single rotation cycle, satisfying Div = 2 v ; Cycle is the number of sampling cycles, which is twice the number of single sampling k The sampling frequency division and the number of sampling cycles can be preset.

[0058] S202, collecting the horizontal displacement amplitude and the vertical displacement amplitude of each measuring point corresponding to the measuring point name in the steam turbine generator set through a vibration displacement sensor based on a sampling frequency, a sampling length and a sampling time, and forming the vibration data of the steam turbine generator set with the horizontal displacement amplitude and the vertical displacement amplitude.

[0059] Among them, the vibration displacement sensor refers to a sensor used to measure the displacement change of an object when it vibrates. For example, the vibration displacement sensor may include but is not limited to a bearing displacement sensor, which may be arranged along the radial X direction and Y direction of the shaft.

[0060] In an embodiment, the measuring point corresponding to each measuring point name in the steam turbine generator set can be determined, and the horizontal displacement amplitude and vertical displacement amplitude of the measuring point of the sampling length are collected by a vibration displacement sensor according to the sampling frequency and sampling time. The horizontal displacement amplitude and the vertical displacement amplitude are arranged in chronological order to obtain an array sequence as the vibration data of the steam turbine generator set.

[0061] S203, performing wavelet transform on the vibration data to obtain a wavelet coefficient matrix.

[0062] The wavelet coefficient matrix refers to a matrix composed of coefficients obtained after performing wavelet transformation on vibration data.

[0063] In the embodiment, the vibration data X(t) in the X direction and the vibration data Y(t) in the Y direction can be respectively subjected to fast wavelet transform to obtain their wavelet coefficient matrix Wx and complex sequence Wy. Taking the vibration data X(t) in the X direction as an example, the calculation process is as follows: the vibration data X(t) in the X direction can be decomposed to obtain the approximate part Bx(j) and the detail part Dx(j) of the vibration data X(t):

[0064] Among them, Bx(j) represents the approximate part of the vibration data obtained by passing the vibration data X(t) through a low-pass filter, which represents the low-frequency component of the vibration data. Dx(j) represents the detailed part of the vibration data obtained by passing the vibration data X(t) through a high-pass filter, which represents the high-frequency component of the vibration data. h(t) is a low-pass filter, g(t) is a high-pass filter, t represents the tth data in the vibration data, and j and m are both integers. The approximate part Bx(j) and the detailed part Dx(j) corresponding to each size j are combined together to obtain the wavelet coefficient matrix Wx=[Ax(J)Dx(j)Dx(j-1)…Dx(j)] of the vibration data X(t) in the X direction; wherein Ax(J) represents the approximate coefficient of the vibration data X(t) at the coarsest scale, and Dx(j)Dx(j-1)…Dx(j) represent the detailed coefficients of the vibration data X(t) at different scales. Similarly, the wavelet coefficient matrix of the Y-direction vibration data Y(t) is: Wy=[Ay(J)Dy(j)Dy(j-1)…Dy(j)]; wherein Ay(J) represents the approximate coefficient of the vibration data Y(t) at the coarsest scale, and Dy(j)Dy(j-1)…Dy(j) represent the detail coefficients of the vibration data Y(t) at different scales.

[0065] S204 , inputting the wavelet coefficient matrix into a preset trajectory determination model to determine the axis trajectory corresponding to the vibration data.

[0066] Among them, the preset trajectory determination model is a convolutional neural network model; the axis trajectory at least includes: figure eight, banana, plum blossom and ellipse.

[0067] The axis trajectory refers to the movement trajectory of the axis of rotating machinery such as the rotor of a steam turbine generator set in a plane.

[0068] In an embodiment, the wavelet coefficient matrix can be input into a preset trajectory determination model, and the preset trajectory determination model outputs the axis trajectory corresponding to the vibration data. In the actual operation process, the preset trajectory determination model is trained in advance with historical data (wavelet coefficient matrix), and no further training is required during the diagnosis. The wavelet matrices Wx and Wy are input, and 0, 1, 2, and 3 are output, which respectively represent that the axis trajectory is an eight-shaped, banana-shaped, plum-shaped, and elliptical.

[0069] S205 , determining the maximum value and the minimum value in the vibration data corresponding to each measuring point, taking the difference between the maximum value and the minimum value as the peak-to-peak value, and taking the maximum value as the peak value.

[0070] In an embodiment, the vibration data corresponding to each measuring point can be extracted, the maximum value and the minimum value in the vibration data can be determined, the difference between the maximum value and the minimum value can be determined, the difference can be used as the peak-to-peak value, and the maximum value can be used as the peak value.

[0071] S206. Use the peak-to-peak value, peak value and axis trajectory as the time domain characteristics of the steam turbine generator set.

[0072] S207, performing Fourier transform on the vibration data to determine the full-band amplitude of the vibration data and the amplitude and phase corresponding to at least one frequency, and using the full-band amplitude and the amplitude and phase corresponding to at least one frequency as frequency domain features.

