Intelligent diagnosis method and system for vibration fault of rotating machinery

Abstract

The application provides a rotating machinery vibration fault intelligent diagnosis method and system, first, the working condition related vibration data is determined; then, the rotating speed synchronous resampling and order spectrum transformation are carried out on the working condition related vibration data, the rotating speed normalized vibration spectrum corresponding to each rotating speed working condition is generated; then, the rotating frequency multiple order energy distribution, bearing characteristic order energy distribution and modulation sideband interval characteristics are extracted according to the rotating speed normalized vibration spectrum, the vibration fault sensitive feature set is constructed; then, the rotating frequency synchronous phase reference is determined based on the rotating speed pulse signal, the vibration fault sensitive feature set is associated with the cross-working condition phase consistency, and the fault response consistency coefficient is obtained; finally, the fault type confidence evaluation matrix is constructed according to the vibration fault sensitive feature set, the fault response consistency coefficient and the measuring point arrangement information, and the rotating machinery vibration fault type and the corresponding measuring point position are output. The above scheme can realize the stable identification and measuring point positioning of the rotating machinery vibration fault under the variable rotating speed working condition.

Description

Technical Field

[0001] This application relates to the field of rotating machinery technology, and more specifically, to an intelligent diagnostic method and system for vibration faults in rotating machinery. Background Technology

[0002] Rotating machinery is a key piece of equipment in pumps, fans, compressors, steam turbines, gearboxes, and motor drives. Its operating status directly affects production continuity and equipment safety. However, during operation, rotating machinery is affected by the combined effects of rotor mass distribution, shaft alignment, bearing contact, gear meshing, and changes in support stiffness. These factors manifest as measurable vibration responses through periodic excitation, impact excitation, and modulation effects.

[0003] Current fault diagnosis of rotating machinery typically relies on fixed frequency spectra, vibration amplitude under a single speed condition, or general classification results. However, when the equipment is accelerating, decelerating, or experiencing load fluctuations, the position of the same fault excitation on the frequency axis drifts with the speed change, making it difficult to consistently correlate fixed frequency characteristics with the fault source. Furthermore, the vibration response received at different measuring points is significantly affected by the structural transmission path. Without a synchronized phase reference for the speed and consistent correlation across operating conditions, changes in operating conditions, structural transmission attenuation, or short-term impact disturbances can easily be misdiagnosed as genuine faults.

[0004] Therefore, how to convert vibration characteristics from frequency coordinates to order coordinates that match the rotational motion under variable speed conditions, and further utilize cross-condition phase consistency and measurement point layout information to achieve stable output of the vibration fault type and corresponding measurement point location of rotating machinery, has become an urgent problem to be solved in the field of intelligent diagnosis of vibration faults in rotating machinery. Summary of the Invention

[0005] This application provides an intelligent diagnostic method and system for vibration faults in rotating machinery, which can achieve stable identification and measurement point location of vibration faults in rotating machinery under variable speed conditions.

[0006] In a first aspect, this application provides an intelligent diagnostic method for vibration faults in rotating machinery, comprising the following steps: The vibration fault monitoring of rotating machinery is initiated under multiple speed conditions. The vibration signals, speed pulse signals and measurement point layout information of each vibration measuring point are collected simultaneously to obtain vibration data related to the operating conditions. The speed synchronous resampling and order spectrum transformation are performed on the vibration data associated with the working conditions to generate the speed normalized vibration spectrum corresponding to each speed working condition. Based on the normalized vibration spectrum of the rotational speed, the energy distribution of the rotational frequency multiple order, the energy distribution of the bearing characteristic order, and the modulation sideband interval features are extracted to construct a set of vibration fault sensitive features under each rotational speed condition. Based on the speed pulse signal, the frequency synchronization phase reference between each speed condition is determined. Based on the frequency synchronization phase reference, the vibration fault sensitive feature set is correlated across operating conditions to obtain the fault response consistency coefficient. Based on the vibration fault sensitive feature set, the fault response consistency coefficient, and the measurement point layout information, a fault type confidence evaluation matrix is ​​constructed, and the rotating machinery vibration fault type and corresponding measurement point location are output based on the fault type confidence evaluation matrix.

