WO2021035638A1

WO2021035638A1 - Fault diagnosis method and system for rotary mechanical device, and storage medium

Info

Publication number: WO2021035638A1
Application number: PCT/CN2019/103418
Authority: WO
Inventors: 徐楠; 邱周钦; 苏明; 沙永健; 叶晨; 王春光
Original assignee: 亿可能源科技（上海）有限公司
Priority date: 2019-08-29
Filing date: 2019-08-29
Publication date: 2021-03-04
Also published as: CN113632026A

Abstract

A fault diagnosis method for a rotary mechanical device, comprising: acquiring operational data of at least one detection point in a rotary mechanical device (S110); performing an abnormality detection analysis on at least one type of abnormality of the detection point by using the operational data, so as to obtain an analysis result corresponding to each type of abnormality (S120); and using the obtained at least one analysis result to perform diagnosis processing on at least one kind of fault in the rotary mechanical device, so as to output a corresponding fault diagnosis result (S130). By continuously performing dimension reduction processing on acquired data and performing feature extraction on the data by presetting key indexes, accuracy is ensured in the case of a small number of data samples.

Description

Rotating machinery equipment fault diagnosis method, system and storage medium

Technical field

This application relates to the technical field of fault detection, and in particular to a method, system and storage medium for fault diagnosis of rotating machinery equipment.

Background technique

When the rotating mechanical equipment in industrial production fails, it will cause abnormal equipment vibration and other phenomena. The possible faults of the equipment can be diagnosed by analyzing the vibration of the equipment. However, a large number of data samples are usually required as input when constructing a fault diagnosis model. However, the number of samples in actual industrial production is limited, and the current fault diagnosis method in the prior art has a low accuracy in the diagnosis of abnormal phenomena.

Summary of the invention

In view of the above-mentioned shortcomings of the prior art, the purpose of this application is to provide a fault diagnosis method, system and storage medium for rotating machinery equipment, which are used to solve the problem of the failure diagnosis method in the prior art because it cannot obtain a large number of sample inputs for training Diagnose problems with low accuracy.

In order to achieve the above and other related purposes, the first aspect of the present application provides a method for diagnosing faults of rotating machinery equipment, which includes the following steps: acquiring operating data on at least one detection point in a rotating machinery equipment; using the operating data pair At least one abnormality type of the detection point is respectively subjected to abnormality detection and analysis to obtain an analysis result corresponding to each abnormality type; and at least one type of fault in the rotating mechanical equipment is diagnosed by using the obtained at least one analysis result To output the corresponding fault diagnosis results.

In some embodiments of the first aspect of the present application, the abnormality type includes at least one of the following abnormalities reflected by the corresponding detection points during the operation of the rotating mechanical equipment based on working conditions and process requirements: based on at least one spatial dimension The abnormal type is set separately for each vibration abnormality, the abnormal type is set based on the temperature abnormality, and the abnormal type is set separately based on each power abnormality.

In some implementations of the first aspect of the present application, the operating data is used to generate detection data and reference data; the operating data is used to perform abnormality detection and analysis on at least one abnormality type of the detection point. The step includes: performing feature extraction on the acquired detection data based on the reference data of the detection point during the normal operation of the rotating mechanical equipment to obtain at least one feature element for detecting at least one abnormal type of the detection point ; Analyze each feature element corresponding to each abnormal type to obtain the analysis result of the corresponding abnormal type.

In some implementations of the first aspect of the present application, the method further includes the following step: further data processing is performed on the obtained at least one characteristic element, so as to analyze the at least one characteristic element after further data processing, so as to obtain the corresponding abnormality type The results of the analysis.

In some implementations of the first aspect of the present application, the reference data includes at least one or more of the following: the operating condition data of the rotating mechanical equipment is extracted from the process data in the acquired operating data The data is extracted from the mechanical vibration data in the acquired operating data.

In some implementations of the first aspect of the present application, the method further includes the following step: analyzing the fundamental frequency data in the reference data corresponding to the normal operation of the rotating mechanical equipment from the acquired operating data.

In some implementations of the first aspect of the present application, the acquired operating data includes mechanical vibration data; the feature extraction is performed on the acquired detection data based on the reference data of the detection points during the normal operation of the rotating mechanical equipment , The step of obtaining at least one feature element for detecting at least one abnormal type of the detection point includes: extracting at least one of the acquired mechanical vibration data of the frequency spectrum related to the fundamental frequency data in the reference data of the detection point A frequency feature element for detecting at least one abnormal type of the detection point.

In some implementations of the first aspect of the present application, the acquired operating data includes process data and/or operating condition data; the method further includes the following steps: analyzing the acquired process data and/or operating condition data, In order to obtain the reference data in the reference data of the detection point during the normal operation of the rotating mechanical equipment.

In some implementations of the first aspect of the present application, the feature extraction is performed on the acquired detection data based on the reference data of the detection point during the normal operation of the rotating mechanical equipment, so as to obtain the detection point for detecting the detection point. The step of obtaining at least one characteristic element of at least one abnormal type of the at least one abnormal type includes: obtaining at least one deviation characteristic element based on the deviation between the reference data in the reference data during normal operation of the rotating mechanical equipment and the acquired detection data, for use in At least one abnormal type of the detection point is detected.

In some implementations of the first aspect of the present application, the step of using the obtained at least one analysis result to diagnose at least one fault in the rotating mechanical equipment to output a corresponding fault diagnosis result includes : Present the obtained fault diagnosis result on the display interface of the control system of the rotating mechanical equipment.

The second aspect of the present application also provides a server, including: an interface unit for data communication with sensors in at least one dimension at the detection point of the rotating mechanical equipment; a storage unit for storing at least one program; and processing The unit is used to call the at least one program to coordinate the execution of the interface unit and the storage unit and implement the method for diagnosing the fault of a rotating machinery device as described in any one of the first aspect of the present application.

The third aspect of the present application also provides a first fault diagnosis system for rotating machinery equipment, including: the server as described in the second aspect of the present application; and detection devices configured at each detection point of the rotating machinery equipment, and The server communication connection is used to provide the operating data of each detection point.

The fourth aspect of the present application also provides a computer-readable storage medium that stores at least one program, and the at least one program executes and implements the rotating machine as described in any one of the first aspect of the present application when called. Equipment fault diagnosis method.

The fifth aspect of the present application also provides a second fault diagnosis system for rotating machinery equipment, including: a data acquisition module to obtain operating data on at least one detection point in a rotating machinery equipment; and a data processing module to use The operating data performs abnormality detection and analysis on at least one abnormality type of the detection point to obtain an analysis result corresponding to each abnormality type, and uses the obtained at least one analysis result to analyze at least one of the rotating mechanical equipment A fault is diagnosed and processed to output the corresponding fault diagnosis result.

In some implementations of the fifth aspect of the present application, the abnormal type includes at least one of the following abnormalities reflected by the corresponding detection points during the operation of the rotating mechanical equipment based on working conditions and process requirements: based on at least one spatial dimension The abnormal type is set separately for each vibration abnormality, the abnormal type is set based on the temperature abnormality, and the abnormal type is set separately based on each power abnormality.

In some implementations of the fifth aspect of the present application, the operating data is used to generate detection data and reference data; the data processing module is based on the reference of the detection point during the normal operation of the rotating mechanical equipment The data performs feature extraction on the acquired detection data to obtain at least one feature element for detecting at least one abnormality type of the detection point; and analyzes each feature element corresponding to each abnormality type to obtain the corresponding abnormality Type of analysis results.

In some implementation manners of the fifth aspect of the present application, the data processing module performs further data processing on at least one feature element obtained, so as to analyze the at least one feature element after further data processing, so as to obtain the corresponding abnormality. Type of analysis results.

In some implementations of the fifth aspect of the present application, the reference data includes at least one or more of the following: the operating condition data of the rotating mechanical equipment is extracted from the process data in the acquired operating data The data is extracted from the mechanical vibration data in the acquired operating data.

In some implementations of the fifth aspect of the present application, the data processing module further analyzes the fundamental frequency data in the reference data corresponding to the normal operation of the rotating mechanical equipment from the acquired operating data.

In some implementations of the fifth aspect of the present application, the acquired operating data includes mechanical vibration data, and the data processing module extracts the frequency spectrum of the acquired mechanical vibration data and the reference data of the detection point. At least one frequency feature element related to the fundamental frequency data is used to detect at least one abnormal type of the detection point.

In some implementations of the fifth aspect of the present application, the acquired operating data includes process data and/or operating condition data; the data processing module also analyzes the acquired process data and/or operating condition data to obtain The reference data in the reference data of the detection point during the normal operation of the rotating mechanical equipment.

In some implementations of the fifth aspect of the present application, the data processing module obtains at least one deviation characteristic based on the deviation between the reference data in the reference data during the normal operation of the rotating mechanical equipment and the acquired detection data Yuan for detecting at least one abnormal type of the detection point.

In some implementations of the fifth aspect of the present application, a fault display device is further included to present the obtained fault diagnosis result.

The sixth aspect of the present application also provides a management system for rotating machinery equipment, including: detection devices arranged at each detection point of the rotating machinery equipment for providing operating data of each detection point; a control system for the rotating machinery equipment, and each The detection device is connected with data to collect and forward each of the operating data; the server as described in the second aspect of the application is connected to the control system in communication, and is used to receive each of the operating data and based on the received operating data. The data executes the corresponding fault diagnosis method.

As mentioned above, the fault diagnosis method, system and storage medium of rotating machinery equipment of the present application have the following beneficial effects: the present application continuously reduces the dimensionality of the acquired data, and presets key indicators to extract features of the data. Guarantee accuracy in the case of a small number of data samples. Secondly, this application can flexibly expand the process parameters. The newly added parameters will not affect the machine learning model of the existing parameters, and there is no need to retrain the existing models. You only need to create a new learning model for the newly added parameters and perform the comprehensive diagnosis model. Just add the relationship model between the new parameter and the existing parameter. In addition, the three-layer model structure of this application includes both the characteristic mechanism model and the machine learning model. It also combines the location and type of detection points with an integrated diagnosis model to integrate management experience with machine learning algorithms to ensure fault diagnosis. The accuracy of the results.

Description of the drawings

Figure 1 shows a schematic diagram of a fault diagnosis method for rotating machinery in this application.

Figure 2 shows a schematic diagram of an embodiment of setting a detection point on the fan in this application.

3a to 3b show schematic diagrams of an embodiment of abnormal detection and analysis performed by the fault diagnosis system in this application.

Figures 4a to 4b show schematic diagrams of converting the time-domain waveforms in Figures 3a and 3b into frequency-domain spectra in this application.

FIG. 5 shows a schematic diagram of an embodiment of using operating data to perform anomaly detection and analysis in this application.

FIG. 6 shows a schematic diagram of an embodiment of the fault diagnosis process in this application.

FIG. 7 shows a schematic diagram of a structural embodiment of a server in this application.

detailed description

The following specific examples illustrate the implementation of this application. Those familiar with this technology can easily understand the other advantages and effects of this application from the content disclosed in this specification.