[0073] In the embodiment, the vibration data X(t) in the X direction and the vibration data Y(t) in the Y direction can be respectively subjected to fast Fourier transform to obtain their frequency domain complex sequence Fx and complex sequence Fy. Taking the vibration data X(t) in the X direction as an example, the calculation process is as follows: the vibration data X(t) in the X direction is decomposed into even indexes X according to the index. even (r) and odd index X odd Two subsequences of (r):

[0074] X(t)=X even (r)+X odd (r);

[0075]

[0076] Where r is a positive integer greater than 0, t = 2r, and using the sum-difference-product formula, the even index sequence X even (r) and the odd index sequence X odd (r) is expanded to obtain the Fourier series expression:

[0077]

[0078] Among them, X even (r) is the even index sequence X of the X-direction vibration dataeven The amplitude of (r), X is the even index sequence of X-direction vibration data even The phase of (r), Ax odd,0 X is the odd index sequence of vibration data in the X direction odd The amplitude of (r), X is the odd index sequence of vibration data in the X direction odd (r) phase. even (r) and the odd index sequence X odd The Fourier series expression of (r) is summed, differentiated and multiplied to obtain the Fourier series expression Fx of the vibration data X(t): Where Ax n is the amplitude of the X-direction vibration data X(t), is the phase of the X-direction vibration data X(t).

[0079] Similarly, the Y-direction vibration data Y(t) is transformed by fast Fourier transform to obtain the Fourier series expression Fy: Where Ay0 is the amplitude of the vibration data Y(t) in the Y direction, is the phase of the vibration data Y(t) in the Y direction. The passband amplitude of the vibration data and the amplitude and phase corresponding to at least one frequency are determined, and the passband amplitude, the amplitude and phase corresponding to at least one frequency are used as frequency domain features. In an embodiment, 1X represents a frequency equal to the rotation frequency, and the frequency domain features may include the passband amplitude in the X direction, 0.5X, 1X, 2X, 3X, 4X, 5X and other amplitudes and phases and the passband amplitude in the Y direction, 0.5X, 1X, 2X, 3X, 4X, 5X and other amplitudes and phases.

[0080] S208. Determine the actual symptom value of each fault symptom corresponding to the time domain characteristics and the frequency domain characteristics according to the preset confidence comparison relationship, and extract the preset confidence upper limit value, preset confidence lower limit value, preset symptom upper limit value and preset symptom lower limit value of each fault symptom.

[0081] Among them, the preset confidence comparison relationship can be understood as the correspondence between the preset time domain characteristics and frequency domain characteristics and the fault symptoms, and the actual symptom value of the corresponding fault symptom can be queried according to the time domain characteristics and frequency domain characteristics. The actual symptom value can be understood as the actual corresponding value of the fault symptom. The preset confidence upper limit value refers to the preset upper limit value of the confidence corresponding to the fault symptom; the preset confidence lower limit value refers to the preset lower limit value of the confidence corresponding to the fault symptom; the preset symptom upper limit value refers to the upper limit value corresponding to the fault symptom; the preset symptom lower limit value refers to the lower limit value corresponding to the fault symptom. The preset confidence upper limit value, the preset confidence lower limit value, the preset symptom upper limit value and the preset symptom lower limit value can all be preset according to user needs.

[0082] In an embodiment, the actual symptom value of each fault symptom can be determined in a preset confidence comparison relationship through time domain characteristics and frequency domain characteristics, and the preset confidence upper limit value, preset confidence lower limit value, preset symptom upper limit value and preset symptom lower limit value of each fault symptom stored in advance can be extracted.

[0083] S209: Determine the difference between the preset upper confidence limit value and the preset lower confidence limit value of each fault symptom as the first difference value, and determine the difference between the preset symptom upper confidence limit value and the preset symptom lower confidence limit value of each fault symptom as the second difference value.

[0084] In an embodiment, the difference between a preset upper confidence limit value and a preset lower confidence limit value of each fault symptom can be calculated and used as the first difference; the difference between a preset symptom upper limit value and a preset symptom lower limit value of each fault symptom can be calculated and used as the second difference.

[0085] S210: Determine a ratio of the first difference to the second difference as a first ratio, and use a product of the first ratio and an actual value of the symptom as a confidence level of the fault symptom.

[0086] In an embodiment, the ratio of the first difference to the second difference may be determined, the ratio is used as the first ratio, the product of the first ratio and the actual value of the symptom is determined, and the product is used as the confidence of the fault symptom. In actual operation, the calculation process of the confidence of the fault symptom may be as follows: Among them, Con j,m Indicates the confidence of the mth sign of the vibration data of the jth measuring point, Value j,m ConUp represents the actual value of the mth sign of the vibration data of the jth measuring point. j,m ConLow represents the preset confidence upper limit value of the mth fault symptom of the vibration data of the jth measuring point. j,m The preset confidence lower limit value of the mth fault symptom of the vibration data of the jth measuring point, Upper j,m Indicates the preset symptom upper limit value of the mth fault symptom of the vibration data of the jth measuring point, Lower j,m Indicates the preset symptom upper limit value of the mth fault symptom of the vibration data of the jth measuring point.