[0007] In some embodiments, rotating machinery vibration fault monitoring is initiated under multiple speed conditions, and vibration signals, speed pulse signals, and measurement point layout information are collected simultaneously from each vibration measuring point to obtain condition-related vibration data, specifically including: Vibration sensors are installed in the bearing housings, housing supports, and transmission ends of rotating machinery, and the installation positions of the corresponding measuring points of each vibration sensor are recorded. A rotational speed pulse signal synchronized with the rotating shaft is acquired by a rotational speed sensor, and the rotational speed pulse signal is used as the synchronization reference for the vibration signal of each measuring point. Multiple speed conditions are divided into stable speed range, acceleration range, and deceleration range. The vibration signals, speed pulse signals, and measurement point layout information of each speed condition are time-scaled to form condition-related vibration data.

[0008] In some embodiments, performing speed synchronous resampling and order spectrum transformation on the vibration data associated with the operating conditions to generate a speed-normalized vibration spectrum corresponding to each speed operating condition specifically includes: The instantaneous rotation angle of the rotating shaft at each sampling moment is calculated based on the rotational speed pulse signal; The vibration signal at the measuring point is resampled at equal angular intervals to obtain the angular domain vibration sequence; The angular domain vibration sequence is subjected to order spectrum transformation to obtain an order spectrum with the frequency multiple as the horizontal axis and the vibration amplitude as the vertical axis. The order spectra are arranged according to the rotational speed conditions and vibration measurement points to generate a rotational speed-normalized vibration spectrum.

[0009] In some embodiments, extracting the rotational frequency order energy distribution, bearing characteristic order energy distribution, and modulation sideband spacing features from the normalized vibration spectrum to construct a vibration fault sensitive feature set for each rotational speed condition specifically includes: The energy distributions corresponding to the first, second, and third rotational frequency orders are extracted from the normalized vibration spectrum to obtain the rotational frequency order energy distributions. Based on the geometric parameters and rotational speed pulse signals of the rolling bearing in rotating machinery, the bearing characteristic order corresponding to the outer ring, inner ring, rolling elements and cage is determined, and the energy distribution of the corresponding order is extracted to obtain the bearing characteristic order energy distribution. Identify the sideband structures that appear on both sides of the gear meshing order or bearing characteristic order with the frequency order as the interval, extract the sideband interval, sideband amplitude ratio and sideband symmetry to obtain the modulation sideband interval characteristics; The frequency multiplication order energy distribution, the bearing characteristic order energy distribution, and the modulation sideband interval feature are combined to construct a vibration fault sensitive feature set under each speed condition.

[0010] In some embodiments, determining a rotational frequency synchronization phase reference between different speed conditions based on the rotational speed pulse signal, and performing cross-condition phase consistency correlation on the vibration fault sensitive feature set according to the rotational frequency synchronization phase reference to obtain the fault response consistency coefficient specifically includes: The zero-phase position of the rotating shaft under each speed condition is determined based on the speed pulse signal, and the frequency synchronization phase reference is obtained. Phase alignment is performed on the vibration fault sensitive feature set of the same measuring point under different speed conditions according to the aforementioned frequency synchronization phase reference. Calculate the order position stability, energy change consistency, and phase repeatability of the same fault-sensitive features under multiple speed conditions after phase alignment; The fault response consistency coefficient is determined based on the order position stability, the energy change consistency, and the phase repeatability.

[0011] In some embodiments, constructing a fault type confidence assessment matrix based on the vibration fault sensitive feature set, the fault response consistency coefficient, and the measuring point layout information specifically includes: Establish fault feature matching rules corresponding to rotor imbalance, rotor misalignment, rolling bearing damage, and abnormal gear meshing; According to the fault feature matching rule, each fault sensitive feature in the vibration fault sensitive feature set is mapped to the corresponding candidate fault type. The feature matching results of each candidate fault type are weighted using the fault response consistency coefficient to obtain the confidence value of the candidate fault type. The confidence values ​​of the candidate fault types are arranged according to the vibration measurement points and the candidate fault types to construct a fault type confidence evaluation matrix.