Although the terms first, second, etc. are used herein to describe various elements in some instances, these elements should not be limited by these terms. These terms are only used to distinguish one element from another element. For example, the first fault diagnosis system may be referred to as the second fault diagnosis system, and similarly, the second fault diagnosis system may be referred to as the first fault diagnosis system without departing from the scope of the various described embodiments. The first fault diagnosis system and the second fault diagnosis system are both describing a fault diagnosis system, but unless the context clearly indicates otherwise, they are not the same fault diagnosis system.

Furthermore, as used herein, the singular forms "a", "an" and "the" are intended to also include the plural forms, unless the context dictates to the contrary. It should be further understood that the terms "comprising" and "including" indicate the presence of the described features, steps, operations, elements, components, items, types, and/or groups, but do not exclude one or more other features, steps, operations, The existence, appearance or addition of elements, components, items, categories, and/or groups. The terms "or" and "and/or" used herein are interpreted as inclusive or mean any one or any combination. Therefore, "A, B or C" or "A, B and/or C" means "any of the following: A; B; C; A and B; A and C; B and C; A, B and C" . An exception to this definition will only occur when the combination of elements, functions, steps, or operations is inherently mutually exclusive in some way.

As mentioned in the background technology section, in industrial production, data such as vibration and temperature of rotating machinery equipment during normal operation and abnormal operation will change. By analyzing the fault characteristics of different faults, a machine learning model can be constructed, so that the data of the detected self-rotating mechanical equipment can be input into the model to perform fault diagnosis. However, in some embodiments, a large amount of data samples are needed to analyze various possible failures of rotating machinery through a model with comprehensive analysis capabilities. For example, in an experimental scenario, the sample size can be sufficient by artificially destroying various parts of the rotating mechanical equipment and obtaining data, so as to meet the requirement of a large number of sample inputs when building a machine learning model. However, rotating machinery and equipment in actual industrial production are expensive. Obviously, this destructive method cannot be applied to actual industrial production. If there are insufficient samples, it will affect the accuracy of the analysis results. Therefore, although these machine models can be used to diagnose faults in rotating machinery in theory, the accuracy of diagnosis in the actual production application process will be greatly reduced when a large number of data samples are lacking.

To this end, the present application provides a fault diagnosis method for rotating machinery equipment, and the fault diagnosis method is mainly executed by a fault diagnosis system. Wherein, the fault diagnosis system can be executed by the server. Wherein, the fault diagnosis system may be a software system configured on the server side. The server includes but is not limited to a single server, server cluster, distributed server cluster, cloud server, etc. Here, according to the actual design, the server where the fault diagnosis system is located can be configured in the server equipment located in the machine room on the side of the rotating machinery equipment. For example, the single server or server cluster where the fault diagnosis system is located is located in the machine room on the side of the rotating mechanical equipment. Or, according to actual design, the fault diagnosis system can also be configured in a cloud server provided by a cloud provider. Wherein, the cloud server includes a public cloud (Public Cloud) server and a private cloud (Private Cloud) server, where the public or private cloud server includes Software-as-a-Service (Software-as-a-Service, SaaS) ), Platform-as-a-Service (Platform-as-a-Service, PaaS) and Infrastructure-as-a-Service (Infrastructure-as-a-Service, IaaS), etc. The private cloud service terminal is, for example, Alibaba Cloud Computing Service Platform, Amazon Cloud Computing Service Platform, Baidu Cloud Computing Platform, Tencent Cloud Computing Platform, and so on.

Here, according to the actual design of the fault diagnosis system, the server can be communicatively connected with the control system of the rotating machinery equipment. Wherein, the control system is a software system running on the computer equipment, which collects the operating data detected by the sensors arranged at the detection points of the rotating machinery equipment by means of the computer equipment, obtains equipment parameters of the rotating machinery equipment, and reports to the Rotating machinery equipment outputs control commands, etc. For example, vibration sensors, temperature sensors, current sensors, etc. are distributed on rotating mechanical equipment, and the control system obtains data provided by any one or more of the aforementioned sensors and transmits it to the fault diagnosis system through a communication network. Or, according to the actual design of the fault diagnosis system, the server can also directly communicate with the rotating machinery equipment and the sensors arranged at each detection point of the rotating machinery equipment, so as to collect the sensors arranged at each detection point of the rotating machinery equipment Detected operating data, obtain equipment parameters of rotating machinery and equipment, etc.

It should be understood that due to the diverse types of failures of rotating machinery and equipment, each type of failure exhibits different abnormalities. Therefore, in some embodiments, although a diagnosis result can be obtained through one detection point, in order to ensure the correct rate of diagnosis , Multiple detection points can be set on rotating machinery and equipment, and a more credible diagnosis result can be obtained through comprehensive analysis of multiple detection points. Among them, the setting position of the detection point can be set according to actual needs. Taking a fan as an example, please refer to FIG. 2, which shows a schematic diagram of an embodiment of setting detection points on the fan in this application. As shown in the figure, the motor 102 drives the fan 100 to rotate through a transmission component. In this embodiment, A vibration sensor is provided in the three directions of the fan bearing 101 (that is, the axial direction, the vertical direction, and the horizontal direction), so that mechanical vibration data in 6 directions can be obtained. If the fan blade fails, it will increase the overall shaking of the equipment, causing the mechanical vibration data on both bearings to be abnormal; and if one of the bearings fails, the mechanical vibration data of the failed bearing is abnormal, and the mechanical vibration data of the other bearing is abnormal. There is no obvious abnormality in the vibration data. For another example, when the rotating mechanical equipment is a generator set, one temperature sensor may be provided on each bearing shell of each bearing of the generator set. As a result, it is possible to more accurately diagnose the malfunction of the rotating machine equipment.

Here, according to the selected detection point and the sensor type, the operating data obtained from the sensor includes at least one of the following: mechanical vibration data, process data, and working condition data. Wherein, the mechanical vibration data is the sensing data provided by a sensor that senses mechanical vibration, examples of which include but are not limited to: vibration velocity, vibration impact, vibration displacement, vibration acceleration, vibration spectrum, etc. of the detection point One or more data. The process data is one or more data related to the production process in the rotating mechanical equipment; wherein, the production process is usually the work and method of processing or processing various raw materials, materials, semi-finished products, etc. and turning them into finished products And technology. The rotating mechanical equipment is equipment in the process of performing processing or processing. The manufacturing method is set according to the physical and chemical characteristics of the finished product, and the rotating mechanical equipment is used to provide specific indicators that need to be achieved at some manufacturing stages in the manufacturing process of the finished product. For example, the fan is used to provide combustion-supporting wind force for the molten iron raw material to reach the temperature index of the molten iron raw material. For another example, the gear machine is used to provide mechanical energy for driving the mechanical arm to grab molten steel. To this end, according to the type of the rotating mechanical equipment and its function in the production process, the process data reflects the working data set to provide corresponding indicators during the manufacturing process of the rotating mechanical equipment performing the corresponding production process. Obtained by the sensor at the detection point or by the detection point itself. Wherein, according to the number and type of process data provided by the rotating mechanical equipment itself, and the number and type of sensors additionally equipped with rotation-related devices in the rotating mechanical equipment, the detection points may be one or more. Taking a fan as an example, multiple detection points can be configured on the fan, such as fan inlet, fan outlet, motor output power (or drive voltage, drive current, etc.), fan blades, etc. Taking the example of a fan and extending it to other rotating machinery and equipment, according to its role in the production process and through built-in or external sensors at each detection point, the acquired process data reflects the period during which the rotating machinery and equipment is performing the corresponding production process. At least one of the following provided by each detection point: temperature, current, speed, flow, air temperature, air door opening, inlet pressure, outlet pressure, etc. For example: when the rotating mechanical equipment is a fan and the detection point is located at the valve that controls the flow rate, the acquired process data may include the opening of the damper, flow rate, etc.; when the rotating mechanical equipment is a fan and the detection point is located at the fan When the transmission bearing is mounted, the acquired process data can include temperature and so on. Among them, the working condition refers to the working state of the equipment under the conditions that are directly related to its action. The rotating mechanical equipment usually has multiple working modes, and the energy generated in each working mode is different, and the energy required for actual production requirements can be met by setting the working mode of the rotating mechanical equipment. For example, the motor can provide different speeds for the wind turbine to meet the demand of wind power, and the electric energy needed in this process is met by the electric current. To this end, according to the type of the rotating mechanical equipment and the energy that it can provide during work, the working condition data reflects the working state of the rotating mechanical equipment during the working process, which can be measured by a sensor set at the detection point. The detection or detection point itself is obtained. For example, the fault diagnosis system obtains the current vibration frequency of the motor through the sensor and analyzes the current speed of the motor from this, and infers the current working mode of the motor from the speed, and obtains the working condition data. Wherein, according to the number and type of working condition data that the rotating mechanical equipment itself can provide, and the number and type of sensors additionally installed in the rotating mechanical equipment, the detection points may be one or more. Taking the water pump as an example, multiple detection points can be configured on the water pump, such as the water outlet, the output power of the motor, and the output shaft of the motor. Taking the example of a water pump and extending it to other rotating machinery and equipment, according to its role in production and through built-in or external sensors at each detection point, the working condition data can be provided by the rotating machinery and equipment in providing the required production Obtained from at least one of the following data provided by each detection point during the energy period: vibration, temperature, speed, flow, water temperature, valve opening, inlet pressure, outlet pressure, etc. For example: when the rotating mechanical equipment is a water pump and the detection point is located at the valve that controls the amount of water, the working condition data can be obtained through flow rate, etc.; when the rotating mechanical equipment is a water pump and the detection point is located on the motor output shaft of the pump When the working condition data can be obtained by rotating speed; when the rotating mechanical equipment is a water pump and the detection point is on the side of the motor driving the water pump, the working condition data can be obtained by electric current, etc. In some embodiments, the operating condition data may also be obtained by the rotating mechanical equipment itself or the management system of the rotating mechanical equipment, or the like. For example, the rotating mechanical equipment directly sends the current working mode of its own work to the fault diagnosis system. For another example, the management system of the rotating mechanical equipment obtains the current working mode of the rotating mechanical equipment and sends it to the fault diagnosis system.

Here, the application uses the fault diagnosis system to diagnose a rotating mechanical device as an example to describe the execution process of diagnosing the fault of the rotating mechanical device.

Please refer to FIG. 1, which shows a schematic diagram of the fault diagnosis method for rotating mechanical equipment in this application. As shown in the figure, in step S110, the fault diagnosis system obtains operating data on at least one detection point in a rotating mechanical equipment.

In some embodiments, the mechanical vibration data and process data are directly acquired by sensors. For example, various types of sensors are set on the detection points to obtain required various types of mechanical vibration data and/or process data. In other embodiments, some mechanical vibration data and/or process data can be calculated from the obtained related data. For example: the vibration acceleration obtained by the sensor can be used to calculate the vibration displacement and vibration velocity, without additional vibration displacement and vibration velocity sensors at the detection point.

Please continue to refer to FIG. 1, in step S120, the fault diagnosis system uses the operating data to perform an abnormality detection and analysis on at least one abnormality type of the detection point respectively to obtain an analysis result corresponding to each abnormality type. It should be understood that because the rotating mechanical equipment may have multiple faults at the same time in some cases, in order to ensure that the fault diagnosis system can diagnose multiple faults of the rotating mechanical equipment, it is necessary to use the acquired operating data to analyze The abnormal type corresponding to the corresponding detection point.