[0087] S211. Extracting a preset fault symptom corresponding to each fault type and a preset weight of each preset fault symptom.

[0088] In an embodiment, the preset fault symptoms corresponding to each fault type may be pre-set, and the preset weights of the preset fault symptoms corresponding to each fault type may also be pre-set. The preset fault symptoms corresponding to the preset fault types and the preset weights of each preset fault symptom may be extracted.

[0089] S212: Determine the sum of the product of the confidence of the preset fault symptom in each fault type and the preset weight, and use the sum as the occurrence probability of the fault type.

[0090] In an embodiment, the confidence of the preset fault symptom in each fault type can be extracted, and the product of the confidence of the preset fault symptom in each fault type and the preset weight can be determined respectively, and the respective products can be summed to obtain the occurrence probability of the fault type. For example, the occurrence probability can be Where P q represents the probability of occurrence of the qth fault type, Con q,m The confidence of the mth fault symptom of the qth fault type, Weight q,m The preset weight of the m-th fault symptom of the q-th fault type, where q is an integer greater than 0.

[0091] S213. Take the fault type with a probability of occurrence greater than a preset probability of occurrence as a target fault.

[0092] The preset occurrence probability may be understood as a preset threshold for determining a target fault. In actual operation, the preset occurrence probability may be set according to user needs.

[0093] In an embodiment, a maximum value among the occurrence probabilities may be determined, and a fault type having an occurrence probability greater than a preset occurrence probability may be taken as a target fault.

[0094] In the embodiment of the present invention, by determining the measuring point name, sampling frequency, sampling length and sampling time included in the measuring point parameters of the steam turbine generator set, the horizontal displacement amplitude and the vertical displacement amplitude of the measuring point corresponding to each measuring point name in the steam turbine generator set are collected by a vibration displacement sensor based on the sampling frequency, sampling length and sampling time, and the horizontal displacement amplitude and the vertical displacement amplitude constitute the vibration data of the steam turbine generator set, so as to achieve accurate acquisition of the vibration data; by performing wavelet transform on the vibration data to obtain a wavelet coefficient matrix, the wavelet coefficient matrix is ​​input into a preset trajectory determination model, the axis trajectory corresponding to the vibration data is determined, the maximum value and the minimum value in the vibration data corresponding to each measuring point are determined, the difference between the maximum value and the minimum value is taken as the peak-to-peak value, the maximum value is taken as the peak value, the peak-to-peak value, the peak-to-peak value and the axis trajectory are taken as the time domain characteristics of the steam turbine generator set, and the vibration data are taken as the time domain characteristics of the steam turbine generator set. According to Fourier transform, the general frequency amplitude of the vibration data and the amplitude and phase corresponding to at least one frequency are determined, and the general frequency amplitude, the amplitude and phase corresponding to at least one frequency are used as frequency domain features. The actual symptom value of each fault symptom corresponding to the time domain features and the frequency domain features is determined according to the preset confidence comparison relationship, and the preset confidence upper limit value, preset confidence lower limit value, preset symptom upper limit value and preset symptom lower limit value of each fault symptom are extracted to determine the confidence of the fault symptom, the preset fault symptoms corresponding to each fault type and the preset weight of each preset fault symptom are extracted, and the sum of the products of the confidence of the preset fault symptoms and the preset weight in each fault type is determined, and the sum is used as the occurrence probability of the fault type, and the fault type with a probability of occurrence greater than the preset occurrence probability is used as the target fault, so as to improve the accuracy and efficiency of fault determination and enhance the user experience.

[0095] Embodiment 3

[0096] Figure 3 is a flow chart of a fault diagnosis method provided according to Embodiment 3 of the present invention. This embodiment takes the fault diagnosis of the inferencer as an example, takes the convolutional neural network and the network as the preset trajectory determination model as an example, takes the symptom as the fault symptom as an example, and further describes a fault diagnosis method, such as Figure 3 As shown, the method includes:

[0097] Step 1: Collect the measurement point parameters and vibration data of the steam turbine generator set.

[0098] In an embodiment, the measuring point name Name, vibration data x(t) and y(t) of the steam turbine generator set can be collected, and the parameters of the steam turbine generator set can be obtained at the same time: the rotor speed (Rotspeed) at the sampling moment, the sampling frequency (Freq) at the sampling moment, the sampled data length (Len), the sampling frequency division (Div), the sampling time (Time) and the number of sampling cycles (Cycle), and the measuring point name and the steam turbine generator set parameters are used as measuring point parameters. Among them, the vibration data x(t) is the vibration waveform sequence measured by the X-direction sensor, and the vibration data y(t) is the vibration waveform sequence measured by the Y-direction sensor. The X-direction sensor and the Y-direction sensor are arranged at the bearing of the steam turbine generator set and are perpendicular to each other. Among them, the sampling frequency can be 2 times the rotor speed of the steam turbine generator set. v times, v can be a positive integer greater than or equal to 6; the sampling length L is the length of the vibration data sequence of a single sampling, L = Div × Cycle = 2 n , where n is a positive integer greater than or equal to 9, Div is the sampling frequency division, which is the number of sampling points in a single rotation cycle, satisfying Div = 2 v ; Cycle is the number of sampling cycles, which is twice the number of single sampling k The sampling frequency division and the number of sampling cycles can be preset.

