Rotating equipment health assessment method and device based on slow variable and electronic equipment

Abstract

The invention provides a rotating equipment health assessment method and device based on a slow variable and electronic equipment. The method comprises the following steps: respectively determining an equipment importance coefficient and a fault hazard coefficient of each key equipment in rotating equipment; acquiring state monitoring data of each key device, and determining a single device health score of each key device; according to the equipment importance coefficient, the fault hazard coefficient and the single equipment health score of at least part of the key equipment, calculating to obtain an overall health score of the rotating equipment; performing feature extraction on the state monitoring data, and determining a slow variable from the extracted feature quantity; inputting the slow variable into the performance degradation model, and predicting to obtain the residual life of the rotating equipment; and taking the total health score and the residual life of the rotating equipment as a health assessment result of the rotating equipment. The overall health monitoring of the rotating equipment is realized, the residual life of the rotating equipment is predicted by extracting and utilizing the slow variable, and the preventive maintenance of the rotating equipment can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment health monitoring, and in particular to a method, device and electronic equipment for health assessment of rotating equipment based on slow variables. Background Art

[0002] In recent years, with the rapid development of big data, digital twins and artificial intelligence, a new round of scientific and technological revolution and industrial transformation has been accelerated. Large and complex equipment systems have complex internal structures, numerous components, and intertwined relationships. The system itself has dynamic uncertainty. Therefore, it is particularly important to carry out health monitoring of such complex equipment.

[0003] Taking rotating equipment as an example, traditional health monitoring solutions can only calculate the health status of a certain component through sensor data, such as the operation status of a motor, the working status of a certain bearing, etc., but cannot know the overall health status of the equipment; at the same time, traditional health monitoring solutions can only judge the current health status of the equipment, which is post-monitoring and cannot achieve preventive maintenance.

[0004] Therefore, traditional health monitoring solutions have technical problems such as insufficient monitoring and only being able to implement post-event monitoring. Summary of the invention

[0005] The present invention provides a method, device and electronic equipment for health assessment of rotating equipment based on slow variables, so as to solve the defects that traditional health monitoring schemes are not comprehensive enough and can only realize post-event monitoring.

[0006] In one aspect, the present invention provides a method for health assessment of rotating equipment based on slow variables, comprising:

[0007] Determine the equipment importance coefficient and failure criticality coefficient of each key equipment in the rotating equipment respectively;

[0008] Acquire status monitoring data of each of the key devices, and determine a single device health score of each of the key devices based on the status monitoring data;

[0009] Calculate the overall health score of the rotating equipment according to the equipment importance coefficient, the failure criticality coefficient and the single equipment health score of at least part of the key equipment;

[0010] Extracting features from the state monitoring data and determining slow variables from the extracted features;

[0011] Inputting the slow variable into a pre-built performance degradation model to predict the remaining life of the rotating equipment;

[0012] The overall health score and remaining life of the rotating equipment are used as the health assessment result of the rotating equipment.

[0013] According to the rotating equipment health assessment method based on slow variables provided by the present invention, the device importance coefficient and the fault criticality coefficient of each key device in the rotating equipment are determined separately, including:

[0014] Construct a structure tree based on the structural hierarchy of the rotating equipment;

[0015] Constructing a fault tree according to the logical relationship between the fault events of each key device in the structure tree;

[0016] Determine the equipment importance coefficient of each key equipment according to the weight value of each key equipment in the structure tree;

[0017] According to the weight values ​​of each key device in the fault tree under different fault events, the failure criticality coefficient of each key device is determined.

[0018] According to the rotating equipment health assessment method based on slow variables provided by the present invention, the structure tree is constructed according to the structural hierarchical relationship of the rotating equipment, including:

[0019] According to the structural hierarchy, the rotating equipment is subdivided and the subordinate relationship between each layer of structure is established;

[0020] Taking each layer structure as a tree node, and determining the weight value of each tree node according to the relative importance of each tree node in the parent node;

[0021] A structure tree is established according to the subordinate relationships, the tree nodes and the weight values ​​of the tree nodes.

[0022] According to the rotating equipment health assessment method based on slow variables provided by the present invention, a fault tree is constructed according to the fault logical relationship of each key equipment in the structure tree, including:

[0023] Decompose the failure causes of rotating equipment into fault events corresponding to key equipment at different levels, and determine the logical relationship between each fault event;

[0024] Determining a weight value of each fault event according to the degree of harm of each fault event;

[0025] A fault tree is established according to the key devices at different levels, the fault events, the logical relationships and the weight values ​​of the fault events.

[0026] According to the rotating equipment health assessment method based on slow variables provided by the present invention, determining the single equipment health score of each of the key equipment according to the state monitoring data includes:

[0027] For each key equipment, obtaining historical measurement point data of the key equipment;

[0028] Performing time series analysis and frequency domain analysis on the historical measurement point data respectively, and performing feature extraction to obtain equipment characteristic parameters;

[0029] Establishing a health assessment model according to the characteristic parameters of the equipment;

[0030] The status monitoring data is input into the health assessment model to obtain the single device health score of the key device output by the health assessment model.