Wherein, the abnormality type includes at least one of the following abnormalities reflected by the corresponding detection points during the operation of the rotating mechanical equipment based on working conditions and process requirements: an abnormality type set separately based on each vibration abnormality in at least one spatial dimension , The abnormal type set based on temperature abnormality, and the abnormal type set separately based on each power abnormality. Wherein, the abnormal types separately set based on each vibration abnormality in at least one spatial dimension include, but are not limited to: impeller unbalance abnormality, coupling misalignment abnormality, rolling bearing vibration abnormality, sliding bearing vibration abnormality, etc.; The abnormal type set based on the abnormal temperature includes, but is not limited to, the abnormal bearing temperature change; the abnormal type set separately based on each power abnormality includes but is not limited to the abnormal change of the motor current.

Here, the fault diagnosis system is preset with an anomaly detection model set according to each type of abnormal performance corresponding to various faults at the detection point, wherein the anomaly detection model includes an abnormality detection model that is determined to belong to or based on input operating data. The algorithm for the possibility of not belonging to the corresponding abnormal type. The fault diagnosis system obtains an analysis result of whether the corresponding detection point shows a type of abnormality by inputting the received operating data of the detection point into the corresponding abnormality detection model. Among them, according to the different types of failures of rotating machinery and equipment reflected in the abnormal performance of each detection point, the abnormal types are examples but not limited to: impeller imbalance abnormality, coupling misalignment abnormality, rolling bearing vibration abnormality, sliding bearing vibration Abnormal, abnormal bearing temperature change, abnormal motor current change, etc. For example, the fault diagnosis system performs the abnormal detection and analysis of the impeller imbalance on the acquired operating data to obtain the analysis result of whether the detection point has the impeller imbalance abnormality.

Here, because each type of abnormality requires different data when performing abnormality detection and analysis, the fault diagnosis system also generates the acquired operating data corresponding to the data required for each abnormal type when performing abnormality detection and analysis. . Among them, in order to diagnose abnormalities, it is necessary to determine the reference data for the normal operation of the rotating mechanical equipment under the current working conditions during the execution of the current production process.

In some embodiments, the reference data is static data pre-stored locally. Here, the reference data includes the initial parameters and/or preset calibration parameters of the rotating mechanical equipment. Wherein, the initial parameters include, for example, inherent mechanical vibration data, rated process data, rated operating condition data, etc. of the rotating mechanical equipment when it leaves the factory or after maintenance. The preset calibration parameters are data obtained after the management personnel of the rotating mechanical equipment calibrate all or part of the mechanical vibration data, process data, etc. of the rotating mechanical equipment based on experience. For example: In some scenarios, due to geographic environment problems (such as insufficient installation foundation strength), the equipment will experience abnormal vibration, resulting in a difference between the mechanical vibration data of the rotating mechanical equipment when working in this environment and the mechanical vibration data at the factory. Larger, but because the abnormality of the mechanical vibration data is not caused by the failure of the rotating mechanical equipment, the management personnel of the rotating mechanical equipment can calibrate the mechanical vibration data of the rotating mechanical equipment based on experience. Use the calibrated mechanical vibration data as reference data.

In other embodiments, the operating data is used to generate detection data and reference data, that is, the reference data is dynamic data. Here, the fault diagnosis system analyzes the reference data corresponding to the normal operation of the rotating mechanical equipment from the acquired operating data. For example, the fault diagnosis system analyzes the mechanical vibration data in the acquired operating data to obtain the fundamental frequency of the rotating mechanical equipment, which is used as reference data, and at the same time, the fault diagnosis system also obtains the acquired mechanical vibration data. Convert it into a vibration frequency spectrum, and use the fundamental frequency to perform anomaly detection and analysis on the mechanical vibration data. In this embodiment, the operating data acquired by the fault diagnosis system is real-time data provided by the detection point, which directly or indirectly provides detection data and at least part of reference data for abnormality detection and analysis. Wherein, the reference data is used to represent normal data when the rotating mechanical equipment is operating normally under the current working conditions during the execution of the current production process. The detection data is used to represent the current data when the rotating mechanical equipment is currently running under the current working conditions during the execution of the current production process. The fault diagnosis system performs anomaly detection and analysis on the detection data based on the reference data and obtains corresponding analysis results. In some specific examples, according to the abnormal type, some of the obtained operating data can be directly used as reference data, for example, the torque of the drive motor.

It should be understood that the fault diagnosis system may obtain the reference data once every time the fault diagnosis is performed; it may also store the reference data obtained once in the storage medium, and call the reference data in the storage medium each time the fault diagnosis needs to be performed. Reference data.

It should be understood that the reference data and the detection data should be under the same operating condition data, that is, the reference data and the detection data are both data when the rotating mechanical equipment is operating in the same working mode. Therefore, the reference data includes at least one or more of the following: the operating condition data of the rotating machinery equipment, the data extracted from the process data in the acquired operating data, and the mechanical data from the acquired operating data. Data extracted from vibration data.

In some embodiments, the method of obtaining the reference data of the detection point during the normal operation of the rotating mechanical equipment may be obtained in a dynamic manner and/or in a static manner in the foregoing embodiment. For example, in some embodiments, the reference data of the detection point is obtained in a dynamic manner in the foregoing embodiment; another example, in some embodiments, the reference data of the detection point is obtained in a static manner in the foregoing embodiment ; For another example, in some embodiments, a part of the reference data of the detection point is obtained in a dynamic manner in the foregoing embodiment, and another part of the reference data of the detection point is obtained in a static manner in the foregoing embodiment.

In an exemplary embodiment, when the acquired operating data includes process data and/or operating condition data; the step S1201 further includes the following step: analyzing the acquired process data and/or operating condition data to obtain The reference data in the reference data of the detection point during the normal operation of the rotating mechanical equipment.

In some specific examples, according to the abnormal type corresponding to the detection point, the fault diagnosis system uses the process data that meets the normal operating conditions to calculate the corresponding benchmark data. For example, the fault diagnosis system detects the abnormal wind at the air outlet of the fan. It uses the temperature, flow, pressure, and density of the inlet air to calculate the baseline data of the fan's energy efficiency under the current operating conditions. In still other specific examples, according to the abnormal type corresponding to the detection point, the fault diagnosis system uses the operating condition data obtained from the detection point or the locally prestored operating condition data to obtain the benchmark data. For example, in order to detect abnormal bearing rotation, the fault diagnosis system determines that the reference data includes the rotation speed data according to the rotation speed data corresponding to the same operating mode that is continuously obtained multiple times. In still other specific examples, according to the abnormal type corresponding to the detection point, the fault diagnosis system uses the working condition data obtained from the detection point, or the working condition data obtained from the local pre-stored, and the process data to obtain the benchmark data. For example, the fault diagnosis system detects the abnormality of the air outlet of the fan, and uses the operating mode of the air outlet and the motor current as the reference data.

Here, when the detection data includes process data, working condition data, or both process data and working condition data, the fault diagnosis system analyzes the process data, working condition data, or process data and working condition data , So as to obtain the reference data used as the standard in the reference data of the detection point during the normal operation of the rotating machinery equipment, so that the reference data can be used to determine whether the detection data is abnormal.

Wherein, the reference data can be obtained in a static way in addition to the above-mentioned dynamic way. Here, the reference data includes the initial parameters and/or preset calibration parameters of the rotating mechanical equipment. Wherein, the initial parameters include, for example, process data of the rotating mechanical equipment when it leaves the factory. The preset calibration parameters are data obtained after the management personnel of the rotating machinery equipment calibrated all or part of the process data of the rotating machinery equipment based on experience. For example, in some scenarios, due to geographical environment problems, If the temperature of the production environment is high, the temperature of the equipment will be abnormal, resulting in a large difference between the temperature data of the rotating machinery when working in this environment and the temperature data at the factory, but the abnormality of the temperature data is not due to the It is caused by the failure of the rotating mechanical equipment, so the management personnel of the rotating mechanical equipment can calibrate the temperature data of the rotating mechanical equipment based on experience, so as to use the calibrated temperature data as the reference data. It should be understood that the fault diagnosis system may obtain the benchmark data every time the fault diagnosis is performed; it may also store the benchmark data obtained once in the storage medium, and call the data in the storage medium each time the fault diagnosis needs to be performed. Benchmark data.

In an exemplary embodiment, please refer to FIG. 5, which shows a schematic diagram of an embodiment of using operating data to perform anomaly detection and analysis in this application. As shown in the figure, in step S1201, the fault diagnosis system is also based on During the normal operation of the rotating mechanical equipment, the reference data of the detection point performs feature extraction on the acquired detection data to obtain at least one feature element for detecting at least one abnormal type of the detection point.

In some embodiments, after the operating data is acquired by the fault diagnosis system, there may be interference or measurement errors in the operating data acquired from the sensor or the like. In this regard, the fault diagnosis system preprocesses the acquired operating data, then generates test data (or test data and reference data) from the running data, and performs feature extraction on the test data. Wherein, the pre-processing method includes, but is not limited to, noise reduction, abnormal value elimination, and the like.

In some embodiments, the fault diagnosis system performs feature extraction on the detection data according to the reference data, so as to obtain at least one feature element for detecting at least one abnormal type of the detection point. The feature extraction methods include, but are not limited to, average calculation, RMS (Root Mean Square), FFT (Fast Fourier Transformation, Fast Fourier Transformation), envelope spectrum extraction, etc. The features obtained through feature extraction Meta is used as input for analyzing abnormal types.

In some embodiments, after the fault diagnosis system performs feature extraction on the detection data based on the reference data, it further performs feature engineering on the feature extraction results to process the feature extraction results into corresponding abnormal types. Each feature element needed. The feature engineering includes, but is not limited to, feature combination, feature dimensionality reduction, feature processing, feature normalization, and the like.

In some embodiments, in order to facilitate abnormality analysis, the fault diagnosis system further includes performing further data processing on the obtained at least one characteristic element, so as to analyze the at least one characteristic element after further data processing, so as to obtain the corresponding abnormality type. Analyze the results. Here, the data processing includes, but is not limited to, mathematical operations. The feature elements processed by the data can also be used to input into the anomaly detection model for analysis, so as to achieve accurate analysis in the case of insufficient data samples. For example, take the component at 2 times the fundamental frequency, the component at 3 times the fundamental frequency, the component at 4 times the fundamental frequency, and the component at 5 times the fundamental frequency. Incremental combination to form a feature element.

Among them, because different abnormal types need to use feature elements of different types and different descriptions as judgments, each obtained feature element can correspond to one or more abnormal types. In other words, the same feature element may be reused in different anomaly types. Among them, the set feature extraction method can be determined based on the experience of the management personnel of the rotating machinery and equipment, thereby ensuring the criticality of the extracted features, thereby ensuring the accuracy and efficiency of the analysis results.