[0012] In some embodiments, outputting the rotating machinery vibration fault type and corresponding measurement point location based on the fault type confidence assessment matrix specifically includes: Candidate fault types with confidence values ​​higher than a predetermined diagnostic threshold are selected from the fault type confidence assessment matrix; Determine the confidence contribution distribution of the candidate fault types at each vibration measurement point; The dominant fault response measurement points are determined based on the confidence value contribution distribution and the measurement point layout information. The candidate fault type is taken as the rotating machinery vibration fault type, and the installation position of the measuring point corresponding to the dominant measuring point of the fault response is output as the corresponding measuring point position.

[0013] Secondly, this application provides an intelligent diagnostic system for vibration faults in rotating machinery, which includes a data acquisition module, a processing module, and an execution module, wherein... The acquisition module is used to initiate vibration fault monitoring of rotating machinery under multiple speed conditions, and simultaneously acquire the vibration signals, speed pulse signals and measurement point layout information of each vibration measurement point to obtain vibration data related to the operating conditions. The processing module is used to perform speed synchronous resampling and order spectrum transformation on the vibration data associated with the working conditions, and generate speed normalized vibration spectrum corresponding to each speed working condition. The processing module is also used to extract the rotational frequency order energy distribution, bearing characteristic order energy distribution and modulation sideband interval features based on the rotational speed normalized vibration spectrum, and to construct a vibration fault sensitive feature set under each rotational speed condition. The processing module is also used to determine the frequency synchronization phase reference between each speed condition based on the speed pulse signal, and to perform cross-condition phase consistency correlation on the vibration fault sensitive feature set according to the frequency synchronization phase reference to obtain the fault response consistency coefficient. The execution module is used to construct a fault type confidence assessment matrix based on the vibration fault sensitive feature set, the fault response consistency coefficient and the measurement point layout information, and output the rotating machinery vibration fault type and the corresponding measurement point location based on the fault type confidence assessment matrix.

[0014] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: In the process of diagnosing vibration faults in rotating machinery, this application firstly forms condition-correlated vibration data by combining vibration signals from measuring points, speed pulse signals, and measuring point layout information, thus providing a unified data foundation for vibration response, rotational phase, and measuring point spatial location. Secondly, through synchronous resampling of speed and order spectrum transformation, the frequency characteristics that drift with speed changes are converted into order characteristics that match the rotational motion, enabling the same mechanical fault to have comparable characteristic locations under different speed conditions. Thirdly, a set of sensitive features for vibration faults is constructed by using the energy distribution of frequency multiples, the energy distribution of bearing characteristic orders, and the modulation sideband interval characteristics, allowing rotor-type faults, rolling bearing-type faults, and gear meshing-type faults to be expressed within the same diagnostic framework. Finally, a fault response consistency coefficient is obtained through cross-condition phase consistency correlation, and together with the measuring point layout information, a fault type confidence assessment matrix is ​​constructed, which can reduce the impact of speed fluctuations and local interference on the diagnostic results, thereby achieving stable identification and measuring point location of vibration faults in rotating machinery under variable speed conditions. Attached Figure Description

[0015] Figure 1 This is an exemplary flowchart of an intelligent diagnostic method for vibration faults in rotating machinery, according to some embodiments of this application. Figure 2 This is a schematic diagram of the structure of an intelligent diagnostic system for vibration faults in rotating machinery, as shown in some embodiments of this application. Figure 3 This is a schematic diagram of the structure of a computer device for implementing an intelligent diagnosis method for vibration faults in rotating machinery, according to some embodiments of this application. Detailed Implementation

[0016] To better understand the above technical solutions, the following will provide a detailed description of the technical solutions in conjunction with the accompanying drawings and specific implementation methods.

[0017] refer to Figure 1 The figure is an exemplary flowchart of an intelligent diagnostic method for vibration faults in rotating machinery according to some embodiments of this application. The method mainly includes the following steps.

[0018] In step S101, the vibration fault monitoring of rotating machinery is started under multiple speed conditions. The vibration signals, speed pulse signals and measurement point layout information of each vibration measuring point are collected simultaneously to obtain vibration data related to the operating conditions.