[0031] According to the rotating equipment health assessment method based on slow variables provided by the present invention, the performance degradation model is established through the following process:

[0032] Obtaining historical measurement point data of each of the key devices;

[0033] Performing time series analysis and frequency domain analysis on the historical measurement point data respectively, and performing feature extraction to obtain equipment characteristic parameters;

[0034] Establishing a prediction model according to the characteristic parameters of the equipment;

[0035] Inputting the slow variable into the prediction model to obtain the parameter change trend within a future preset time period output by the prediction model;

[0036] A performance degradation model is established according to the slow variable and the parameter change trend.

[0037] According to the rotating equipment health assessment method based on slow variables provided by the present invention, the method further includes:

[0038] Comparing the overall health score of the rotating equipment with a preset score threshold to obtain a comparison result;

[0039] An equipment operation and maintenance plan is generated according to the comparison result.

[0040] According to the rotating equipment health assessment method based on slow variables provided by the present invention, the method further includes:

[0041] The maintenance time and maintenance plan of the rotating equipment are determined according to the remaining life of the rotating equipment.

[0042] On the other hand, the present invention also provides a rotating equipment health assessment device based on slow variables, comprising:

[0043] A determination module, used to respectively determine the equipment importance coefficient and the failure criticality coefficient of each key equipment in the rotating equipment;

[0044] An acquisition module, used to acquire status monitoring data of each of the key devices, and determine a single device health score of each of the key devices based on the status monitoring data;

[0045] A calculation module, configured to calculate an overall health score of the rotating equipment according to equipment importance coefficients, failure criticality coefficients and single equipment health scores of at least some of the key equipment;

[0046] An extraction module, used for extracting features from the state monitoring data and determining slow variables from the extracted features;

[0047] A prediction module, used for inputting the slow variable into a pre-built performance degradation model to predict the remaining life of the rotating equipment;

[0048] The evaluation module is used to use the overall health score and remaining life of the rotating equipment as the health evaluation result of the rotating equipment.

[0049] On the other hand, the present invention also provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the program, a rotating equipment health assessment method based on slow variables as described above is implemented.

[0050] The health assessment method, device and electronic device of rotating equipment based on slow variables provided by the present invention obtain the state monitoring data of each key equipment by determining the equipment importance coefficient and fault criticality coefficient of each key equipment in the rotating equipment, and then determine the single equipment health score of each key equipment. According to the equipment importance coefficient, fault criticality coefficient and single equipment health score of at least part of the key equipment, the overall health score of the rotating equipment is calculated, and the state monitoring data is extracted. The slow variables are determined from the extracted feature quantities, and the slow variables are input into the pre-built performance degradation model to predict the remaining life of the rotating equipment. Finally, the overall health score and the remaining life of the rotating equipment are used as the health assessment result of the rotating equipment. Since the overall health score of the rotating equipment can be calculated according to the equipment importance coefficient, fault criticality coefficient and single equipment health score of at least part of the key equipment, the overall health monitoring of the rotating equipment can be realized, and the monitoring is more comprehensive. At the same time, by extracting slow variables and using slow variables to predict the remaining life of the rotating equipment, the remaining life and health state of the rotating equipment can be predicted, so that preventive maintenance of the rotating equipment can be realized, and the accuracy and reliability of the health monitoring link are improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the present invention or the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0052] Figure 1 It is a flow chart of a method for health assessment of rotating equipment based on slow variables provided by an embodiment of the present invention;

[0053] Figure 2 It is a schematic diagram of the principle of the V-shaped overall evaluation framework;

[0054] Figure 3 It is a schematic diagram of the implementation principle of the health assessment method of rotating equipment based on slow variables;

[0055] Figure 4 is a structural schematic diagram of a rotating equipment health assessment device based on slow variables provided by an embodiment of the present invention;

[0056] Figure 5 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0057] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be clearly and completely described below in conjunction with the drawings of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0058] This embodiment relates to the field of equipment health monitoring, and can be specifically applied to health monitoring and intelligent operation and maintenance scenarios of rotating equipment such as centrifuges. At present, the health monitoring and intelligent operation and maintenance methods commonly used in the industry generally have the following steps:

[0059] (1) Data collection: Temperature, vibration, acceleration, voltage and other sensors are installed in key components of key equipment to collect data in real time during operation.

[0060] (2) Real-time data display: Build a three-dimensional model and use data-driven methods to reproduce and simulate the equipment operation status in a virtual three-dimensional space, or use video surveillance images to achieve visual observation of the equipment operation status.

[0061] (3) Real-time data analysis: Analyze the data collected by the sensor, generally making judgments within the set threshold, or using fault identification or abnormality judgment models.

[0062] (4) Post-operation data analysis: After operation, the data from each sensor of the equipment is analyzed to detect whether there is any abnormal data.

[0063] However, this type of health monitoring and intelligent operation and maintenance method only solves the automation of operation and maintenance and monitoring, displays the current operating status and the status of some sensor data of this operation, and can only judge the status of a certain sensor data. Through sensor data calculation, it can generally only obtain the health status of a certain component, such as the operation status of a motor, the working status of a certain bearing, etc., and cannot know the overall health status of the equipment; at the same time, this method can only judge the current health status of the equipment, cannot predict the future trend of the equipment data, cannot predict the remaining life and health status of the equipment, and cannot achieve preventive maintenance.

[0064] Based on this, this embodiment provides a solution to the above technical problem. Figures 1 to 5 The detailed scheme of the method, device and electronic device for health assessment of rotating equipment based on slow variables provided in the embodiments of the present invention is described.

[0065] Figure 1 It is a flow chart of a method for health assessment of rotating equipment based on slow variables provided in an embodiment of the present invention.