In some embodiments, the fault diagnosis system further includes a characteristic mechanism model, the input of the characteristic mechanism model is detection data and reference data, and the output of the characteristic mechanism model is each characteristic element corresponding to each abnormal type. Here, the feature mechanism model uses reference data to perform feature extraction on the detection data. Wherein, the feature mechanism model determines the features to be extracted through preset rules, and combines algorithms such as average calculation, effective value calculation, spectrum extraction, envelope spectrum extraction, etc., and uses reference data to perform feature extraction on the detection data to generate features value. Wherein, the reference data may be obtained in a static manner and/or obtained in a dynamic manner. At the same time, the characteristic mechanism model further processes the characteristic value into each characteristic element corresponding to each abnormal type, so that the analysis result of the corresponding abnormal type is obtained after each characteristic element is analyzed. The characteristic mechanism model can be constructed through the mechanism model and using pre-marked historical operating data, reference data, and the like. For example, the historical operating data of the rotating mechanical equipment is obtained from the management personnel or the control system of the rotating mechanical equipment, and the initial parameters of the rotating mechanical equipment are obtained from the management personnel or the control system or the network of the rotating mechanical equipment. It should be understood that the mechanism model is also called the white box model. It is an accurate mathematical model established based on the object, the internal mechanism of the production process, or the transfer mechanism of the material flow. Wherein, the algorithms in the feature mechanism model include but are not limited to feature value calculation, feature engineering, etc. The feature value calculation includes, but is not limited to: average value calculation, effective value calculation, frequency spectrum extraction, and data conversion according to preset conversion formulas. For conversion, taking mechanical vibration data as an example, the characteristic value calculation includes processing the acquired historical mechanical vibration data into a frequency domain spectrum, obtaining a fundamental frequency from the mechanical vibration data, etc. to obtain at least one characteristic value. The feature engineering includes, but is not limited to, subjecting the at least one feature value to feature combination dimensionality reduction, feature processing, feature normalization, etc., to obtain at least one feature element. If the accuracy of the output result of the trained characteristic mechanism model reaches the preset accuracy threshold, the training is completed.

Here, examples are given to illustrate the process of calculating feature elements when the acquired operating data includes mechanical vibration data, or when the acquired detection data includes process data and/or working condition data.

Wherein, when the acquired operating data includes mechanical vibration data, the step S1201 includes: extracting at least one frequency feature element in the frequency spectrum of the acquired mechanical vibration data that is related to the fundamental frequency data in the reference data of the detection point, For detecting at least one type of abnormality of the detection point. Wherein, the fundamental frequency data corresponds to a low-frequency and high-strength frequency or frequency range in the vibration spectrum generated by the rotating mechanical equipment during normal operation under current working conditions during the execution of the current production process. According to the actual working conditions during the production process performed by the rotating machinery and equipment, the fundamental frequency data of different production processes and different working conditions are not completely consistent. For this reason, in some specific examples, the fundamental frequency data is extracted from operating data. For example, the frequency domain conversion is performed on the mechanical vibration data of a certain dimension of the blade, and the fundamental frequency data is obtained through the analysis of the frequency spectrum distribution. In another specific example, the fundamental frequency data is selected from a plurality of correspondences between locally stored processes, operating conditions, and fundamental frequency data according to the acquired operating condition data, process data, and the like.

It should be understood that rotating mechanical equipment (for example: fans, motors, water pumps, gearboxes, etc.) in industrial production rotates and performs work to output energy under the drive of a motor. In this way, motors, rotating parts, etc. have their own rotation speeds, and this rotation speed will cause the equipment to vibrate to a certain extent, and the fundamental frequency of vibration can be calculated by the rotation speed of the equipment. At the same time, the vibration frequency spectrum caused by the rotation speed will be based on the fundamental frequency, and additional spectrum components such as integer multiples, fractional multiples, characteristic multiples, and high-frequency modulation of the fundamental frequency are superimposed. These spectral components are often small when the equipment is operating normally (the vibration energy is small during normal operation), and when the equipment fails, different faults will reflect different vibration frequency spectrums.

It should be understood that, in some embodiments, the fundamental frequency can be calculated by the rotation speed, for example: f=n/60, where f is the frequency (unit: HZ) and n is the rotation speed (unit: rpm). In other embodiments, the fundamental frequency can also be obtained from the vibration frequency spectrum by means of frequency spectrum analysis.

The frequency feature element includes, but is not limited to, the multiplying frequency spectrum component of the fundamental frequency and the like. For example: by performing frequency domain conversion operations on the acquired mechanical vibration data and analyzing the frequency spectrum distribution, frequency characteristic elements such as components at 2 times of the fundamental frequency data and components at 4 times the fundamental frequency are obtained. In some implementations, the frequency feature element may also be a spectrum component of another specific frequency, for example, a component at a frequency of 0.5 times a frequency of 3 times, and so on.

When the acquired detection data includes process data and/or working condition data, the step S1201 further includes: a comparison between the reference data and the acquired detection data in the reference data during the normal operation of the rotating mechanical equipment The deviation obtains at least one deviation feature element, which is used to detect at least one abnormal type of the detection point.

Here, the acquired detection data is compared with the reference data to obtain at least one deviation feature element, the deviation feature element including but not limited to: increment, percentage of increment, and the like. For example: the acquired detection data is temperature data, the reference temperature data in the reference data during the normal operation of the rotating machinery equipment is compared with the temperature data in the detection data, so as to extract the difference in temperature change, and/or the difference The percentage of the value relative to the reference temperature data. The difference and/or the percentage of the difference relative to the reference temperature data are respectively used as a deviation feature element to detect at least one abnormal type of the detection point.

After obtaining the at least one characteristic element in various ways in the foregoing embodiment, the fault diagnosis system provides the at least one characteristic element to step S1202.

Please continue to refer to FIG. 5, in step S1202, each feature element corresponding to each abnormality type is analyzed to obtain the analysis result of the corresponding abnormality type.

Wherein, the method for analyzing each feature element corresponding to each anomaly type includes, but is not limited to, using an anomaly detection model for analysis. The anomaly detection model is a machine learning model, and the machine learning model includes, but is not limited to, KNN-based (k-Nearest Neighbor, k-nearest algorithm) machine learning model, etc. In some embodiments, each abnormality type has an independent machine learning model, such as: impeller imbalance abnormality, coupling misalignment abnormality, rolling bearing vibration abnormality, and sliding bearing vibration abnormality corresponding to the impeller imbalance model and joint Shaft misalignment model, rolling bearing fault model, sliding bearing fault model; abnormal bearing temperature, abnormal motor current corresponding to abnormal bearing temperature model, abnormal motor current model, etc. Since the required data is different when analyzing each abnormal type, the fault diagnosis system converts the detected data into the input required by these models, that is, the feature element, to obtain the corresponding analysis result. For example: the coupling misalignment model requires a combination of a component at 2 times the fundamental frequency, a component at 3 times the fundamental frequency, and a component at 4 times the fundamental frequency as input, then the fault diagnosis system will After feature extraction of the detection data, a combination of the components at 2 times the fundamental frequency, the components at 3 times the fundamental frequency, the components at 4 times the fundamental frequency, and the fundamental frequency itself are two characteristic elements. As input, enter the coupling misalignment model to analyze, and obtain the analysis result of the coupling misalignment model.

At least the machine learning algorithm in the anomaly detection model can obtain the parameters of the algorithm through training such as pre-labeled historical operating data. For example, the historical operating data of the rotating mechanical equipment and the historical fault corresponding to the historical operating data are obtained from the management personnel or the control system of the rotating mechanical equipment. The obtained data is processed into sample data required by the corresponding algorithm and the algorithm is trained to obtain an anomaly detection model. Wherein, the processing process is used to convert the acquired data into data that can be processed by the algorithm, which includes, but is not limited to: normalization processing, data conversion according to a preset conversion formula, and the like. If the accuracy of the trained anomaly detection model reaches the preset accuracy threshold, the training is completed. The input of the abnormality detection model is the feature element required by the model, and the fault diagnosis system uses the abnormality detection model to determine the analysis result of the abnormality type corresponding to each abnormality of the rotating mechanical equipment.

Here, the fault diagnosis system executes step S130 after obtaining the analysis result corresponding to each abnormal type at at least one detection point.

Please continue to refer to FIG. 1, in step S130, the fault diagnosis system uses the obtained at least one analysis result to perform diagnosis processing on at least one fault in the rotating mechanical equipment to output a corresponding fault diagnosis result.

Here, the fault diagnosis system obtains at least one analysis result according to actual conditions such as the amount of operating data provided by the detection point and the network transmission efficiency, and the fault diagnosis system performs fault diagnosis processing according to the at least one analysis result. For example, the fault diagnosis system can diagnose blade imbalance faults or coupling misalignment based on the analysis results of the abnormal types of blade vibrations corresponding to the blade detection points obtained in step S120, or perform blade imbalance faults and couplings respectively. Incorrect fault diagnosis, and obtain the fault diagnosis result through a diagnosis evaluation system.

It should be understood that when different parts of the rotating mechanical equipment fail, there will be different abnormal phenomena. Take the fan as an example. When the fan's bearing fails, it will cause the bearing part to vibrate strongly and the temperature will rise, but it will not cause the abnormal current of the fan motor; another example, when the fan blade fails, it will not cause the bearing Part of the temperature rises, but it will cause the overall fan to shake. Therefore, if only the abnormal type analysis result of a single detection point is used as the fault diagnosis result, the accuracy of the result will be low, and if the abnormal type analysis results of multiple detection points are combined with a comprehensive diagnosis process, the correctness will be obtained. Higher rate of diagnosis results.

Here, the fault diagnosis system also includes a comprehensive diagnosis model, so as to comprehensively diagnose and process the abnormal type analysis results of each detection point on the rotating machinery equipment, and combine the position and type of each detection point to comprehensively process it to obtain The fault diagnosis result of the rotating mechanical equipment. Wherein, the comprehensive diagnosis model is a mechanism model, the input of the comprehensive diagnosis model is an analysis result corresponding to each abnormal type at at least one detection point, and the output of the comprehensive diagnosis model is a fault diagnosis result of the rotating machinery equipment.

In an exemplary embodiment, the comprehensive diagnosis model includes multiple independent fault models, such as: impeller imbalance fault model, coupling misalignment fault model, fan side bearing fault model, motor side bearing fault model, etc. The fault diagnosis system inputs the analysis results of the abnormal types of each detection point into the fault model correspondingly. Among them, the input required for each fault type is different, and the same input may also be used for different fault types. For example: the input of the impeller unbalance fault model corresponds to the analysis result of the abnormal type of impeller unbalance at the detection point, and the input of the coupling misalignment fault model corresponds to the analysis result of the abnormal type of the coupling misalignment at the detection point and the detection point The analysis result of the abnormal type of upper current, the input of the fan-side bearing fault model corresponds to the analysis result of the abnormal type of rolling bearing at the detection point and/or the analysis result of the abnormal type of sliding bearing, the input of the motor-side bearing fault model corresponds to the detection point Analysis results of abnormal types and/or analysis results of abnormal types of sliding bearings, and analysis results of abnormal types of temperature at detection points, etc.