[0099] In one embodiment, the sampling method includes sampling length and sampling frequency. In the embodiment of the present invention, synchronous full cycle sampling can be adopted, n is 8, and k is 3. Then the sampling frequency of synchronous full cycle sampling is 128 times the rotation speed of the rotating machine, and 8 cycles of vibration signals are collected each time.

[0100] Step 2: pre-process the collected vibration data, remove null value data, and obtain processed vibration data.

[0101] In an embodiment, the collected vibration data x(t) and y(t) can be preprocessed to eliminate null value data, and the vibration data during the start and stop of the steam turbine unit can be eliminated according to the rotor speed (Rotspeed). The data is arranged according to the sampling time (Time), and the data with the same sampling time in the X direction and the Y direction are retained to obtain the processed vibration data X(t) and Y(t).

[0102] In one embodiment, the vibration data collected can be based on the preset working conditions, the vibration data when the equipment is started and stopped is eliminated, and then the vibration data with the same sampling time is retained to ensure that the data volume of each measuring point is the same. Specifically, in this embodiment, the working condition is that the speed reaches 3000 revolutions per minute.

[0103] Step 3: Perform wavelet transform on the preprocessed vibration data to obtain the wavelet coefficient matrix, and then input the obtained matrix into the convolutional neural network to obtain the axis trajectory and determine the fault signs of the axis trajectory.

[0104] In the embodiment, the pre-processed vibration data X(t) and Y(t) are subjected to wavelet transformation to obtain wavelet coefficient matrices Wx and Wy, and then the obtained matrices are input into the pre-trained convolutional neural network to obtain the axis trajectory and judge the fault signs of the axis trajectory. In actual operation, the X-direction vibration data X(t) and the Y-direction vibration data Y(t) are subjected to fast wavelet transformation to obtain their wavelet coefficient matrices Wx and complex sequence Wy. Taking the X-direction vibration data X(t) as an example, the calculation process is as follows:

[0105] The X-direction vibration data X(t) is decomposed to obtain the approximate part Bx(j) and the detailed part Dx(j) of the vibration data X(t): Among them, Bx(j) represents the approximate part of the vibration data obtained by passing the vibration data X(t) through a low-pass filter, which represents the low-frequency component of the vibration data. Dx(j) represents the detailed part of the vibration data obtained by passing the vibration data X(t) through a high-pass filter, which represents the high-frequency component of the vibration data. h(t) is a low-pass filter, g(t) is a high-pass filter, t represents the tth data in the vibration data, and j and m are both integers. The approximate part Bx(j) and the detailed part Dx(j) corresponding to each size j are combined together to obtain the wavelet coefficient matrix Wx=[Ax(J)Dx(j)Dx(j-1)…Dx(j)] of the vibration data X(t) in the X direction; wherein Ax(J) represents the approximate coefficient of the vibration data X(t) at the coarsest scale, and Dx(j)Dx(j-1)…Dx(j) represent the detailed coefficients of the vibration data X(t) at different scales. Similarly, the wavelet coefficient matrix of the Y-direction vibration data Y(t) is: Wy=[Ay(J)Dy(j)Dy(j-1)…Dy(j)]; wherein Ay(J) represents the approximate coefficient of the vibration data Y(t) at the coarsest scale, and Dy(j)Dy(j-1)…Dy(j) represent the detail coefficients of the vibration data Y(t) at different scales.

[0106] Step 4: Perform Fourier transform on the preprocessed vibration data to obtain the time domain characteristics and frequency domain characteristics of the steam turbine generator set vibration fault.

[0107] In the embodiment, the vibration data X(t) in the X direction and the vibration data Y(t) in the Y direction can be respectively subjected to fast Fourier transform to obtain their frequency domain complex sequence Fx and complex sequence Fy. Taking the vibration data X(t) in the X direction as an example, the calculation process is as follows: the vibration data X(t) in the X direction is decomposed into even indexes X according to the index. even (r) and odd index Xodd Two subsequences of (r):

[0108] X(t)=X even (r)+X odd (r);

[0109]

[0110] Where r is a positive integer greater than 0, t = 2r, and using the sum-difference-product formula, the even index sequence X even (r) and the odd index sequence X odd (r) is expanded to obtain the Fourier series expression:

[0111]

[0112] Among them, X even (r) is the even index sequence X of the X-direction vibration data even The amplitude of (r), X is the even index sequence of X-direction vibration data even The phase of (r), Ax odd,0 X is the odd index sequence of vibration data in the X direction odd The amplitude of (r), X is the odd index sequence of vibration data in the X direction odd (r) phase. even (r) and the odd index sequence X odd The Fourier series expression of (r) is summed, differentiated and multiplied to obtain the Fourier series expression Fx of the vibration data X(t): Where Ax n is the amplitude of the X-direction vibration data X(t), is the phase of the X-direction vibration data X(t).