[0066] like Figure 1 As shown, the rotating equipment health assessment method based on slow variables provided by the embodiment of the present invention can be performed by a computer or server with data receiving and data processing functions. The method mainly includes the following steps:

[0067] Step 110: Determine the equipment importance coefficient and failure criticality coefficient of each key equipment in the rotating equipment respectively.

[0068] It can be understood that the equipment importance coefficient is mainly used to characterize the role and influence of each key equipment in the rotating equipment, and then each key equipment can be classified and evaluated for importance.

[0069] In practical applications, the equipment importance analysis method based on the hierarchical analysis method can be adopted to divide the rotating equipment according to the hierarchical structure of "system-subsystem-equipment-measurement point", build a structure tree, identify key equipment and key measurement points, and divide the components in the rotating equipment into three important levels according to expert experience and equipment structure, namely high level, medium level and low level.

[0070] Among them, high-level components refer to equipment that is crucial to the performance and safety of rotating equipment, such as bearings, spindles, motors, etc.; intermediate components refer to equipment that has a greater impact on the performance and safety of rotating equipment, such as shaft seals, sleeves, etc.; low-level components refer to equipment that has a smaller impact on the performance and safety of rotating equipment, such as pipes, valves, instruments, etc.

[0071] The fault criticality coefficient is mainly used to characterize the degree of harm to the entire rotating equipment when a key equipment fails. In practical applications, a fault tree can be constructed first, and then a fault criticality analysis can be performed based on the severity of the fault itself and the level in the fault tree, and the faults can be classified and sorted to determine the key faults and risk levels.

[0072] This embodiment specifically adopts a fault hazard analysis method based on the failure mode and effect analysis method, constructs a fault tree according to the causes and consequences of various possible faults of the rotating equipment, evaluates and scores the faults according to the probability and consequences of the faults, and then divides the faults into three hazard levels, namely high, medium and low.

[0073] Among them, high-level faults refer to faults that may cause the rotating equipment to shut down or be damaged, resulting in serious economic losses or personal injury, such as bearing damage, motor short circuit, etc.; intermediate-level faults refer to faults that may cause the performance of the rotating equipment to decline or become abnormal, resulting in certain economic losses or personal discomfort, such as bearing wear, shaft seal leakage, etc.; low-level faults refer to faults that may cause the noise or vibration of the rotating equipment to increase, resulting in minor economic losses or personal discomfort, such as loose pipes, stuck valves, etc.

[0074] Step 120: Obtain status monitoring data of each key device, and determine a single device health score of each key device based on the status monitoring data.

[0075] In this embodiment, the state monitoring data of key equipment may include data related to the working state of key equipment, such as temperature, pressure, current, vibration, etc. In practical applications, each key equipment may be used as a measuring point, and the above state monitoring data may be obtained through sensors associated with each measuring point.

[0076] It is understandable that the single device health score is mainly used to characterize the local health status of each key device itself. The higher the single device health score, the more normal and stable the operation process of the key device is.

[0077] Step 130: Calculate the overall health score of the rotating equipment according to the equipment importance coefficient, the failure criticality coefficient and the single equipment health score of at least some of the key equipment.

[0078] It can be understood that the overall health score of the rotating equipment can represent the overall health status of the rotating equipment. The overall health score is a comprehensive score result obtained by comprehensively considering the importance of each key equipment, the degree of failure hazard, and the health status of the equipment itself.

[0079] Step 140: extract features from the condition monitoring data and determine slow variables from the extracted features.

[0080] It should be noted that slow variables refer to characteristic quantities that are stable in the short term but change slowly in the long term, such as changes in measurement point parameters caused by wear, aging, etc. In practical applications, some filtering methods or smoothing methods can be used to extract slow variables, such as Kalman filtering, moving average method, etc.

[0081] Step 150: Input the slow variables into the pre-built performance degradation model to predict the remaining life of the rotating equipment.

[0082] It should be noted that the remaining life refers to the length of time that the rotating equipment can still work normally in the current state. Based on the usage of the rotating equipment, the remaining number of uses and seconds are predicted to obtain the remaining life.

[0083] Step 160: The overall health score and remaining life of the rotating equipment are used as the health assessment result of the rotating equipment.

[0084] The solution provided in this embodiment can reflect the overall health status of the rotating equipment through the overall health score, and can predict the normal working conditions of the rotating equipment in the future through the remaining life, thereby realizing the overall health assessment and status prediction functions of the rotating equipment, thereby improving the comprehensiveness and reliability of the equipment health monitoring link.

[0085] In one embodiment, the equipment importance coefficient and the failure criticality coefficient of each key equipment in the rotating equipment are determined respectively, specifically including:

[0086] The first step is to construct a structure tree based on the structural hierarchy of the rotating equipment.

[0087] The structure tree is a graphical tool used to represent the hierarchical structure and relationships of rotating equipment. It can divide the structure of rotating equipment from different perspectives such as function, location, and components. The structure tree can help analyze the failure mode, failure impact, and failure propagation of rotating equipment, and can provide data basis for building a fault tree.

[0088] In a specific implementation, a structure tree is constructed according to the structural hierarchical relationship of the rotating equipment, which specifically includes:

[0089] Firstly, the rotating equipment is subdivided according to the structural hierarchy, and the subordinate relationship between each layer of the structure is established.