[0019] It should be noted that the vibration response of rotating machinery essentially originates from the propagation of periodic excitation of rotating components, rolling contact impact excitation, and gear meshing modulation excitation within the structure. If only the vibration amplitude is collected without recording the rotational speed pulse signal, the drift in vibration frequency caused by changes in rotational speed cannot be correctly interpreted; if only a single measuring point signal is collected without recording the measuring point layout information, it is difficult to determine whether the fault response mainly originates from the bearing end, the transmission end, or the housing support. Therefore, this application first binds the measuring point vibration signal, the rotational speed pulse signal, and the measuring point layout information into operating condition-related vibration data.

[0020] In practical implementation, vibration sensors can be installed on the drive-end bearing housing, non-drive-end bearing housing, gearbox housing, and near the support of the coupling in rotating machinery. Vibration sensors can collect radial or axial acceleration signals, while speed sensors can collect speed pulse signals corresponding one-to-one with the rotating shaft. The acquisition system uniformly records the installation location, sampling time, vibration amplitude, and speed pulse time of each vibration measuring point, and divides the system into multiple speed conditions according to the speed stabilization section, acceleration section, and deceleration section, thereby obtaining condition-correlated vibration data.

[0021] In step S102, the speed synchronous resampling and order spectrum transformation are performed on the vibration data associated with the operating conditions to generate the speed normalized vibration spectrum corresponding to each speed operating condition.

[0022] In rotating machinery, many fault excitations have a fixed relationship with the angular position of the rotating shaft. For example, rotor imbalance is usually amplified near the first-order rotational frequency; rotor misalignment can lead to amplification at the second-order rotational frequency and related multiples; local damage to rolling bearings will generate impact energy near characteristic orders related to bearing geometry and rotational speed; abnormal gear meshing will form modulation structures near the meshing order and its sidebands. Therefore, this application does not directly compare vibration signals under fixed frequency coordinates, but uses rotational speed pulse signals to calculate instantaneous rotation angles and converts time-domain vibration signals into angular-domain vibration sequences.

[0023] In some embodiments, the instantaneous rotation angle of the rotating shaft at each sampling moment can be calculated based on the rotational speed pulse signal. The vibration signal at the measuring point is then resampled at equal rotational speed intervals to obtain an angular domain vibration sequence. Subsequently, an order spectrum transformation is performed on the angular domain vibration sequence to obtain an order spectrum with the rotational frequency multiple as the horizontal coordinate. Since the order coordinate has been normalized to the rotational frequency, the same rotor or bearing fault can maintain a relatively stable order position under different rotational speed conditions, thereby forming a rotational speed-normalized vibration spectrum.

[0024] In step S103, the rotational frequency order energy distribution, bearing characteristic order energy distribution, and modulation sideband interval features are extracted from the normalized vibration spectrum to construct a vibration fault sensitive feature set under each rotational speed condition.

[0025] Among them, the frequency multiplication order energy distribution is used to characterize periodic excitation synchronized with the rotating shaft; the bearing characteristic order energy distribution is used to characterize periodic impacts caused by damage to the outer ring, inner ring, rolling elements, or cage of the rolling bearing; and the modulation sideband interval feature is used to characterize amplitude modulation phenomena caused by abnormal gear meshing, bearing impact modulation, or rotor loosening. These three types of features correspond to different physical excitation sources, avoiding misjudgments caused by judging fault types solely based on a single peak amplitude.

[0026] In practice, the energy distributions corresponding to the first, second, and third rotational frequency orders can be extracted from the normalized vibration spectrum to form a rotational frequency order energy distribution. Then, based on the number of rolling elements, rolling element diameter, pitch circle diameter, and contact angle of the rolling bearing, the bearing characteristic orders corresponding to the outer ring, inner ring, rolling elements, and cage are determined, and the energy distributions near the corresponding orders are extracted. Subsequently, sideband structures appearing at intervals of rotational frequency orders are identified on both sides of the gear meshing order or bearing characteristic order, and the sideband interval, sideband amplitude ratio, and sideband symmetry are extracted, thereby forming a vibration fault sensitive feature set.

[0027] In step S104, the rotational frequency synchronization phase reference between each rotational speed condition is determined based on the rotational speed pulse signal. The vibration fault sensitive feature set is then correlated across operating conditions according to the rotational frequency synchronization phase reference to obtain the fault response consistency coefficient.