[0113] Similarly, the Y-direction vibration data Y(t) is transformed by fast Fourier transform to obtain the Fourier series expression Fy: Where Ay0 is the amplitude of the vibration data Y(t) in the Y direction, is the phase of the vibration data Y(t) in the Y direction. The passband amplitude of the vibration data and the amplitude and phase corresponding to at least one frequency are determined, and the passband amplitude, the amplitude and phase corresponding to at least one frequency are used as frequency domain features. In an embodiment, 1X represents a frequency equal to the rotation frequency, and the frequency domain features may include the passband amplitude in the X direction, 0.5X, 1X, 2X, 3X, 4X, 5X and other amplitudes and phases and the passband amplitude in the Y direction, 0.5X, 1X, 2X, 3X, 4X, 5X and other amplitudes and phases.

[0114] In one embodiment, the peak-to-peak value xppv and ppv The calculation method is as follows:

[0115] x ppv =x max -x min ;

[0116] y ppv =y max -y min ;

[0117] Among them, x ppv is the peak-to-peak value of the vibration data sequence in the X direction, x max is the maximum value of the vibration amplitude of the vibration data sequence in the X direction, x min is the minimum value of the vibration amplitude of the vibration data sequence in the X direction, y ppv is the peak-to-peak value of the vibration data sequence in the Y direction, y max is the maximum value of the vibration amplitude of the vibration data sequence in the Y direction, y min It is the minimum value of the vibration amplitude of the vibration data sequence in the Y direction.

[0118] In one embodiment, in order to suppress the interference of accidental factors on the peak-to-peak value of the signal, the vibration data collected in the X direction and the Y direction can be divided into two n Molecular sequence, n is greater than or equal to 2 and less than 2 k-1 A positive integer. After calculating the peak-to-peak value for each subsequence, the average is calculated as follows:

[0119]

[0120] Among them, x ppv,m is the peak-to-peak value of the m-th subsequence, It is divided into 2 n The average of the peak-to-peak values ​​of the molecular sequence.

[0121] In one embodiment, the peak value x pv and pv The calculation method is as follows:

[0122] x pv =x max ;y pv =y max ;

[0123] Among them, x pv is the peak value of the vibration data sequence in the X direction, x max is the maximum value of the vibration amplitude of the vibration data sequence in the X direction, y pv is the peak value of the vibration data sequence in the Y direction, y max It is the maximum value of the vibration amplitude of the vibration data sequence in the Y direction.

[0124] Step 5: Input the obtained measurement point parameters, time domain features, and frequency domain features into the inference engine to diagnose the fault type and corresponding occurrence probability, fault symptoms, and maintenance suggestions of the steam turbine generator set.

[0125] Among them, the reasoner includes a symptom judge, a fault judge and a conclusion output, and the symptom judge, the fault judge and the conclusion output are supported by a fault knowledge base. In one embodiment, in order to accurately use the data of multiple homogeneous or heterogeneous measuring point sensors arranged on the equipment, the fault knowledge base divides the entire rotating machinery equipment system into subsystems such as steam turbines, induced draft fans, primary fans, and steam-driven water supply pumps when it is established. The knowledge is summarized by in-depth research and understanding of the structure and working principle of rotating machinery to obtain principle knowledge, as well as summarizing the historical diagnosis experience of rotating machinery or repeatedly communicating with experts and equipment maintenance personnel to obtain practical diagnostic methods and experience knowledge, which are summarized into the fault types and corresponding fault symptoms and maintenance suggestions of the rotating machinery equipment subsystems.

[0126] In one embodiment, Figure 4 FIG. 1 is a flowchart of the operation of a symptom judgement unit and a fault judgement unit in a reasoner according to the third embodiment of the present invention. Figure 4 As shown, firstly, the measuring point parameters, time domain characteristics, and frequency domain characteristics in the X direction and the Y direction are input into the symptom judger, and each symptom value of the measuring point is determined to determine whether the symptom value exceeds the lower limit of the fault symptom value. If not, the confidence of the symptom at the measuring point is determined to be 0; if so, the confidence of the measuring point is determined according to the linear confidence rule, and all the symptoms of the steam turbine generator set and their corresponding confidences are obtained.