[0090] Then, each layer structure is used as a tree node, and the weight value of each tree node is determined according to the relative importance of each tree node in the parent node.

[0091] Finally, a structure tree is established based on the subordinate relationships, tree nodes, and the weight values ​​of the tree nodes.

[0092] In practical applications, in order to clarify the specific composition, key equipment and key measuring points of rotating equipment, rate the importance of each key equipment, and provide a basis for the subsequent construction of a fault tree, this embodiment establishes a structure tree of rotating equipment, and specifically uses the hierarchical analysis method to model the structure tree of rotating equipment, subdivides the rotating equipment according to the hierarchical relationship of "system-subsystem-equipment-measuring point", and constructs the subordinate relationship between each measuring point and the system.

[0093] In this embodiment, the rotating equipment is divided according to function, and consists of five main subsystems: the rotating subsystem, the transmission subsystem, the drive subsystem, the auxiliary subsystem, and the safety monitoring subsystem. The safety monitoring subsystem is a high-level software system, which is not included in the operation and maintenance subject and structure tree of the rotating equipment, but its monitoring and alarm results will serve as an important reference and standard for health monitoring and operation and maintenance.

[0094] In some embodiments, the structure tree can be divided according to three modes: function, position, and structure. This embodiment mainly adopts a function-based division method. Combined with a structure-based division method, the rotating equipment can be divided into a rotating subsystem, a transmission subsystem, a drive subsystem, and an auxiliary subsystem, and each subsystem can be further subdivided.

[0095] Each node in the structure tree has a weight value, which indicates the relative importance of the node in its parent node. Through the structure tree, the global weight value of each node can be calculated, that is, the relative importance of the node in the entire system. The larger the global weight value, the more important the node device is.

[0096] Based on the structure tree of rotating equipment, key equipment and key measuring points can be further determined. For cost control and sensor installation feasibility considerations, not all physical quantities of all structures need corresponding sensor deployment and data monitoring. It is necessary to identify all key equipment in the structure tree, and identify the physical quantities and sensor deployment plans that need to be monitored based on the structural characteristics. If the current sensor deployment plan cannot meet the monitoring needs of the key equipment's measurement points, the sensor deployment plan needs to be further optimized.

[0097] The following is the process of identifying some key equipment and determining the sensor deployment plan:

[0098] In the rotating subsystem, the swing arm and the dynamic balancing adjustment device determine the monitoring range of the required unbalanced force of the rotating subsystem, and directly or indirectly determine the allowable unbalanced force of the rotating equipment test. Their damage will directly affect the normal use of the rotating equipment. Therefore, the swing arm and the dynamic balancing adjustment device are one of the key equipment, and the key measuring points are the strain between the swing arm and the hanging basket, the installation vibration at the typical vibration position of the swing arm, and the unbalanced force in the middle of the swing arm.

[0099] In the transmission subsystem, the shaft system is the core part of the rotating equipment, and the shaft system includes the bearing system, the main shaft and the coupling. The bearing system is the key component connecting the main shaft and the base, and its performance directly affects the stability of the rotating equipment during operation. All equipment in the shaft system is key equipment. The key measurement points include the spindle swing, guide bearing temperature, guide bearing vibration, oil station temperature, and the fault word and fault alarm provided by the safety monitoring word system.

[0100] In the drive subsystem, the main motor and its auxiliary cooling system are key equipment that determine whether the power is sufficient. Key measurement points include motor temperature, motor speed, output torque, cooling water flow, cooling water temperature, etc.

[0101] The second step is to construct a fault tree based on the logical relationship between the failure events of each key device in the structure tree.

[0102] In a specific implementation, a fault tree is constructed according to the fault logical relationship of each key device in the structure tree, including:

[0103] First, the failure causes of rotating equipment are decomposed into fault events corresponding to key equipment at different levels, and the logical relationship between each fault event is determined;

[0104] Then, according to the degree of harm of each fault event, the weight value of each fault event is determined;

[0105] Finally, a fault tree is established based on key equipment at different levels, fault events, logical relationships, and weight values ​​of fault events.

[0106] It can be understood that fault tree analysis is a system reliability analysis method based on logical operations. It takes the top event of the rotating equipment as the root node, and connects it with logical gates such as AND gates, OR gates, and NOT gates to decompose the failure causes of the rotating equipment into intermediate events and bottom events at different levels to form a fault tree. The fault tree can be used to analyze the failure probability, sensitivity, and importance of the rotating equipment, and then determine the weak links of the rotating equipment and the specific causes of a certain fault.

[0107] This embodiment mainly combines the structure tree of the rotating equipment to build a fault tree, including the key measurement points of all key equipment, and then identifies the main fault of the key equipment. In practical applications, this embodiment can realize two aspects of fault diagnosis and alarm functions: one is the fault diagnosis of the whole process of the rotating equipment, mainly for the abnormal cause analysis and fault location of the start-up and shutdown process; the other is the fault diagnosis of the rotating equipment itself, establishing a fault diagnosis calculation model based on the fault tree, covering the typical faults of key components, key subsystems and the whole machine.

[0108] The third step is to determine the equipment importance coefficient of each key equipment according to the weight value of each key equipment in the structure tree.

[0109] It is understandable that identifying key equipment and key measuring points can provide data basis for sorting out key faults and is a reference for constructing fault trees. However, in order to obtain the overall health score of rotating equipment, it is necessary to further subdivide the key equipment and give the equipment importance parameters of each key equipment.