[0028] It should be noted that real mechanical faults generally exhibit a certain degree of phase repetition with rotational motion, while random shocks, external disturbances, and short-term operating condition fluctuations typically cannot maintain a stable order position and phase relationship across multiple speed conditions. Therefore, this application further utilizes speed pulse signals to determine the zero-phase position of the rotating shaft, comparing the vibration fault sensitivity characteristics under different operating conditions under a unified frequency synchronous phase reference.

[0029] In some embodiments, the vibration fault sensitive feature sets at the same measuring point under different speed conditions can be phase aligned according to the rotational frequency synchronization phase reference, and the order position stability, energy change consistency, and phase repeatability of the same fault sensitive feature under multiple speed conditions can be calculated. If a feature is stable in order position, has continuity in energy change with load change, and has a repeating pattern in phase, then the fault response consistency coefficient corresponding to that feature is high; conversely, if the feature only appears briefly in a single operating condition, the corresponding coefficient is low.

[0030] In step S105, a fault type confidence assessment matrix is ​​constructed based on the vibration fault sensitive feature set, the fault response consistency coefficient, and the measurement point layout information. Based on the fault type confidence assessment matrix, the rotating machinery vibration fault type and the corresponding measurement point location are output.

[0031] In practical implementation, fault characteristic matching rules can be established for rotor imbalance, rotor misalignment, rolling bearing damage, and gear meshing abnormalities. For rotor imbalance, the focus is on the stable enhancement of energy at the first frequency order in adjacent speed conditions; for rotor misalignment, the focus is on the energy response at the second frequency order and related multiples; for rolling bearing damage, the focus is on the bearing characteristic order and its modulation sidebands; for gear meshing abnormalities, the focus is on the meshing order and the sideband structure spaced at frequency orders.

[0032] In addition, after mapping the vibration fault sensitive feature set to the candidate fault type, the feature matching results of each candidate fault type can be weighted using the fault response consistency coefficient to obtain the confidence value of the candidate fault type. The confidence values ​​of the candidate fault type are then arranged according to the vibration measurement point and the candidate fault type to form a fault type confidence evaluation matrix.

[0033] In practice, candidate fault types with confidence values ​​higher than a predetermined diagnostic threshold are selected from the fault type confidence assessment matrix, their confidence value contribution distribution at each vibration measurement point is determined, and the dominant measurement point of the fault response is determined in combination with the measurement point layout information. Finally, the vibration fault type of rotating machinery and the corresponding measurement point location are output.

[0034] This application's solution conforms to the basic mechanical principles of vibration diagnosis in rotating machinery. Rotational speed synchronous resampling, based on the rotational shaft angular position, eliminates frequency coordinate drift caused by variable speed. Order spectrum transformation converts frequency-dependent fault excitations to stable order coordinates. The bearing characteristic order is jointly determined by the bearing geometry and rotational speed, while the modulation sideband is caused by periodic impacts or meshing abnormalities. Cross-condition phase consistency correlation utilizes the repeatability of real mechanical faults to distinguish random disturbances. Therefore, the fault type confidence assessment matrix provides a physically constrained data foundation for fault type identification and measurement point location.

[0035] Furthermore, in another aspect of this application, in some embodiments, this application provides an intelligent diagnostic system for vibration faults in rotating machinery, referring to... Figure 2 The system includes a data acquisition module 101, a processing module 102, and an execution module 103, which are described below: The acquisition module 101 is used to start the vibration fault monitoring of rotating machinery under multiple speed conditions, and simultaneously acquire the vibration signal of each vibration measuring point, the speed pulse signal and the measuring point layout information to obtain the vibration data related to the operating conditions. Processing module 102 is used to perform speed synchronous resampling and order spectrum transformation on the vibration data associated with the working conditions, and generate speed normalized vibration spectrum corresponding to each speed working condition. In specific implementation, the processing module 102 is also used to extract the frequency multiple order energy distribution, bearing characteristic order energy distribution and modulation sideband interval features based on the normalized vibration spectrum of the rotational speed, and to construct a vibration fault sensitive feature set under each rotational speed condition; In addition, the processing module 102 is also used to determine the frequency synchronization phase reference between each speed condition based on the speed pulse signal, and to perform cross-condition phase consistency correlation on the vibration fault sensitive feature set according to the frequency synchronization phase reference to obtain the fault response consistency coefficient. The execution module 103 is used to construct a fault type confidence assessment matrix based on the vibration fault sensitive feature set, the fault response consistency coefficient and the measurement point layout information, and output the rotating machinery vibration fault type and the corresponding measurement point location based on the fault type confidence assessment matrix.