Wherein, the fault diagnosis system presets the diagnosis rules of multiple fault models in the comprehensive diagnosis model, so as to perform fault diagnosis on the rotating mechanical equipment through the analysis result of each abnormal type at at least one detection point. For example, taking a fan as an example, the fault diagnosis system presets that when the analysis results of the abnormal type of rolling bearing at the detection point and the analysis results of the abnormal temperature type are abnormal, and the analysis results of other abnormal types at the detection point are normal, the comprehensive diagnosis The model output is the fault diagnosis result of the motor side bearing fault. It should be understood that, since there may be multiple faults in the rotating mechanical equipment at the same time, in some embodiments, there are multiple fault diagnosis results output by the comprehensive diagnosis model. In some other embodiments, when the fault diagnosis system cannot diagnose the fault of the rotating mechanical equipment, it will output a diagnosis result of an unknown fault.

In an exemplary embodiment, in order to facilitate the management or operation of the management personnel of the rotating machinery and equipment, the fault diagnosis system displays the fault diagnosis result. To this end, the step S130 further includes: presenting the obtained fault diagnosis result on the display interface of the control system of the rotating mechanical equipment. In some embodiments, the computer equipment where the control system is located may be connected to a display, and the management personnel of the rotating machinery equipment can view the fault diagnosis result through the display. In some other embodiments, after obtaining the fault diagnosis result, the management personnel of the rotating machinery equipment may overhaul or check the rotating machinery equipment according to the fault diagnosis result, and feed back the conclusion of whether the fault diagnosis result is correct to the control system through the control system. Fault diagnosis system.

To facilitate understanding, the following will illustrate the fault diagnosis process of the rotating machinery equipment by the fault diagnosis system with 5 detection points in conjunction with Figures 3a-6. It should be understood that this embodiment is only used to explain the application, not to limit the present application. According to the application, according to actual needs, the detection points configured on the rotating mechanical equipment may further include a sixth detection point, a seventh detection point, an eighth detection point, and the like.

In this embodiment, the fault diagnosis system first obtains the operating data of the sensor at the first detection point, where the sensor is a vibration sensor, and the vibration sensor provides mechanical vibration data to the fault diagnosis system. After preprocessing the mechanical vibration data, the fault diagnosis system inputs the preprocessed mechanical vibration data into the characteristic mechanism model. The characteristic mechanism model first generates detection data from vibration machine data. At the same time, the fault diagnosis system obtains reference data corresponding to the detection data, that is, mechanical vibration data of the rotating mechanical equipment during normal operation, and inputs the reference data into the characteristic mechanism model. The characteristic mechanism model records the detection data and reference data into time-domain waveforms. Please refer to FIGS. 3a to 3b, which show a schematic diagram of an embodiment of an abnormality detection and analysis performed by the fault diagnosis system in this application, in which, FIG. 3b It is displayed as the time-domain waveform of the detection data at the detection point in the rotating machinery equipment, that is, the waveform of the real-time mechanical vibration velocity data of the detection point (unit: mm/s); Figure 3a shows the reference data corresponding to the detection data The time-domain waveform, that is, the time-domain waveform of the mechanical vibration data at the detection point when the rotating mechanical equipment is operating normally. It can be seen that it is difficult to analyze the reference data and the detection data by relying solely on FIG. 3a and FIG. 3b, so the characteristic mechanism model further converts the time-domain waveforms of the reference data and the detection data into a frequency-domain spectrum.

Please refer to Figures 4a to 4b, which show schematic diagrams of converting the time domain waveforms in Figures 3a to 3b into frequency domain spectra in this application. As shown in the figure, it is obvious from the frequency domain spectrum that the detection data The characteristic mechanism model processes the characteristic values corresponding to these abnormalities to obtain multiple first characteristic elements. Wherein, the processing method includes converting the characteristic value into a frequency spectrum component of a specific frequency magnification, so as to be input into each abnormality detection model for analysis. Here, after a plurality of first feature elements are obtained, some of the first feature metadata is further processed, so as to reduce the dimensionality to form a plurality of second feature elements.

The feature mechanism model provides the first feature element and/or the second feature element to the corresponding abnormality detection model. Here, the abnormality detection model includes an impeller unbalance model, a coupling misalignment model, a rolling bearing failure model, Sliding bearing fault model, each anomaly detection model outputs the analysis results separately, that is, whether there is the possibility of the fault corresponding to the model. For example, the impeller unbalance model outputs the impeller unbalance abnormality probability of 30%, and the coupling is misaligned The possibility of model output coupling misalignment is 80%, etc. Here, the fault diagnosis system obtains the analysis result output by each abnormality detection model at the first detection point.

The second detection point and the third detection point are also vibration sensors, and the way to obtain the analysis result is the same as that of the first detection point, so it will not be repeated one by one.

Secondly, the fault diagnosis system also acquires operating data of the sensor at the fourth detection point, where the sensor is a temperature sensor, and the vibration sensor provides temperature data to the fault diagnosis system. After preprocessing the temperature data, the fault diagnosis system inputs the preprocessed temperature data into the characteristic mechanism model. The characteristic mechanism model first generates detection data from temperature data. At the same time, the fault diagnosis system obtains reference data corresponding to the detection data, that is, temperature data during normal operation of the rotating mechanical equipment, and inputs the reference data into the characteristic mechanism model. The feature mechanism model compares the detection data with the reference data, thereby calculating the incremental percentage between the detection data and the reference data, and uses this as a feature element. For example, the detection data is 30% higher than the reference data. The characteristic mechanism model provides the characteristic element to the corresponding abnormality detection model. Here, the abnormality detection model includes a temperature abnormality model, and the temperature abnormality model outputs an analysis result that is whether there is a possibility of a fault corresponding to the model For example, the temperature abnormality model outputs a 10% probability of temperature abnormality. Here, the fault diagnosis system obtains the analysis result output by the abnormality detection model at the fourth detection point.

At the same time, the fault diagnosis system also acquires operating data of the sensor at the fifth detection point, where the sensor is a current sensor, and the vibration sensor provides current data to the fault diagnosis system. After preprocessing the current data, the fault diagnosis system inputs the preprocessed current data into the characteristic mechanism model. The characteristic mechanism model first generates detection data from current data. At the same time, the fault diagnosis system obtains reference data corresponding to the detection data, that is, current data when the rotating mechanical equipment is operating normally, and inputs the reference data into the characteristic mechanism model. The feature mechanism model compares the detection data with the reference data, thereby calculating the incremental percentage between the detection data and the reference data, and uses this as a feature element. For example, the detection data is 30% higher than the reference data. The characteristic mechanism model provides the characteristic element to the corresponding abnormality detection model. Here, the abnormality detection model includes a current change model, and the current change model outputs an analysis result that is whether there is a possibility of a fault corresponding to the model For example, the probability that the output current of the current change model is abnormal is 10%. Here, the fault diagnosis system obtains the analysis result output by the abnormality detection model at the fifth detection point.

Please refer to FIG. 6, which shows a schematic diagram of an embodiment of the fault diagnosis process in this application. Here, the analysis results output by each abnormality detection model at each of the five detection points are provided to the comprehensive diagnosis model for The abnormal type analysis result of each detection point is comprehensively diagnosed and processed, combined with the comprehensive processing of the position and type of each detection point, to obtain the fault diagnosis result of the rotating mechanical equipment.

The fault diagnosis solution provided by this application splits a complex single machine learning model into a single anomaly detection model, which not only reduces the complexity of the model, but also uses fewer sample learning to obtain high-precision fault detection effects; in addition, this application The provided solution does not rely on the collection of comprehensive abnormal types, and can provide corresponding fault diagnosis results based on the analysis results of the actual abnormal types that can be collected, thereby effectively improving the flexibility of cooperation between the machine learning models.

The embodiment of the second aspect of the present application provides a server. Please refer to FIG. 7, which shows a schematic structural diagram of a server. As shown in the figure, the server includes an interface unit 11, a storage unit 12, and a processing unit 13. Among them, the storage unit 12 includes a non-volatile memory, a storage server, and the like. Wherein, the non-volatile memory is, for example, a solid state hard disk or a U disk. The storage server is used to store various acquired operating data, reference data, etc. The interface unit 11 includes a network interface, a data line interface, and the like. The network interface includes, but is not limited to: an Ethernet network interface device, a network interface device based on mobile networks (3G, 4G, 5G, etc.), a network interface device based on short-distance communication (WiFi, Bluetooth, etc.), and the like. The data line interface includes but is not limited to: USB interface, RS232, etc. The interface unit is connected with the sensors, the fault diagnosis system, the Internet and other data arranged at the detection points on the rotating mechanical equipment. The processing unit 13 is connected to the interface unit 11 and the storage unit 12, and includes at least one of a CPU or a chip integrated with the CPU, a programmable logic device (FPGA), and a multi-core processor. The processing unit 13 also includes a memory for temporarily storing data, such as a memory and a register.

The interface unit 11 is used for data communication with sensors in at least one dimension at the detection point of the rotating mechanical equipment. Here, the interface unit 11 is an example of a network card, which can communicate with the computer equipment via the Internet or a built-up dedicated network.

The storage unit 12 is used to store at least one program. Here, the storage unit 12 includes, for example, a hard disk set on the server side and stores the at least one program. In addition, according to the external data that needs to be acquired during the running of the program, to the interface unit 11 Various information is stored in the storage unit 12. Wherein, the various information includes the aforementioned reference data of the rotating mechanical equipment and the like.

The processing unit 13 is configured to call the at least one program to coordinate the interface unit and the storage unit to execute the method for diagnosing the failure of the rotating machinery device mentioned in any of the foregoing examples. The fault diagnosis method of the rotating machinery equipment is shown in FIG. 1 and the corresponding description, and will not be repeated here.

The embodiment of the third aspect of the present application provides a first fault diagnosis system for rotating machinery equipment. The first fault diagnosis system includes the server described in the embodiment of the second aspect of the present application and is configured on the rotating machinery. The detection device of each detection point of the equipment. The detection device of each detection point of the rotating machinery equipment is in communication connection with the server, so as to provide the server with the operating data of each detection point, so that the server can diagnose through the provided operating data of each detection point Failure of rotating machinery.

In addition, it should be noted that through the description of the above implementation manners, those skilled in the art can clearly understand that part or all of this application can be implemented by means of software in combination with a necessary general hardware platform. Based on this understanding, the embodiment of the fourth aspect of the present application provides a computer-readable storage medium, the storage medium stores at least one program, and the at least one program executes and implements any one of the foregoing when invoked. The described method for fault diagnosis of rotating machinery equipment.

At the same time, based on this understanding, the technical solution of this application essentially or the part that contributes to the existing technology can also be embodied in the form of a software product, and the computer software product can include a machine executable instruction stored thereon. One or more machine-readable media, when these instructions are executed by one or more machines, such as a computer, a computer network, or other electronic devices, can cause the one or more machines to perform operations according to the embodiments of the present application. For example, perform the steps in the fault diagnosis method of rotating machinery equipment, etc. Machine-readable media may include, but are not limited to, floppy disks, optical disks, CD-ROM (compact disk-read only memory), magneto-optical disks, ROM (read only memory), RAM (random access memory), EPROM (erasable Except programmable read-only memory), EEPROM (electrically erasable programmable read-only memory), magnetic or optical cards, flash memory, or other types of media/machine-readable media suitable for storing machine-executable instructions. Wherein, the storage medium may be located in the server or a third-party server, such as a server that provides an application mall. There are no restrictions on specific application stores, such as Xiaomi App Store, Huawei App Store, and Apple App Store.