[0127] Among them, the linear confidence rule can be as follows:

[0128] Among them, Con j,m Indicates the confidence of the mth sign of the vibration data of the jth measuring point, Value j,m ConUp represents the actual value of the mth sign of the vibration data of the jth measuring point. j,m ConLow represents the preset confidence upper limit value of the mth fault symptom of the vibration data of the jth measuring point. j,m The preset confidence lower limit value of the mth fault symptom of the vibration data of the jth measuring point, Upper j,m Indicates the preset symptom upper limit value of the mth fault symptom of the vibration data of the jth measuring point, Lower j,m The preset symptom upper limit value of the mth fault symptom of the vibration data of the jth measuring point is represented. Then the obtained symptom and its confidence are input into the fault judgement device, and the fault type and corresponding probability of the steam turbine generator set can be diagnosed.

[0129] Then input the obtained signs and their confidence levels into the fault judgement device, determine all possible fault types of the equipment according to the measuring point name, determine the probability of the fault type occurring, and judge whether the probability of the fault occurring is greater than a threshold. If so, the fault type is judged to occur, and the process continues until all fault types have been judged. Judge whether the fault type that occurs is an empty set. If so, determine that no fault has occurred. If not, arrange the fault types according to their probability.

[0130] The calculation formula for the occurrence probability is as follows:

[0131] Where P q represents the probability of occurrence of the qth fault type, Con q,m The confidence of the mth fault symptom of the qth fault type, Weight q,m The preset weight of the m-th fault symptom of the q-th fault type, where q is an integer greater than 0.

[0132] In one embodiment, Figure 5 FIG. 1 is a flowchart of the operation of a conclusion output unit in a reasoner according to the third embodiment of the present invention. Figure 5 As shown in the figure, the measurement point name, the fault type, symptom and corresponding confidence level can be input into the conclusion output device, and the maintenance suggestion corresponding to the fault type can be determined in the fault knowledge base. The sign with a confidence level of 0 under the fault type is deleted, and the remaining ones are the target fault signs. Finally, the fault type and the corresponding probability of occurrence, fault symptom and maintenance suggestion are integrated and output.

[0133] Embodiment 4

[0134] Figure 6 Schematic diagram of a fault diagnosis device according to Embodiment 3 of the present invention. Figure 6 As shown, the device includes: a data acquisition module 61, a feature determination module 62, a confidence determination module 63 and a fault determination module 64.

[0135] Wherein, the data acquisition module 61 is used to acquire the vibration data of the steam turbine generator set according to the measurement point parameters of the steam turbine generator set;

[0136] A feature determination module 62, for determining the time domain features and frequency domain features of the steam turbine generator set according to the vibration data;

[0137] A confidence determination module 63, used to determine the confidence that the measurement point belongs to each fault sign according to the time domain characteristics and the frequency domain characteristics;

[0138] The fault determination module 64 is used to determine the occurrence probability corresponding to each fault type of the steam turbine generator set according to the confidence level, and determine the target fault according to the occurrence probability.

[0139] In the embodiment of the present invention, the vibration data of the steam turbine generator set is acquired according to the measuring point parameters of the steam turbine generator set by the data acquisition module, the feature determination module determines the time domain characteristics and frequency domain characteristics of the steam turbine generator set according to the vibration data, the confidence determination module determines the confidence that the measuring point belongs to each fault symptom according to the time domain characteristics and the frequency domain characteristics, the fault determination module determines the occurrence probability corresponding to each fault type of the steam turbine generator set according to the confidence, determines the target fault according to the occurrence probability, realizes automatic diagnosis of the steam turbine generator set fault, and determines the occurrence probability corresponding to each fault type of the steam turbine generator set by determining the confidence of each fault symptom, thereby improving the accuracy of fault diagnosis and enhancing the user experience.

[0140] In one embodiment, the data acquisition module 61 includes:

[0141] A parameter determination unit, used to determine the measurement point name, sampling frequency, sampling length and sampling time included in the measurement point parameters of the steam turbine generator set;

[0142] The data determination unit is used to collect the horizontal displacement amplitude and the vertical displacement amplitude of each measuring point corresponding to the measuring point name in the steam turbine generator set through a vibration displacement sensor based on the sampling frequency, sampling length and sampling time, and form the vibration data of the steam turbine generator set with the horizontal displacement amplitude and the vertical displacement amplitude.

[0143] In one embodiment, the feature determination module 62 includes:

[0144] A matrix determination unit, used for performing wavelet transformation on the vibration data to obtain a wavelet coefficient matrix;

[0145] A trajectory determination unit is used to input the wavelet coefficient matrix into a preset trajectory determination model to determine the axis trajectory corresponding to the vibration data; wherein the preset trajectory determination model is a convolutional neural network model; the axis trajectory includes at least: figure eight, banana, plum blossom and ellipse;

[0146] A peak value determination unit is used to determine the maximum value and the minimum value in the vibration data corresponding to each measuring point, and the difference between the maximum value and the minimum value is taken as the peak-to-peak value, and the maximum value is taken as the peak value;

[0147] A time domain feature determination unit, used for taking the peak-to-peak value, the peak value and the axis trajectory as the time domain features of the steam turbine generator set;

[0148] The frequency domain feature determination unit is used to perform Fourier transform on the vibration data, determine the full-band amplitude of the vibration data and the amplitude and phase corresponding to at least one frequency, and use the full-band amplitude and the amplitude and phase corresponding to at least one frequency as frequency domain features.