[0036] In addition, this application also provides a computer device, which may be a server, and its internal structure diagram may be as follows: Figure 3 As shown, the computer device includes a processor, memory, and a network interface connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and a database. The internal memory provides the environment for the operation of the operating system and computer programs stored in the non-volatile storage media. The database stores operating condition-related vibration data, speed-normalized order spectrum sets, vibration fault sensitivity feature sets, and fault diagnosis results. The network interface communicates with external terminals via a network. When executed by the processor, the computer program implements an intelligent method for diagnosing vibration faults in rotating machinery.

[0037] Those skilled in the art will understand that Figure 3 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0038] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

Claims

1. A method for intelligent diagnosis of vibration faults in rotating machinery, characterized in that, Includes the following steps: The vibration fault monitoring of rotating machinery is initiated under multiple speed conditions. The vibration signals, speed pulse signals and measurement point layout information of each vibration measuring point are collected simultaneously to obtain vibration data related to the operating conditions. The speed synchronous resampling and order spectrum transformation are performed on the vibration data associated with the working conditions to generate the speed normalized vibration spectrum corresponding to each speed working condition. Based on the normalized vibration spectrum of the rotational speed, the energy distribution of the rotational frequency multiple order, the energy distribution of the bearing characteristic order, and the modulation sideband interval features are extracted to construct a set of vibration fault sensitive features under each rotational speed condition. Based on the speed pulse signal, the frequency synchronization phase reference between each speed condition is determined. Based on the frequency synchronization phase reference, the vibration fault sensitive feature set is correlated across operating conditions to obtain the fault response consistency coefficient. Based on the vibration fault sensitive feature set, the fault response consistency coefficient, and the measurement point layout information, a fault type confidence evaluation matrix is ​​constructed, and the rotating machinery vibration fault type and corresponding measurement point location are output based on the fault type confidence evaluation matrix.

2. The method according to claim 1, characterized in that, Vibration fault monitoring of rotating machinery is initiated under multiple speed conditions. Vibration signals, speed pulse signals, and measurement point layout information are collected simultaneously from each vibration measuring point to obtain vibration data related to the operating conditions, including: Vibration sensors are installed in the bearing housings, housing supports, and transmission ends of rotating machinery, and the installation positions of the corresponding measuring points of each vibration sensor are recorded. A rotational speed pulse signal synchronized with the rotating shaft is acquired by a rotational speed sensor, and the rotational speed pulse signal is used as the synchronization reference for the vibration signal of each measuring point. Multiple speed conditions are divided into stable speed range, acceleration range, and deceleration range. The vibration signals, speed pulse signals, and measurement point layout information of each speed condition are time-scaled to form condition-related vibration data.

3. The method according to claim 1, characterized in that, The process of resampling the vibration data associated with the aforementioned operating conditions and performing order spectrum transformation to generate normalized vibration spectra for each operating condition specifically includes: The instantaneous rotation angle of the rotating shaft at each sampling moment is calculated based on the rotational speed pulse signal; The vibration signal at the measuring point is resampled at equal angular intervals to obtain the angular domain vibration sequence; The angular domain vibration sequence is subjected to order spectrum transformation to obtain an order spectrum with the frequency multiple as the horizontal axis and the vibration amplitude as the vertical axis. The order spectra are arranged according to the rotational speed conditions and vibration measurement points to generate a rotational speed-normalized vibration spectrum.

4. The method according to claim 1, characterized in that, Based on the normalized vibration spectrum, the rotational frequency order energy distribution, bearing characteristic order energy distribution, and modulation sideband spacing features are extracted to construct a vibration fault sensitive feature set for each rotational speed condition. Specifically, this includes: The energy distributions corresponding to the first, second, and third rotational frequency orders are extracted from the normalized vibration spectrum to obtain the rotational frequency order energy distributions. Based on the geometric parameters and rotational speed pulse signals of the rolling bearing in rotating machinery, the bearing characteristic order corresponding to the outer ring, inner ring, rolling elements and cage is determined, and the energy distribution of the corresponding order is extracted to obtain the bearing characteristic order energy distribution. Identify the sideband structures that appear on both sides of the gear meshing order or bearing characteristic order with the frequency order as the interval, extract the sideband interval, sideband amplitude ratio and sideband symmetry to obtain the modulation sideband interval characteristics; The frequency multiplication order energy distribution, the bearing characteristic order energy distribution, and the modulation sideband interval feature are combined to construct a vibration fault sensitive feature set under each speed condition.