This application can be used in many general or special computing system environments or configurations. For example: personal computers, server computers, handheld devices or portable devices, tablet devices, multi-processor systems, microprocessor-based systems, set-top boxes, programmable consumer electronic devices, network PCs, small computers, large computers, including Distributed computing environment of any of the above systems or equipment, etc.

This application may be described in the general context of computer-executable instructions executed by a computer, such as a program module. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform specific tasks or implement specific abstract data types. This application can also be practiced in distributed computing environments. In these distributed computing environments, tasks are performed by remote processing devices connected through a communication network. In a distributed computing environment, program modules can be located in local and remote computer storage media including storage devices.

The embodiment of the fifth aspect of the present application provides a second fault diagnosis system.

In an exemplary embodiment, first, the second fault diagnosis system obtains operating data on at least one detection point in a rotating mechanical equipment.

The second fault diagnosis system uses the operating data to perform abnormality detection and analysis on at least one abnormality type of the detection point respectively to obtain an analysis result corresponding to each abnormality type. It should be understood that, because the rotating mechanical equipment may have multiple faults at the same time in some cases, in order to ensure that the second fault diagnosis system can diagnose multiple faults of the rotating mechanical equipment, it is necessary to use the obtained operation The abnormal type corresponding to the detection point corresponding to the data analysis.

Here, the second fault diagnosis system is preset with an anomaly detection model that is set according to each type of abnormal performance corresponding to various faults at the detection point, wherein the anomaly detection model includes an abnormality detection model that is determined based on input operating data. The algorithm for the possibility of belonging or not belonging to the corresponding abnormal type. The second fault diagnosis system obtains an analysis result of whether the corresponding detection point exhibits a type of abnormality by inputting the received operating data of the detection point into the corresponding abnormality detection model. Among them, according to the different types of failures of rotating machinery and equipment reflected in the abnormal performance of each detection point, the abnormal types are examples but not limited to: impeller imbalance abnormality, coupling misalignment abnormality, rolling bearing vibration abnormality, sliding bearing vibration Abnormal, abnormal bearing temperature change, abnormal motor current change, etc. For example, the second fault diagnosis system performs the abnormal detection and analysis of the impeller imbalance on the acquired operating data to obtain the analysis result of whether the detection point has the impeller imbalance abnormality.

Here, because each type of abnormality requires different data when performing abnormality detection and analysis, the second fault diagnosis system also generates corresponding operating data for each abnormal type that is required for abnormality detection and analysis. The data. Among them, in order to diagnose abnormalities, it is necessary to determine the reference data for the normal operation of the rotating mechanical equipment under the current working conditions during the execution of the current production process.

In other embodiments, the operating data is used to generate detection data and reference data, that is, the reference data is dynamic data. Here, the second fault diagnosis system also analyzes the reference data corresponding to the normal operation of the rotating mechanical equipment from the acquired operating data. For example, the second fault diagnosis system analyzes the mechanical vibration data in the acquired operating data to obtain the fundamental frequency of the rotating mechanical equipment, which is used as reference data, and at the same time, the second fault diagnosis system also obtains the fundamental frequency The mechanical vibration data of is converted into a vibration frequency spectrum, and the fundamental frequency is used to perform abnormality detection and analysis on the mechanical vibration data. In this embodiment, the operating data acquired by the second fault diagnosis system is real-time data provided by the detection point, which directly or indirectly provides detection data and at least part of reference data for abnormality detection and analysis. Wherein, the reference data is used to represent normal data when the rotating mechanical equipment is operating normally under the current working conditions during the execution of the current production process. The detection data is used to represent the current data when the rotating mechanical equipment is currently running under the current working conditions during the execution of the current production process. The second fault diagnosis system performs abnormality detection and analysis on the detection data based on the reference data and obtains corresponding analysis results. In some specific examples, according to the abnormal type, some of the obtained operating data can be directly used as reference data, for example, the torque of the drive motor.

It should be understood that the second fault diagnosis system may obtain the reference data once every time the fault diagnosis is performed; it may also store the reference data obtained once in the storage medium, and call the storage medium every time the fault diagnosis needs to be performed. Reference data in.

In an exemplary embodiment, when the acquired operating data includes process data and/or operating condition data; the second fault diagnosis system also analyzes the acquired process data and/or operating condition data to obtain the The reference data in the reference data of the detection point during the normal operation of the rotating machinery equipment.

In some specific examples, according to the abnormal type corresponding to the detection point, the second fault diagnosis system uses the process data that meets the normal operating conditions to calculate the corresponding benchmark data. For example, the second fault diagnosis system is to detect the abnormal wind at the air outlet of the fan, which uses the temperature, flow, pressure, density, etc. of the inlet air to calculate the baseline data of the fan's energy efficiency under the current operating conditions. In still other specific examples, according to the abnormal type corresponding to the detection point, the second fault diagnosis system obtains the benchmark data by using the working condition data obtained from the detection point or the locally pre-stored working condition data. For example, the second fault diagnosis system detects the abnormality of bearing rotation, and determines that the reference data includes the rotation speed data according to the rotation speed data corresponding to the same operating mode mode continuously obtained multiple times. In still other specific examples, according to the abnormal type corresponding to the detection point, the second fault diagnosis system uses the operating condition data obtained from the detection point, or the locally prestored operating condition data, and the process data to obtain the benchmark data. For example, the second fault diagnosis system detects the abnormality of the air outlet of the fan, and uses the operating mode of the air outlet and the motor current as reference data.

Here, when the detection data includes process data, working condition data, or both process data and working condition data, the second fault diagnosis system performs a check on the process data, working condition data, or process data and working condition data. The analysis is carried out to obtain the reference data used as the standard in the reference data of the detection point when the rotating machinery is operating normally, so that the reference data can be used to determine whether the detection data is abnormal.

Wherein, the reference data can be obtained in a static way in addition to the above-mentioned dynamic way. Here, the reference data includes the initial parameters and/or preset calibration parameters of the rotating mechanical equipment. Wherein, the initial parameters include, for example, process data of the rotating mechanical equipment when it leaves the factory. The preset calibration parameters are data obtained after the management personnel of the rotating machinery equipment calibrated all or part of the process data of the rotating machinery equipment based on experience. For example, in some scenarios, due to geographical environment problems, If the temperature of the production environment is high, the temperature of the equipment will be abnormal, resulting in a large difference between the temperature data of the rotating machinery when working in this environment and the temperature data at the factory, but the abnormality of the temperature data is not due to the It is caused by the failure of the rotating mechanical equipment, so the management personnel of the rotating mechanical equipment can calibrate the temperature data of the rotating mechanical equipment based on experience, so as to use the calibrated temperature data as the reference data. It should be understood that the second fault diagnosis system may obtain the benchmark data every time a fault diagnosis is performed; it may also store the benchmark data obtained once in a storage medium, and call the storage medium every time the fault diagnosis needs to be performed. Benchmark data in.

In an exemplary embodiment, the second fault diagnosis system further performs feature extraction on the acquired detection data based on the reference data of the detection points during the normal operation of the rotating mechanical equipment, so as to obtain the At least one feature element of at least one abnormal type of the detection point.

In some embodiments, after the operating data is acquired by the second fault diagnosis system, the operating data acquired from the sensor or the like may have interference or measurement errors. In this regard, the second fault diagnosis system preprocesses the acquired operating data, and then generates testing data (or testing data and reference data) from the operating data, and performs feature extraction on the testing data. Wherein, the pre-processing method includes, but is not limited to, noise reduction, abnormal value elimination, and the like.

In some embodiments, the second fault diagnosis system performs feature extraction on the detection data according to the reference data, so as to obtain at least one feature element for detecting at least one abnormal type of the detection point. The feature extraction methods include but are not limited to average calculation, effective value calculation, frequency spectrum extraction, envelope spectrum extraction, etc. The feature elements obtained by feature extraction are used as input for analyzing abnormal types.

In some embodiments, after the second fault diagnosis system performs feature extraction on the detection data based on the reference data, it further performs feature engineering on the feature extraction results to process the feature extraction results into corresponding to each abnormality. Each feature element required by the type. The feature engineering includes, but is not limited to, feature combination, feature dimensionality reduction, feature processing, feature normalization, and the like.

In some embodiments, in order to facilitate abnormality analysis, the second fault diagnosis system further includes performing further data processing on the obtained at least one characteristic element, so as to analyze the at least one characteristic element after further data processing, so as to obtain the corresponding abnormality. Type of analysis results. Here, the data processing includes, but is not limited to, mathematical operations. The feature elements processed by the data can also be used to input into the anomaly detection model for analysis, so as to achieve accurate analysis in the case of insufficient data samples. For example, take the component at 2 times the fundamental frequency, the component at 3 times the fundamental frequency, the component at 4 times the fundamental frequency, and the component at 5 times the fundamental frequency. Incremental combination to form a feature element.

In some embodiments, the second fault diagnosis system further includes a characteristic mechanism model. The input of the characteristic mechanism model is detection data and reference data, and the output of the characteristic mechanism model is each characteristic corresponding to each abnormal type. yuan. Here, the feature mechanism model uses reference data to perform feature extraction on the detection data. Wherein, the feature mechanism model determines the features to be extracted through preset rules, and combines algorithms such as average calculation, effective value calculation, spectrum extraction, envelope spectrum extraction, etc., and uses reference data to perform feature extraction on the detection data to generate features value. Wherein, the reference data may be obtained in a static manner and/or obtained in a dynamic manner. At the same time, the characteristic mechanism model further processes the characteristic value into each characteristic element corresponding to each abnormal type, so that the analysis result of the corresponding abnormal type is obtained after each characteristic element is analyzed. The characteristic mechanism model can be constructed through the mechanism model and using pre-marked historical operating data, reference data, and the like. For example, the historical operating data of the rotating mechanical equipment is obtained from the management personnel or the control system of the rotating mechanical equipment, and the initial parameters of the rotating mechanical equipment are obtained from the management personnel or the control system or the network of the rotating mechanical equipment. It should be understood that the mechanism model is also called the white box model. It is an accurate mathematical model established based on the object, the internal mechanism of the production process, or the transfer mechanism of the material flow. Wherein, the algorithms in the feature mechanism model include but are not limited to feature value calculation, feature engineering, etc. The feature value calculation includes, but is not limited to: average value calculation, effective value calculation, frequency spectrum extraction, and data conversion according to preset conversion formulas. For conversion, taking mechanical vibration data as an example, the characteristic value calculation includes processing the acquired historical mechanical vibration data into a frequency domain spectrum, obtaining a fundamental frequency from the mechanical vibration data, etc. to obtain at least one characteristic value. The feature engineering includes, but is not limited to, subjecting the at least one feature value to feature combination dimensionality reduction, feature processing, feature normalization, etc., to obtain at least one feature element. If the accuracy of the output result of the trained characteristic mechanism model reaches the preset accuracy threshold, the training is completed.