[0110] Each tree node in the structure tree of rotating equipment should have a weight value, which indicates the relative importance of the tree node in its parent node. The equipment importance coefficient of the tree node can be initially rated by the following method:

[0111] First, the rating method based on numerical ranking: according to the size of the equipment importance value, sort it from large to small or from small to large to obtain the equipment importance ranking table.

[0112] Second, the rating method based on grade division: According to the distribution characteristics of the equipment importance values, the equipment is graded according to certain standards or thresholds to obtain an equipment importance grade table.

[0113] Third, the rating method based on comprehensive evaluation: According to the multiple importance index values ​​of the equipment, the relative weights and comprehensive scores are comprehensively considered to evaluate and obtain the equipment comprehensive importance evaluation table.

[0114] In this embodiment, the weight value is divided into three levels based on the rating method of comprehensive evaluation, which is given by experts or historical data. Some methods can be used for calculation or adjustment, such as hierarchical single sorting method, hierarchical total sorting method, consistency test method, etc. In the early stage of rotating equipment application, due to the lack of fault data and operation data, the weight value is determined by experts based on the structure and experience of the rotating equipment. The weight value of the equipment at this tree node, that is, the relative importance of the tree node in the entire system, is mainly determined based on the importance index method based on the equipment failure rate, the consequences of equipment failure, and the propagation of equipment failure. The larger the global weight value, the more important the tree node. In practical applications, the nodes can be divided into three important levels according to the global weight value, namely, high level (weight value within 0.4 to 1), medium level (weight value within 0.2 to 0.4), and low level (weight value within 0 to 0.2).

[0115] After accumulating enough measurement point data and failure experience, or simulating and calculating based on the mechanism simulation model of the rotating equipment, the weight value of the tree node can be updated, and its contribution to the system failure rate or failure time, such as the failure rate importance coefficient, the failure time importance coefficient, etc., can be calculated according to the failure rate of the tree node; according to the failure consequences of the tree node, its impact on the system function, economy, environment, etc. can be evaluated, such as the failure consequence importance coefficient, the failure cost importance coefficient, etc.; according to the fault propagation path of the tree node, its impact range and intensity on other equipment or the entire system can be determined, such as the fault propagation importance coefficient, the fault impact importance coefficient, etc.

[0116] After obtaining the above-mentioned multiple importance coefficients, the equipment importance coefficient of each key equipment can be obtained through comprehensive analysis.

[0117] The fourth step is to determine the failure criticality coefficient of each key device according to the weight value of each key device in the fault tree under different failure events.

[0118] In this embodiment, after obtaining the fault tree of the rotating equipment, similar to the structure tree, in order to obtain the overall health score of the rotating equipment, it is necessary to calculate the fault criticality parameters. Specifically, the weight value of each fault event node can be given in the fault tree, and the fault criticality analysis method can be used to obtain the fault criticality parameters.

[0119] Fault criticality analysis is based on the failure mode impact analysis, introduces criticality indicators, and conducts quantitative or qualitative evaluation of each failure mode to determine its impact on the safety and reliability of the entire rotating equipment. Fault criticality analysis usually uses the risk matrix method or risk priority number method to evaluate the risk level and formulate corresponding risk control measures.

[0120] If the rotating equipment manufacturer or equipment producer can provide an FMECA report or FMEA report, the failure criticality can be determined directly based on the above report, or the failure criticality coefficient can be determined based on the risk matrix method and risk priority.

[0121] The risk matrix method is a qualitative risk assessment method that takes the frequency of occurrence of failure modes and the severity of consequences as two dimensions to construct a risk matrix table and determine the risk level based on the combination of the two. This embodiment specifically divides the risk level into high risk, medium risk and low risk, and corresponds to different treatment methods.

[0122] The risk priority number method is a quantitative risk assessment method that takes the frequency of occurrence of failure modes, severity of consequences, and probability of detection as three factors, and calculates their product to obtain the risk priority number. The larger the risk priority number, the higher the risk, and the priority should be given. It is usually necessary to set a risk priority number threshold. Failure modes exceeding this threshold require improvement measures, and failure modes below this threshold can be accepted or monitored. Using the risk priority number can obtain a more accurate failure hazard coefficient, but it relies heavily on FMECA reports, or requires a large amount of data and simulation tests to determine the frequency of occurrence of failure modes, severity of consequences, and probability of detection.

[0123] In view of the current status of the structure and engineering of general rotating equipment, this embodiment uses the risk matrix method and refers to the idea of ​​the risk priority number method. The risk level of the fault event node in the fault tree is given by expert experience. The frequency of occurrence, severity of consequences and detection probability of the fault mode are comprehensively considered to give a reference value of the risk priority number. The overall principle is that the shallower the node level, the greater the detection probability and the higher the severity. This reference value is used as the fault criticality coefficient to calculate the overall health score of the rotating equipment.

[0124] It should be noted that, on the one hand, the rotating equipment is subdivided based on the hierarchical analysis method to form a rotating equipment structure tree; on the other hand, various possible faults of the rotating equipment are constructed into a fault tree according to the causes and consequences based on the failure mode and effect analysis method. Based on the structure tree and the fault tree, a V-shaped overall evaluation framework can be constructed. The V-shaped overall evaluation framework can be found in Figure 2 , taking the equipment importance coefficient and fault criticality coefficient as the horizontal axis and vertical axis respectively, a health status assessment matrix is ​​constructed to determine the overall health score of the rotating equipment.