5. The method according to claim 1, characterized in that, Based on the speed pulse signal, a frequency synchronization phase reference is determined between different speed conditions. Then, based on the frequency synchronization phase reference, a cross-condition phase consistency correlation is performed on the vibration fault sensitive feature set to obtain the fault response consistency coefficient, specifically including: The zero-phase position of the rotating shaft under each speed condition is determined based on the speed pulse signal, and the frequency synchronization phase reference is obtained. Phase alignment is performed on the vibration fault sensitive feature set of the same measuring point under different speed conditions according to the aforementioned frequency synchronization phase reference. Calculate the order position stability, energy change consistency, and phase repeatability of the same fault-sensitive features under multiple speed conditions after phase alignment; The fault response consistency coefficient is determined based on the order position stability, the energy change consistency, and the phase repeatability.

6. The method according to claim 1, characterized in that, The construction of a fault type confidence assessment matrix based on the vibration fault sensitive feature set, the fault response consistency coefficient, and the measuring point layout information specifically includes: Establish fault feature matching rules corresponding to rotor imbalance, rotor misalignment, rolling bearing damage, and abnormal gear meshing; According to the fault feature matching rule, each fault sensitive feature in the vibration fault sensitive feature set is mapped to the corresponding candidate fault type. The feature matching results of each candidate fault type are weighted using the fault response consistency coefficient to obtain the confidence value of the candidate fault type. The confidence values ​​of the candidate fault types are arranged according to the vibration measurement points and the candidate fault types to construct a fault type confidence evaluation matrix.

7. The method according to claim 1, characterized in that, Based on the aforementioned fault type confidence assessment matrix, the output of rotating machinery vibration fault types and corresponding measurement point locations specifically includes: Candidate fault types with confidence values ​​higher than a predetermined diagnostic threshold are selected from the fault type confidence assessment matrix; Determine the confidence contribution distribution of the candidate fault types at each vibration measurement point; The dominant fault response measurement points are determined based on the confidence value contribution distribution and the measurement point layout information. The candidate fault type is taken as the rotating machinery vibration fault type, and the installation position of the measuring point corresponding to the dominant measuring point of the fault response is output as the corresponding measuring point position.

8. An intelligent diagnostic system for vibration faults in rotating machinery, characterized in that, include: The acquisition module is used to initiate vibration fault monitoring of rotating machinery under multiple speed conditions, and simultaneously acquire the vibration signals, speed pulse signals and measurement point layout information of each vibration measurement point to obtain vibration data related to the operating conditions. The processing module is used to perform speed synchronous resampling and order spectrum transformation on the vibration data associated with the operating conditions, and generate speed normalized vibration spectrum diagrams corresponding to each speed operating condition. The processing module is also used to extract the rotational frequency order energy distribution, bearing characteristic order energy distribution and modulation sideband interval features based on the rotational speed normalized vibration spectrum, and to construct a vibration fault sensitive feature set under each rotational speed condition. The processing module is also used to determine the frequency synchronization phase reference between each speed condition based on the speed pulse signal, and to perform cross-condition phase consistency correlation on the vibration fault sensitive feature set according to the frequency synchronization phase reference to obtain the fault response consistency coefficient. The execution module is used to construct a fault type confidence assessment matrix based on the vibration fault sensitive feature set, the fault response consistency coefficient and the measurement point layout information, and output the rotating machinery vibration fault type and the corresponding measurement point location based on the fault type confidence assessment matrix.

9. A computer device, characterized in that, The computer device includes a memory and a processor, the memory storing code, and the processor being configured to retrieve the code and execute the intelligent diagnostic method for vibration faults in rotating machinery as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the intelligent diagnostic method for vibration faults of rotating machinery as described in any one of claims 1 to 7.

Images