Wherein, when the acquired operating data includes mechanical vibration data, the second fault diagnosis system extracts at least one frequency feature element in the frequency spectrum of the acquired mechanical vibration data that is related to the fundamental frequency data in the reference data of the detection point , For detecting at least one abnormal type of the detection point. Wherein, the fundamental frequency data corresponds to a low-frequency and high-strength frequency or frequency range in the vibration spectrum generated by the rotating mechanical equipment during normal operation under current working conditions during the execution of the current production process. According to the actual working conditions during the production process performed by the rotating machinery and equipment, the fundamental frequency data of different production processes and different working conditions are not completely consistent. For this reason, in some specific examples, the fundamental frequency data is extracted from operating data. For example, the frequency domain conversion is performed on the mechanical vibration data of a certain dimension of the blade, and the fundamental frequency data is obtained through the analysis of the frequency spectrum distribution. In another specific example, the fundamental frequency data is selected from a plurality of correspondences between locally stored processes, operating conditions, and fundamental frequency data according to the acquired operating condition data, process data, and the like.

It should be understood that rotating mechanical equipment (for example: fans, motors, water pumps, gearboxes, etc.) in industrial production rotates and performs work to output energy under the drive of a motor. In this way, motors, rotating parts, etc. have their own rotation speeds, and this rotation speed will cause the equipment to vibrate to a certain extent. The fundamental frequency of vibration can be calculated by the rotation speed of the equipment. At the same time, the vibration frequency spectrum caused by the rotation speed will be based on the fundamental frequency, and additional spectrum components such as integer multiples, fractional multiples, characteristic multiples, and high-frequency modulation of the fundamental frequency are superimposed. These spectral components are often small when the equipment is operating normally (the vibration energy is small during normal operation), and when the equipment fails, different faults will reflect different vibration frequency spectrums.

When the acquired detection data includes process data and/or working condition data, the second fault diagnosis system is also based on the reference data in the reference data during normal operation of the rotating mechanical equipment and the acquired detection data Obtain at least one deviation feature element for detecting at least one abnormal type of the detection point.

After obtaining the at least one characteristic element in various ways in the foregoing embodiment, the second fault diagnosis system analyzes each characteristic element corresponding to each abnormality type to obtain an analysis result of the corresponding abnormality type.

Wherein, the method for analyzing each feature element corresponding to each anomaly type includes, but is not limited to, using an anomaly detection model for analysis. The anomaly detection model is a machine learning model, and the machine learning model includes, but is not limited to, KNN-based (k-Nearest Neighbor, k-nearest algorithm) machine learning model, etc. In some embodiments, each abnormality type has an independent machine learning model, such as: impeller imbalance abnormality, coupling misalignment abnormality, rolling bearing vibration abnormality, and sliding bearing vibration abnormality corresponding to the impeller imbalance model and joint Shaft misalignment model, rolling bearing fault model, sliding bearing fault model; abnormal bearing temperature, abnormal motor current corresponding to abnormal bearing temperature model, abnormal motor current model, etc. Since the data required when analyzing each abnormal type is different, the second fault diagnosis system converts the detected data into the input required by these models, that is, the feature element, so as to obtain the corresponding analysis result. For example: the coupling misalignment model requires a combination of a component at 2 times the fundamental frequency, a component at 3 times the fundamental frequency, and a component at 4 times the fundamental frequency as input, then the fault diagnosis system will After feature extraction of the detection data, a combination of the components at 2 times the fundamental frequency, the components at 3 times the fundamental frequency, the components at 4 times the fundamental frequency, and the fundamental frequency itself are two characteristic elements. As input, enter the coupling misalignment model for analysis, and obtain the analysis result of the coupling misalignment model. At least the machine learning algorithm in the anomaly detection model can obtain the parameters of the algorithm through training such as pre-marked historical operating data. For example, the historical operating data of the rotating mechanical equipment and the historical fault corresponding to the historical operating data are obtained from the management personnel or the control system of the rotating mechanical equipment. The obtained data is processed into sample data required by the corresponding algorithm and the algorithm is trained to obtain an anomaly detection model. Wherein, the processing process is used to convert the acquired data into data that can be processed by the algorithm, which includes, but is not limited to: normalization processing, data conversion according to a preset conversion formula, and the like. If the accuracy of the trained anomaly detection model reaches the preset accuracy threshold, the training is completed. The input of the abnormality detection model is the feature element required by the model, and the second fault diagnosis system uses the abnormality detection model to determine the analysis result of the abnormality type corresponding to each abnormality of the rotating mechanical equipment.

Here, after the second fault diagnosis system obtains the analysis result corresponding to each abnormal type at at least one detection point, the second fault diagnosis system uses the obtained at least one analysis result to analyze the At least one of the faults is diagnosed to output the corresponding fault diagnosis result.

Here, the second fault diagnosis system obtains at least one analysis result according to actual conditions such as the amount of operating data provided by the detection point, network transmission efficiency, etc., and the second fault diagnosis system performs fault diagnosis processing according to the at least one analysis result . For example, the second fault diagnosis system can diagnose blade imbalance faults or coupling mismatches based on the obtained analysis results of the corresponding blade detection points on the abnormal types of blade vibrations, or perform blade imbalance faults and couplings respectively. Incorrect fault diagnosis, and obtain the fault diagnosis result through a diagnosis evaluation system.

Here, the second fault diagnosis system also includes a comprehensive diagnosis model, so as to comprehensively diagnose and process the abnormal type analysis results of each detection point on the rotating machinery equipment, and combine the position and type of each detection point to comprehensively process , To obtain the fault diagnosis result of the rotating mechanical equipment. Wherein, the comprehensive diagnosis model is a mechanism model, the input of the comprehensive diagnosis model is an analysis result corresponding to each abnormal type at at least one detection point, and the output of the comprehensive diagnosis model is a fault diagnosis result of the rotating machinery equipment.

In an exemplary embodiment, the comprehensive diagnosis model includes multiple independent fault models, such as: impeller imbalance fault model, coupling misalignment fault model, fan side bearing fault model, motor side bearing fault model, etc. The second fault diagnosis system inputs the analysis results of the abnormal types of each detection point into the fault model correspondingly. Among them, the input required for each fault type is different, and the same input may also be used for different fault types. For example: the input of the impeller unbalance fault model corresponds to the analysis result of the abnormal type of impeller unbalance at the detection point, and the input of the coupling misalignment fault model corresponds to the analysis result of the abnormal type of the coupling misalignment at the detection point and the detection point The analysis result of the abnormal type of upper current, the input of the fan-side bearing fault model corresponds to the analysis result of the abnormal type of rolling bearing at the detection point and/or the analysis result of the abnormal type of sliding bearing, the input of the motor-side bearing fault model corresponds to the detection point of the rolling bearing Analysis results of abnormal types and/or analysis results of abnormal types of sliding bearings, and analysis results of abnormal types of temperature at detection points, etc.

Wherein, the second fault diagnosis system presets the diagnosis rules of multiple fault models in the comprehensive diagnosis model, so as to perform fault diagnosis on the rotating mechanical equipment through the analysis result of each abnormal type at at least one detection point. For example, taking a fan as an example, the second fault diagnosis system presupposes that when the analysis results of the abnormal type of rolling bearing at the detection point and the analysis results of the abnormal temperature type are abnormal, and the analysis results of other abnormal types at the detection point are normal, The output of the comprehensive diagnosis model is the fault diagnosis result of the motor-side bearing fault. It should be understood that, since there may be multiple faults in the rotating mechanical equipment at the same time, in some embodiments, there are multiple fault diagnosis results output by the comprehensive diagnosis model. In some other embodiments, when the second fault diagnosis system cannot diagnose the fault of the rotating mechanical equipment, it will output a diagnosis result of an unknown fault.

In an exemplary embodiment, in order to facilitate the management or operation of the management personnel of the rotating machinery and equipment, the second fault diagnosis system displays the fault diagnosis result. To this end, the second fault diagnosis system also presents the obtained fault diagnosis results on the display interface of the control system of the rotating mechanical equipment. In some embodiments, the computer equipment where the control system is located may be connected to a display, and the management personnel of the rotating machinery equipment can view the fault diagnosis result through the display. In some other embodiments, after obtaining the fault diagnosis result, the management personnel of the rotating machinery equipment may overhaul or check the rotating machinery equipment according to the fault diagnosis result, and feed back the conclusion of whether the fault diagnosis result is correct to the control system through the control system. The second fault diagnosis system.

For ease of understanding, the following five detection points will be used to illustrate the fault diagnosis process of the rotating machinery equipment by the second fault diagnosis system. It should be understood that this embodiment is only used to explain the present application, not to limit the present application.

In this embodiment, the second fault diagnosis system first obtains the operating data of the sensor at the first detection point, where the sensor is a vibration sensor, and the vibration sensor provides mechanical vibration data to the second fault diagnosis system. After preprocessing the mechanical vibration data, the second fault diagnosis system inputs the preprocessed mechanical vibration data into the characteristic mechanism model. The characteristic mechanism model first generates detection data from vibration machine data. At the same time, the second fault diagnosis system obtains reference data corresponding to the detection data, that is, mechanical vibration data of the rotating mechanical equipment during normal operation, and inputs the reference data into the characteristic mechanism model. The characteristic mechanism model records the detection data and reference data into a time domain waveform, and the characteristic mechanism model further converts the time domain waveform of the reference data and the detection data into a frequency domain spectrum.

The characteristic mechanism model processes the characteristic values corresponding to these abnormalities to obtain a plurality of first characteristic elements. Wherein, the processing method includes converting the characteristic value into a frequency spectrum component of a specific frequency magnification, so as to be input into each abnormality detection model for analysis. Here, after a plurality of first feature elements are obtained, some of the first feature metadata is further processed, so as to reduce the dimensionality to form a plurality of second feature elements.

The feature mechanism model provides the first feature element and/or the second feature element to the corresponding abnormality detection model. Here, the abnormality detection model includes an impeller unbalance model, a coupling misalignment model, a rolling bearing failure model, Sliding bearing fault model, each anomaly detection model outputs the analysis results separately, that is, whether there is the possibility of the fault corresponding to the model. For example, the impeller unbalance model outputs the impeller unbalance abnormality probability of 30%, and the coupling is misaligned The possibility of model output coupling misalignment is 80%, etc. Here, the second fault diagnosis system obtains the analysis results output by each abnormality detection model at the first detection point.

Secondly, the second fault diagnosis system also acquires operating data of the sensor at the fourth detection point, where the sensor is a temperature sensor, and the vibration sensor provides temperature data to the second fault diagnosis system. After preprocessing the temperature data, the second fault diagnosis system inputs the preprocessed temperature data into the characteristic mechanism model. The characteristic mechanism model first generates detection data from temperature data. At the same time, the second fault diagnosis system obtains reference data corresponding to the detection data, that is, temperature data of the rotating mechanical equipment during normal operation, and inputs the reference data into the characteristic mechanism model. The feature mechanism model compares the detection data with the reference data, thereby calculating the incremental percentage between the detection data and the reference data, and uses this as a feature element. For example, the detection data is 30% higher than the reference data. The characteristic mechanism model provides the characteristic element to the corresponding abnormality detection model. Here, the abnormality detection model includes a temperature abnormality model, and the temperature abnormality model outputs an analysis result that is whether there is a possibility of a fault corresponding to the model For example, the temperature abnormality model outputs a 10% probability of temperature abnormality. Here, the second fault diagnosis system obtains the analysis result output by the abnormality detection model at the fourth detection point.