[0149] In one embodiment, the confidence determination module 63 includes:

[0150] An actual value determination unit, used to determine the actual value of each fault symptom corresponding to the time domain feature and the frequency domain feature according to a preset confidence comparison relationship, and extract a preset confidence upper limit value, a preset confidence lower limit value, a preset symptom upper limit value and a preset symptom lower limit value of each fault symptom;

[0151] a difference determination unit, configured to determine a difference between a preset upper confidence limit value and a preset lower confidence limit value of each fault symptom as a first difference value, and determine a difference between a preset symptom upper limit value and a preset symptom lower limit value of each fault symptom as a second difference value;

[0152] The confidence determination unit is used to determine the ratio of the first difference to the second difference as a first ratio, and to use the product of the first ratio and the actual value of the symptom as the confidence of the symptom.

[0153] In one embodiment, the fault determination module 64 includes:

[0154] A weight extraction unit, used to extract a preset fault symptom corresponding to each fault type and a preset weight of each preset fault symptom;

[0155] A probability determination unit, used to determine the sum of the product of the confidence of the preset fault symptom in each fault type and the preset weight, and use the sum as the occurrence probability of the fault type;

[0156] The fault determination unit is used to take the fault type with a probability of occurrence greater than a preset probability of occurrence as a target fault.

[0157] In one embodiment, the fault diagnosis device further includes:

[0158] The data preprocessing module is used to eliminate the vibration data during the start and stop process of the steam turbine generator set according to the rotor speed of the steam turbine generator set, and eliminate the null value data in the vibration data.

[0159] In one embodiment, the fault diagnosis device further includes:

[0160] A target symptom determination module is used to determine the fault symptom corresponding to the target fault, and determine the fault symptom with a confidence level greater than zero as the target fault symptom;

[0161] The suggestion determination module is used to extract the preset maintenance suggestions corresponding to the target fault and output the target fault symptoms and preset maintenance suggestions.

[0162] The fault diagnosis device provided in the embodiment of the present invention can execute the fault diagnosis method provided in any embodiment of the present invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0163] Embodiment 5

[0164] Figure 7 1 is a schematic diagram of the structure of an electronic device for implementing a fault diagnosis method of an embodiment of the present invention. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device may also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present invention described and / or required herein.

[0165] like Figure 7 As shown, the electronic device 10 includes at least one processor 11, and a memory connected to the at least one processor 11, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., wherein the memory stores a computer program that can be executed by at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 to the random access memory (RAM) 13. In the RAM 13, various programs and data required for the operation of the electronic device 10 can also be stored. The processor 11, the ROM 12, and the RAM 13 are connected to each other through a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0166] A number of components in the electronic device 10 are connected to the I / O interface 15, including: an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0167] The processor 11 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various special artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as a fault diagnosis method.

[0168] In some embodiments, the fault diagnosis method may be implemented as a computer program, which is tangibly contained in a computer-readable storage medium, such as a storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on the electronic device 10 via the ROM 12 and / or the communication unit 19. When the computer program is loaded into the RAM 13 and executed by the processor 11, one or more steps of the fault diagnosis method described above may be performed. Alternatively, in other embodiments, the processor 11 may be configured to perform the fault diagnosis method in any other appropriate manner (e.g., by means of firmware).

[0169] Various implementations of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on chips (SOCs), load programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various implementations can include: being implemented in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0170] Computer programs for implementing the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, so that when the computer program is executed by the processor, the functions / operations specified in the flow chart and / or block diagram are implemented. The computer program may be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0171] In the context of the present invention, a computer-readable storage medium may be a tangible medium that may contain or store a computer program for use by or in combination with an instruction execution system, device or equipment. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, devices or equipment, or any suitable combination of the foregoing. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0172] To provide interaction with a user, the systems and techniques described herein may be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or trackball) through which the user can provide input to the electronic device. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0173] The systems and techniques described herein may be implemented in a computing system that includes backend components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes frontend components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such backend components, middleware components, or frontend components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a local area network (LAN), a wide area network (WAN), a blockchain network, and the Internet.

[0174] A computing system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The client and server relationship is generated by computer programs running on the corresponding computers and having a client-server relationship with each other. The server may be a cloud server, also known as a cloud computing server or cloud host, which is a host product in the cloud computing service system to solve the defects of difficult management and weak business scalability in traditional physical hosts and VPS services.

[0175] It should be understood that the various forms of processes shown above can be used to reorder, add or delete steps. For example, the steps described in the present invention can be executed in parallel, sequentially or in different orders, as long as the desired results of the technical solution of the present invention can be achieved, and this document does not limit this.