[0125] In one embodiment, the single device health score of each key device is determined based on the status monitoring data, specifically including:

[0126] The first step is to obtain the historical measurement point data of each key equipment.

[0127] It is understandable that historical measurement point data are some status monitoring data detected by sensors deployed at key measurement points in the previous period of time, such as temperature, pressure, current, vibration, etc. These data reflect the working status and performance of rotating equipment and are an important basis for operation and maintenance. By collecting and accumulating historical measurement point data in real time and regularly, a sufficiently large data set can be formed to provide a basis for subsequent analysis and prediction.

[0128] In the second step, the historical measurement point data are analyzed in time series and frequency domain, and feature extraction is performed to obtain the equipment characteristic parameters.

[0129] It can be understood that time domain analysis refers to the analysis of the temporal variation patterns of historical measurement point data, such as mean, variance, peak, waveform, etc. Frequency domain analysis refers to the analysis of the distribution characteristics of historical measurement point data in frequency, such as power spectrum, harmonics, noise, etc.

[0130] Feature extraction refers to extracting equipment characteristic parameters related to the working status and performance of rotating equipment from historical measurement point data. It can be achieved using some machine learning methods or feature engineering methods, such as kurtosis, skewness, margin, principal component analysis, linear discriminant analysis, etc. In practical applications, the analysis and extraction links can be selected and optimized according to different types of data and models of different targets to improve the representativeness and discrimination of equipment characteristic parameters.

[0131] The third step is to establish a health assessment model based on the equipment characteristic parameters.

[0132] When evaluating the health of a single device, the essence is to subdivide the key devices into a set consisting of multiple measurement points. It is also necessary to comprehensively consider the joint impact of the health status of multiple measurement points under a single device. The following methods can be used to obtain the health score of a single device:

[0133] The first method is the weighted summation of indicators: a weight is assigned to each performance indicator collected at each measurement point, and then the value of each performance indicator is multiplied by the corresponding weight and summed to obtain a comprehensive score. The weight allocation can be based on expert knowledge, historical data analysis, or model-based importance analysis.

[0134] The second type is scoring based on abnormality: accumulate historical measurement point data, establish a normal data model for key equipment, and calculate the abnormality between the data at each moment and the normal data model. This abnormality can be used as a single device health score for key equipment. A higher abnormality indicates that the working status of the key equipment may be more unhealthy.

[0135] The third type is threshold-based scoring: a series of thresholds are set to represent different levels of health status of the measuring points. According to the relationship between the performance indicator data of the key equipment and these thresholds, the key equipment is divided into different status levels, and a corresponding score is assigned to each status level.

[0136] The fourth type is machine learning model prediction: using machine learning models, such as regression, classification, or clustering models, to predict the future health status of equipment through historical measurement point data. According to the prediction results output by the model, key equipment is given corresponding scores.

[0137] In the fourth step, the condition monitoring data is input into the health assessment model to obtain the single equipment health score of the key equipment output by the health assessment model.

[0138] In this embodiment, a machine learning model prediction scheme is adopted to build a health assessment model through equipment characteristic parameters and preliminary training. The health assessment model takes status monitoring data as input and single equipment health score as output, which can more conveniently realize the health assessment of each key equipment.

[0139] It should be noted that the first condition for long-term health monitoring of equipment is to identify slow variables. This embodiment sorts out the slow variables of various key equipment and key measuring points based on the characteristics of rotating equipment and the operation and maintenance experience of rotating structures. Taking the rotating equipment as a centrifuge as an example, the corresponding relationship between key equipment, fault subjects, key faults and slow variables in some subsystems is shown in Table 1 below:

[0140] Table 1 Key information record of some subsystems in the centrifuge

[0141]

[0142]

[0143] In one embodiment, the performance degradation model may be established through the following process:

[0144] The first step is to obtain the historical measurement point data of each key equipment.

[0145] In the second step, the historical measurement point data are analyzed in time series and frequency domain, and feature extraction is performed to obtain the equipment characteristic parameters.

[0146] The third step is to establish a prediction model based on the equipment characteristic parameters.

[0147] The fourth step is to input the slow variables into the prediction model to obtain the parameter change trend within the future preset time period output by the prediction model.

[0148] The fifth step is to establish a performance degradation model based on the slow variables and parameter change trends.

[0149] It should be noted that, in this embodiment, based on the historical measurement point data obtained through long-term monitoring and the extracted equipment characteristic parameters, a performance degradation model capable of evaluating the degree of performance degradation and the remaining life can be established.

[0150] The performance degradation model is a model that evaluates the performance degradation of rotating equipment based on slow variables and set life standards, and gives the remaining life of the rotating equipment and recommended maintenance modes. In practical applications, the overall health score of the rotating equipment can be adjusted in combination with the health assessment model. For example, the performance degradation model can be established based on the long short-term memory network, and the long short-term memory network can be used to extract features and sequence model the vibration signals of key structures, so as to predict the remaining life of key structures.