At the same time, the second fault diagnosis system also acquires the operating data of the sensor at the fifth detection point, where the sensor is a current sensor, and the vibration sensor provides the current data to the second fault diagnosis system. After preprocessing the current data, the second fault diagnosis system inputs the preprocessed current data into the characteristic mechanism model. The characteristic mechanism model first generates detection data from current data. At the same time, the second fault diagnosis system obtains reference data corresponding to the detection data, that is, current data when the rotating mechanical equipment is operating normally, and inputs the reference data into the characteristic mechanism model. The feature mechanism model compares the detection data with the reference data, thereby calculating the incremental percentage between the detection data and the reference data, and uses this as a feature element. For example, the detection data is 30% higher than the reference data. The characteristic mechanism model provides the characteristic element to the corresponding abnormality detection model. Here, the abnormality detection model includes a current change model, and the current change model outputs an analysis result that is whether there is a possibility of a fault corresponding to the model For example, the probability that the output current of the current change model is abnormal is 10%. Here, the second fault diagnosis system obtains the analysis result output by the abnormality detection model at the fifth detection point.

Here, the analysis results output by each abnormality detection model at each of the 5 detection points are provided to the comprehensive diagnosis model, so that the analysis results of the abnormality types of each detection point can be comprehensively diagnosed and combined with the location of each measurement point , Type comprehensive processing, and obtain the fault diagnosis result of the rotating mechanical equipment.

The second fault diagnosis system in this application splits the complex single machine learning model into a single abnormality detection model, which not only reduces the complexity of the model, but also uses fewer sample learning to obtain a high-precision fault detection effect; in addition, this The second fault diagnosis system in the application does not rely on the collection of comprehensive exception types, and can provide corresponding fault diagnosis results based on the analysis results of the exception types that can actually be collected, thereby effectively improving the coordination between the machine learning models flexibility.

The embodiment of the sixth aspect of the present application provides a management system for rotating machinery equipment. The management system for rotating machinery equipment includes detection devices arranged at each detection point of the rotating machinery equipment, a control system for the rotating machinery equipment, and The server described in the embodiment of the second aspect.

Wherein, the detection device arranged at each detection point of the rotating mechanical equipment is used to provide operating data of each detection point on the rotating mechanical equipment. The control system of the rotating machinery equipment is data-connected with each of the detection devices, so as to collect and forward each of the operating data to the server, so that the server that is in communication with the control system receives each of the operating data and receives Execute the corresponding fault diagnosis method for the running data. Wherein, the fault diagnosis method is the same as the fault diagnosis method in the foregoing embodiment, so it will not be repeated one by one.

The foregoing embodiments only exemplarily illustrate the principles and effects of the present application, and are not used to limit the present application. Anyone familiar with this technology can modify or change the above-mentioned embodiments without departing from the spirit and scope of this application. Therefore, all equivalent modifications or changes made by persons with ordinary knowledge in the technical field without departing from the spirit and technical ideas disclosed in this application should still be covered by the claims of this application.

Claims

A fault diagnosis method for rotating machinery equipment is characterized in that it comprises the following steps:

Acquiring operating data on at least one detection point in a rotating mechanical equipment;

Performing anomaly detection and analysis on at least one abnormality type of the detection point by using the operating data to obtain an analysis result corresponding to each abnormality type;

Using the obtained at least one analysis result to perform diagnosis processing on at least one type of fault in the rotating mechanical equipment, so as to output a corresponding fault diagnosis result.
The method for diagnosing faults of rotating machinery equipment according to claim 1, wherein the abnormal type includes at least one of the following abnormalities reflected by corresponding detection points during the operation of the rotating machinery equipment based on working conditions and process requirements: An anomaly type set separately for each vibration anomaly in at least one spatial dimension, an anomaly type set based on an abnormal temperature, and an anomaly type set separately based on each power anomaly.
The fault diagnosis method for rotating machinery equipment according to claim 1, wherein the operating data is used to generate detection data and reference data; and the use of the operating data is performed on at least one abnormal type of the detection point. The steps of anomaly detection and analysis include:

Performing feature extraction on the acquired detection data based on the reference data of the detection point during the normal operation of the rotating mechanical equipment to obtain at least one feature element for detecting at least one abnormal type of the detection point;

Analyze each feature element corresponding to each abnormal type to obtain the analysis result of the corresponding abnormal type.
The fault diagnosis method for rotating machinery equipment according to claim 3, further comprising the following step: further data processing is performed on the obtained at least one characteristic element, so as to analyze the at least one characteristic element after further data processing, thereby Get the analysis result of the corresponding abnormal type.
The fault diagnosis method for rotating machinery equipment according to claim 3 or 4, wherein the reference data includes at least one or more of the following: operating condition data of the rotating machinery equipment is obtained from the obtained operating data The data is extracted from the process data, and the data is extracted from the mechanical vibration data in the acquired operating data.
The method for diagnosing faults of rotating machinery equipment according to claim 3, further comprising the step of analyzing the fundamental frequency data in the reference data corresponding to the normal operation of the rotating machinery equipment from the acquired operating data.
The fault diagnosis method for rotating machinery equipment according to claim 3 or 6, characterized in that the acquired operating data includes mechanical vibration data; the reference data pair based on the detection point during the normal operation of the rotating machinery equipment is acquired The step of performing feature extraction on the detection data to obtain at least one feature element for detecting at least one abnormal type of the detection point includes:

At least one frequency feature element related to the fundamental frequency data in the reference data of the detection point in the frequency spectrum of the acquired mechanical vibration data is extracted to be used for detecting at least one abnormal type of the detection point.
The fault diagnosis method for rotating machinery equipment according to claim 3, wherein the acquired operating data includes process data and/or working condition data; the method further comprises the following steps: analyzing the acquired process data and/or Working condition data to obtain the reference data in the reference data of the detection point during the normal operation of the rotating mechanical equipment.
The fault diagnosis method for rotating machinery equipment according to claim 8, characterized in that, based on the reference data of the detection points during the normal operation of the rotating machinery equipment, feature extraction is performed on the acquired detection data to obtain The step of detecting at least one feature element of at least one abnormal type of the point includes:

At least one deviation feature element is obtained based on the deviation between the reference data in the reference data during the normal operation of the rotating mechanical equipment and the acquired detection data, which is used to detect at least one abnormal type of the detection point.
The method for diagnosing faults of rotating machinery equipment according to claim 1, wherein said using the obtained at least one analysis result performs diagnosis processing on at least one fault in said rotating machinery equipment to output corresponding fault diagnosis The result step includes: presenting the obtained fault diagnosis result on the display interface of the control system of the rotating mechanical equipment.
A server, which is characterized in that it includes:

The interface unit is used for data communication with sensors in at least one dimension at the detection point of the rotating mechanical equipment;

A storage unit for storing at least one program; and

The processing unit is configured to call the at least one program to coordinate the execution of the interface unit and the storage unit and implement the method for diagnosing the fault of a rotating machinery device according to any one of claims 1-10.
A first fault diagnosis system for rotating machinery equipment, characterized in that it comprises:

The server according to claim 11; and

The detection device arranged at each detection point of the rotating machinery equipment is communicatively connected with the server to provide the operating data of each detection point.
A computer-readable storage medium, characterized in that it stores at least one program, and the at least one program executes and implements the method for diagnosing the fault of a rotating machinery device according to any one of claims 1 to 10 when called.
A second fault diagnosis system for rotating machinery equipment, characterized in that it comprises:

The data acquisition module is used to acquire operating data on at least one detection point in a rotating mechanical equipment;

The data processing module is used to perform anomaly detection and analysis on at least one abnormality type of the detection point by using the operating data to obtain an analysis result corresponding to each abnormality type, and use the obtained at least one analysis result to analyze all the abnormalities. At least one fault in the rotating mechanical equipment is diagnosed to output a corresponding fault diagnosis result.
The second fault diagnosis system for rotating machinery equipment according to claim 14, wherein the abnormal type includes at least one of the following reflected by corresponding detection points during the operation of the rotating machinery equipment based on working conditions and process requirements Anomaly: An anomaly type set separately based on each vibration anomaly in at least one spatial dimension, an anomaly type set based on an abnormal temperature, an anomaly type set separately based on an abnormality of each power.
The second fault diagnosis system for rotating machinery equipment according to claim 14, wherein the operating data is used to generate detection data and reference data; and the data processing module is based on the normal operation of the rotating machinery equipment. When the reference data of the detection point performs feature extraction on the acquired detection data to obtain at least one feature element for detecting at least one abnormality type of the detection point; and for each feature element corresponding to each abnormality type Perform analysis to get the analysis result of the corresponding abnormal type.
The second fault diagnosis system for rotating machinery equipment according to claim 16, wherein the data processing module performs further data processing on the obtained at least one characteristic element, so as to perform further data processing on the at least one characteristic element after further data processing. Perform analysis to obtain the analysis result of the corresponding abnormal type.
The second fault diagnosis system for rotating machinery equipment according to claim 16 or 17, wherein the reference data includes at least one or more of the following: operating condition data of the rotating machinery equipment is obtained from Data extracted from the process data in the operating data, and data extracted from the mechanical vibration data in the acquired operating data.
The second fault diagnosis system for rotating machinery equipment according to claim 16, wherein the data processing module further analyzes from the acquired operating data the reference data corresponding to the normal operation of the rotating machinery equipment. Fundamental frequency data.
The second fault diagnosis system for rotating machinery equipment according to claim 16 or 19, wherein the acquired operating data includes mechanical vibration data, and the data processing module extracts the acquired mechanical vibration data from the frequency spectrum and At least one frequency feature element related to the fundamental frequency data in the reference data of the detection point is used to detect at least one abnormal type of the detection point.
The second fault diagnosis system for rotating machinery equipment according to claim 16, wherein the acquired operating data includes process data and/or working condition data; the data processing module also analyzes the acquired process data and/ Or working condition data to obtain the reference data in the reference data of the detection point during the normal operation of the rotating mechanical equipment.
The second fault diagnosis system for rotating machinery equipment according to claim 21, wherein the data processing module is based on the reference data in the reference data during normal operation of the rotating machinery equipment and the acquired detection data. Obtain at least one deviation feature element for detecting at least one abnormal type of the detection point.
22. The second fault diagnosis system for rotating machinery equipment according to claim 21, further comprising a fault display device for presenting the obtained fault diagnosis result.
A management system for rotating machinery equipment, characterized in that it comprises:

The detection devices arranged at each detection point of the rotating machinery equipment are used to provide the operating data of each detection point;

The control system of the rotating machinery equipment is connected to each of the detection devices to collect and forward each of the operating data;

The server according to claim 11, in communication connection with the control system, for receiving each of the operating data and executing a corresponding fault diagnosis method based on the received operating data.