[0176] The above specific implementations do not constitute a limitation on the protection scope of the present invention. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A fault diagnosis method, characterized in that: include: Acquiring vibration data of the steam turbine generator set according to measurement point parameters of the steam turbine generator set; Determine the time domain characteristics and frequency domain characteristics of the steam turbine generator set according to the vibration data; Determine the confidence level that the measurement point belongs to each fault sign according to the time domain characteristics and the frequency domain characteristics; The occurrence probability corresponding to each fault type of the steam turbine generator set is determined according to the confidence level, and the target fault is determined according to the occurrence probability.

2. The method according to claim 1, characterized in that The step of obtaining vibration data of the steam turbine generator set according to the measurement point parameters of the steam turbine generator set comprises: Determine the measuring point name, sampling frequency, sampling length and sampling time included in the measuring point parameters of the steam turbine generator set; The horizontal displacement amplitude and the vertical displacement amplitude of each measuring point corresponding to the measuring point name in the steam turbine generator set are collected by a vibration displacement sensor based on the sampling frequency, the sampling length and the sampling time, and the horizontal displacement amplitude and the vertical displacement amplitude constitute the vibration data of the steam turbine generator set.

3. The method according to claim 1, characterized in that Determining the time domain characteristics and frequency domain characteristics of the steam turbine generator set according to the vibration data includes: Perform wavelet transform on vibration data to obtain wavelet coefficient matrix; Input the wavelet coefficient matrix into a preset trajectory determination model to determine the axis trajectory corresponding to the vibration data; wherein the preset trajectory determination model is a convolutional neural network model; the axis trajectory includes at least: an eight-shaped shape, a banana shape, a plum blossom shape, and an ellipse; Determine the maximum value and the minimum value in the vibration data corresponding to each measuring point, take the difference between the maximum value and the minimum value as the peak-to-peak value, and take the maximum value as the peak value; Taking the peak-to-peak value, the peak value and the axis trajectory as time domain features of the steam turbine generator set; The vibration data is subjected to Fourier transformation to determine a passband amplitude of the vibration data and an amplitude and a phase corresponding to at least one frequency, and the passband amplitude, the amplitude and the phase corresponding to at least one frequency are used as frequency domain features.

4. The method according to claim 1, characterized in that: The step of determining the confidence level of the measurement point belonging to each fault symptom according to the time domain characteristics and the frequency domain characteristics includes: Determine the actual value of each fault symptom corresponding to the time domain feature and the frequency domain feature according to the preset confidence comparison relationship, and extract the preset confidence upper limit value, preset confidence lower limit value, preset symptom upper limit value and preset symptom lower limit value of each fault symptom; Determine the difference between the preset upper confidence limit value and the preset lower confidence limit value of each fault symptom as the first difference value, and determine the difference between the preset symptom upper limit value and the preset symptom lower limit value of each fault symptom as the second difference value; The ratio of the first difference to the second difference is determined as a first ratio, and the product of the first ratio and the actual value of the symptom is used as the confidence of the symptom.

5. The method according to claim 1, characterized in that Determining the occurrence probability corresponding to each fault type of the steam turbine generator set according to the confidence level, and determining the target fault according to the occurrence probability, includes: Extracting a preset fault symptom corresponding to each fault type and a preset weight of each of the preset fault symptom; Determine the sum of the product of the confidence of the preset fault symptom in each fault type and the preset weight, and use the sum as the occurrence probability of the fault type; The fault type with a probability of occurrence greater than a preset probability of occurrence is taken as a target fault.

6. The method according to claim 1, characterized in that After obtaining the vibration data of the steam turbine generator set according to the measuring point parameters of the steam turbine generator set, the method further includes: The vibration data during the start and stop process of the steam turbine generator set is eliminated according to the rotor speed of the steam turbine generator set, and the null value data in the vibration data is eliminated.

7. The method according to claim 1, characterized in that After determining the occurrence probability corresponding to each fault type of the steam turbine generator set according to the confidence level and determining the target fault according to the occurrence probability, the method further includes: Determine a fault symptom corresponding to the target fault, and determine a fault symptom with a confidence level greater than zero as the target fault symptom; Extract the preset maintenance suggestion corresponding to the target fault, and output the target fault symptom and the preset maintenance suggestion.

8. A fault diagnosis device, characterized in that: include: A data acquisition module, used for acquiring vibration data of the steam turbine generator set according to the measurement point parameters of the steam turbine generator set; A feature determination module, used to determine the time domain features and frequency domain features of the steam turbine generator set according to the vibration data; A confidence determination module, used to determine the confidence that the measurement point belongs to each fault sign according to the time domain characteristics and the frequency domain characteristics; The fault determination module is used to determine the occurrence probability corresponding to each fault type of the steam turbine generator set according to the confidence level, and determine the target fault according to the occurrence probability.

9. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the fault diagnosis method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the fault diagnosis method according to any one of claims 1 to 7 when executed.

Images