[0151] See also Figure 3 After the prediction model and the health evaluation model are constructed in this embodiment, the prediction model and the health evaluation model can be used to analyze the real-time measurement point data (i.e., the status monitoring data) to provide an evaluation of the operation status of each key equipment in the test, and to achieve data trend estimation and recommended maintenance operations. By using the prediction model and the health evaluation model to analyze the real-time measurement point data, it is possible to promptly discover and handle anomalies or faults, and provide an evaluation of the operation status of key equipment, such as normal operation, need for adjustment, need for maintenance, etc. At the same time, it is also possible to estimate the data trend in the future based on the trend of data changes, such as stability, increase, decrease, etc. Based on these analyses and predictions, recommended maintenance operations can also be given, such as continued observation, adjustment of parameters, replacement of parts, etc.

[0152] At the same time, the performance degradation model can be used to give the remaining life of the rotating equipment, and combined with the health assessment model to give the health score of the single equipment, and then the overall health score can be obtained. In practical applications, the overall health score can be expressed as a score from 0 to 100, such as 100 points for intact and 0 points for complete damage.

[0153] In this embodiment, the overall health score of the rotating equipment can be calculated by the following formula:

[0154]

[0155] Among them, H is the overall health score of the rotating equipment, d is the equipment importance coefficient of a certain key equipment, h is the failure criticality coefficient of a certain key equipment, g is the single equipment health score of a certain key equipment, and 100 is the highest score of the single equipment health score.

[0156] In one embodiment, the above-mentioned rotating equipment health assessment method based on slow variables may further include:

[0157] First, the overall health score of the rotating equipment is compared with the preset score threshold to obtain a comparison result.

[0158] Then, based on the comparison results, an equipment operation and maintenance plan is generated.

[0159] In this embodiment, the preset scoring threshold can be reasonably set according to actual needs. For example, the preset scoring threshold can be set to 80 points. In actual applications, if the comparison result is that the overall health score is above the preset scoring threshold, it means that the rotating equipment is operating normally, and the rotating equipment can continue to operate according to the current operating conditions; if the comparison result is that the overall health score is lower than the preset scoring threshold, it means that the rotating equipment is operating abnormally, and the rotating equipment can be shut down, or the single device health score of each key equipment in the rotating equipment can be used to further determine which specific key equipment has an abnormality, and according to the target key equipment with the abnormality, the corresponding equipment operation and maintenance plan can be allocated.

[0160] In one embodiment, the above-mentioned rotating equipment health assessment method based on slow variables may further include:

[0161] Determine the maintenance time and maintenance plan for the rotating equipment based on the remaining life of the rotating equipment.

[0162] In this embodiment, on the one hand, a maintenance duration can be set in advance, and after determining the length of time that the current rotating equipment can continue to work normally based on the remaining life, the normal working critical moment of the rotating equipment is calculated, that is, the moment when normal work ends, and then the time point before the critical moment with a maintenance duration is used as the maintenance time of the rotating equipment.

[0163] On the other hand, multiple time length intervals of remaining life can be set in advance, and maintenance plans corresponding to different time length intervals can be set. Subsequently, the target interval to which the time length belongs can be determined based on the time length corresponding to the remaining life of the current rotating equipment, and the maintenance plan corresponding to the target interval can be used as the maintenance plan for the current rotating equipment.

[0164] In summary, the health assessment method for rotating equipment based on slow variables provided by the embodiment of the present invention can extract slow variables based on the structural characteristics and data features of the equipment, and conduct long-term monitoring of slow variables. It uses artificial intelligence technology to train prediction models, predict and extrapolate slow variable data, and thus achieve prediction of equipment life and health status, solving the problem that traditional health monitoring solutions cannot achieve preventive maintenance. At the same time, this embodiment establishes a health evaluation system based on fault trees and structure trees, which can systematically and structuredly make an intuitive evaluation of the overall health status of rotating equipment, and give maintenance and operation suggestions based on the evaluation results, solving the problem of not being able to know the overall health status of the equipment.

[0165] Based on the same general inventive concept, the present invention also protects a rotating equipment health assessment device based on slow variables. The rotating equipment health assessment device based on slow variables provided by the present invention is described below. The rotating equipment health assessment device based on slow variables described below and the rotating equipment health assessment method based on slow variables described above can be referenced to each other.

[0166] Figure 4 It is a structural schematic diagram of a rotating equipment health assessment device based on slow variables provided in an embodiment of the present invention.

[0167] like Figure 4 As shown, the rotating equipment health assessment device based on slow variables provided by an embodiment of the present invention specifically includes:

[0168] The determination module 210 is used to respectively determine the equipment importance coefficient and the failure criticality coefficient of each key equipment in the rotating equipment.

[0169] The acquisition module 220 is used to acquire the status monitoring data of each key device and determine the single device health score of each key device according to the status monitoring data.

[0170] The calculation module 230 is used to calculate the overall health score of the rotating equipment according to the equipment importance coefficient, the failure criticality coefficient and the single equipment health score of at least some key equipment.

[0171] The extraction module 240 is used to extract features from the condition monitoring data and determine slow variables from the extracted features.

[0172] The prediction module 250 is used to input the slow variables into the pre-built performance degradation model to predict the remaining life of the rotating equipment.

[0173] The evaluation module 260 is used to use the overall health score and remaining life of the rotating equipment as the health evaluation result of the rotating equipment.

[0174] Regarding the device in the above embodiment, the specific manner in which each module performs operations has been described in detail in the embodiment of the method, and will not be elaborated again here.

[0175] Figure 5 It is a schematic diagram of the structure of an electronic device provided by an embodiment of the present invention.

[0176] like Figure 5As shown, the electronic device may include: a processor 310, a communication interface 320, a memory 330 and a communication bus 340, wherein the processor 310, the communication interface 320 and the memory 330 communicate with each other through the communication bus 340. The processor 310 may call the logic instructions in the memory 330 to execute the rotating equipment health assessment method based on slow variables provided in the above embodiments.

[0177] In addition, the logic instructions in the above-mentioned memory 330 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0178] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the rotating equipment health assessment method based on slow variables provided in the above-mentioned embodiments.

[0179] On the other hand, the present invention further provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the rotating equipment health assessment method based on slow variables provided in the above-mentioned embodiments.

[0180] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without paying creative labor.

[0181] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or some parts of the embodiments.

[0182] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for health assessment of rotating equipment based on slow variables, characterized in that: include: Determine the equipment importance coefficient and failure criticality coefficient of each key equipment in the rotating equipment respectively; Acquire status monitoring data of each of the key devices, and determine a single device health score of each of the key devices based on the status monitoring data; Calculate the overall health score of the rotating equipment according to the equipment importance coefficient, the failure criticality coefficient and the single equipment health score of at least part of the key equipment; Extracting features from the state monitoring data and determining slow variables from the extracted features; Inputting the slow variable into a pre-built performance degradation model to predict the remaining life of the rotating equipment; The overall health score and remaining life of the rotating equipment are used as the health assessment result of the rotating equipment.

2. The method for health assessment of rotating equipment based on slow variables according to claim 1, characterized in that: The respectively determining the equipment importance coefficient and the failure criticality coefficient of each key equipment in the rotating equipment comprises: Construct a structure tree based on the structural hierarchy of the rotating equipment; Constructing a fault tree according to the logical relationship between the fault events of each key device in the structure tree; Determine the equipment importance coefficient of each key equipment according to the weight value of each key equipment in the structure tree; According to the weight values ​​of each key device in the fault tree under different fault events, the failure criticality coefficient of each key device is determined.

3. The method for health assessment of rotating equipment based on slow variables according to claim 2 is characterized in that: The structure tree is constructed according to the structural hierarchical relationship of the rotating equipment, including: According to the structural hierarchy, the rotating equipment is subdivided and the subordinate relationship between each layer of structure is established; Taking each layer structure as a tree node, and determining the weight value of each tree node according to the relative importance of each tree node in the parent node; A structure tree is established according to the subordinate relationships, the tree nodes and the weight values ​​of the tree nodes.

4. The method for health assessment of rotating equipment based on slow variables according to claim 2, characterized in that: According to the fault logical relationship of each key device in the structure tree, a fault tree is constructed, including: Decompose the failure causes of rotating equipment into fault events corresponding to key equipment at different levels, and determine the logical relationship between each fault event; Determining a weight value of each fault event according to the degree of harm of each fault event; A fault tree is established according to the key devices at different levels, the fault events, the logical relationships and the weight values ​​of the fault events.

5. The method for health assessment of rotating equipment based on slow variables according to claim 1, characterized in that: Determining the single device health score of each of the key devices according to the status monitoring data includes: For each key equipment, obtaining historical measurement point data of the key equipment; Performing time series analysis and frequency domain analysis on the historical measurement point data respectively, and performing feature extraction to obtain equipment characteristic parameters; Establishing a health assessment model according to the characteristic parameters of the equipment; The status monitoring data is input into the health assessment model to obtain the single device health score of the key device output by the health assessment model.

6. The method for health assessment of rotating equipment based on slow variables according to claim 1, characterized in that: The performance degradation model is established through the following process: Obtaining historical measurement point data of each of the key devices; Performing time series analysis and frequency domain analysis on the historical measurement point data respectively, and performing feature extraction to obtain equipment characteristic parameters; Establishing a prediction model according to the characteristic parameters of the equipment; Inputting the slow variable into the prediction model to obtain the parameter change trend within a future preset time period output by the prediction model; A performance degradation model is established according to the slow variable and the parameter change trend.

7. The method for health assessment of rotating equipment based on slow variables according to any one of claims 1 to 6, characterized in that: The method further comprises: Comparing the overall health score of the rotating equipment with a preset score threshold to obtain a comparison result; An equipment operation and maintenance plan is generated according to the comparison result.

8. The method for health assessment of rotating equipment based on slow variables according to any one of claims 1 to 6, characterized in that: The method further comprises: The maintenance time and maintenance plan of the rotating equipment are determined according to the remaining life of the rotating equipment.

9. A rotating equipment health assessment device based on slow variables, characterized in that: include: A determination module, used to respectively determine the equipment importance coefficient and the failure criticality coefficient of each key equipment in the rotating equipment; An acquisition module, used to acquire status monitoring data of each of the key devices, and determine a single device health score of each of the key devices based on the status monitoring data; A calculation module, configured to calculate an overall health score of the rotating equipment according to equipment importance coefficients, failure criticality coefficients and single equipment health scores of at least some of the key equipment; An extraction module, used for extracting features from the state monitoring data and determining slow variables from the extracted features; A prediction module, used for inputting the slow variable into a pre-built performance degradation model to predict the remaining life of the rotating equipment; The evaluation module is used to use the overall health score and remaining life of the rotating equipment as the health evaluation result of the rotating equipment.

10. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein: When the processor executes the program, the rotating equipment health assessment method based on slow variables as described in any one of claims 1 to 8 is implemented.

Images