CN118030409A - Method and system for detecting abnormal operation performance of fan unit - Google Patents
Method and system for detecting abnormal operation performance of fan unit Download PDFInfo
- Publication number
- CN118030409A CN118030409A CN202410153117.7A CN202410153117A CN118030409A CN 118030409 A CN118030409 A CN 118030409A CN 202410153117 A CN202410153117 A CN 202410153117A CN 118030409 A CN118030409 A CN 118030409A
- Authority
- CN
- China
- Prior art keywords
- unit
- data
- sequence
- time
- wind pressure
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 230000002159 abnormal effect Effects 0.000 title claims abstract description 65
- 238000000034 method Methods 0.000 title claims abstract description 45
- 230000005856 abnormality Effects 0.000 claims abstract description 105
- 238000001514 detection method Methods 0.000 claims abstract description 84
- 238000010977 unit operation Methods 0.000 claims abstract description 64
- 238000012502 risk assessment Methods 0.000 claims abstract description 39
- 238000007781 pre-processing Methods 0.000 claims abstract description 15
- 230000009466 transformation Effects 0.000 claims description 13
- 238000000605 extraction Methods 0.000 claims description 12
- 238000009499 grossing Methods 0.000 claims description 8
- 238000010606 normalization Methods 0.000 claims description 8
- 238000010276 construction Methods 0.000 claims description 7
- 238000012549 training Methods 0.000 claims description 7
- 238000010801 machine learning Methods 0.000 claims description 6
- 238000007619 statistical method Methods 0.000 claims description 5
- 238000004140 cleaning Methods 0.000 claims description 4
- 238000010183 spectrum analysis Methods 0.000 claims description 4
- 238000004458 analytical method Methods 0.000 abstract description 14
- 238000011156 evaluation Methods 0.000 abstract description 8
- 230000008859 change Effects 0.000 abstract description 7
- 238000012544 monitoring process Methods 0.000 abstract description 7
- 230000006399 behavior Effects 0.000 description 7
- 238000012545 processing Methods 0.000 description 7
- 238000012423 maintenance Methods 0.000 description 6
- 238000001228 spectrum Methods 0.000 description 4
- 238000012706 support-vector machine Methods 0.000 description 4
- 238000004422 calculation algorithm Methods 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 3
- 230000007613 environmental effect Effects 0.000 description 3
- 230000008569 process Effects 0.000 description 3
- 230000001052 transient effect Effects 0.000 description 3
- 206010000117 Abnormal behaviour Diseases 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 230000015556 catabolic process Effects 0.000 description 2
- 238000006731 degradation reaction Methods 0.000 description 2
- 238000003745 diagnosis Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000007257 malfunction Effects 0.000 description 2
- 238000004445 quantitative analysis Methods 0.000 description 2
- 238000007637 random forest analysis Methods 0.000 description 2
- 230000003595 spectral effect Effects 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000013178 mathematical model Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 230000003449 preventive effect Effects 0.000 description 1
- 238000011002 quantification Methods 0.000 description 1
- 238000012731 temporal analysis Methods 0.000 description 1
- 238000000700 time series analysis Methods 0.000 description 1
- 238000002759 z-score normalization Methods 0.000 description 1
Landscapes
- Testing And Monitoring For Control Systems (AREA)
Abstract
The application provides a method and a system for detecting abnormal operation performance of a fan unit, and relates to the technical field of abnormality detection, wherein the method comprises the following steps: collecting real-time operation data, preprocessing to obtain trusted operation data, extracting unit operation characteristics to obtain a unit operation characteristic data set, collecting historical operation data, constructing a unit abnormality detection model to obtain unit abnormality information, performing unit operation risk assessment on the unit abnormality information, and performing unit potential fault early warning according to an operation risk assessment result. The application mainly solves the problems that the lack of real-time monitoring and analysis of the operation parameters can not find abnormal conditions in time and can not adapt to the change under various working conditions, thereby causing the inaccuracy of detection. The operation risk type judgment and the risk grade evaluation are carried out by presetting an operation risk evaluation constraint interval and combining abnormal information to carry out early warning, so that the operation efficiency and reliability of the unit are improved, and the occurrence and influence of potential faults are reduced.
Description
Technical Field
The application relates to the technical field of abnormality detection, in particular to a method and a system for detecting abnormal operation performance of a fan unit.
Background
Fan units are widely used in many fields, such as energy, construction, traffic, etc. However, the operation performance of the fan unit may be affected by various factors such as equipment aging, environmental changes, improper operation, etc., thereby causing abnormal performance. Conventional methods of performance anomaly detection are typically based on fixed thresholds or empirical determinations, but such methods may not be accurate and flexible enough for complex blower unit systems.
However, in the process of implementing the technical scheme of the embodiment of the application, the above technology is found to have at least the following technical problems:
the lack of real-time monitoring and analysis of operation parameters can not find abnormal conditions in time, and can not adapt to changes under various working conditions, so that the detection inaccuracy is caused.
Disclosure of Invention
The application mainly solves the problems that the lack of real-time monitoring and analysis of the operation parameters can not find abnormal conditions in time and can not adapt to the change under various working conditions, thereby causing the inaccuracy of detection.
In view of the foregoing, the present application provides a method and a system for detecting abnormal operation performance of a fan unit, and in a first aspect, the present application provides a method for detecting abnormal operation performance of a fan unit, where the method includes: based on the sensor module, collecting real-time operation data of the target unit in a preset period, wherein the real-time operation data comprise the rotating speed, power, flow and wind pressure data of the target unit; preprocessing the real-time operation data to obtain trusted operation data; extracting unit operation characteristics based on the trusted operation data to obtain a unit operation characteristic data set; collecting historical operation data of a target unit, and constructing a unit anomaly detection model; performing unit abnormality detection according to the unit abnormality detection model and the unit operation feature set to obtain unit abnormality information; and aiming at the unit abnormal information, performing unit operation risk assessment, and performing unit potential fault early warning according to an operation risk assessment result.
In a second aspect, the present application provides an operational performance anomaly detection system for a fan unit, the system comprising: the real-time operation data acquisition module is based on the sensor module and acquires real-time operation data of the target unit in a preset period, wherein the real-time operation data comprise the rotating speed, the power, the flow and the wind pressure data of the target unit; the preprocessing module is used for preprocessing the real-time operation data to obtain trusted operation data; the characteristic data set acquisition module is used for extracting unit operation characteristics based on the trusted operation data to obtain a unit operation characteristic data set; the abnormality detection model construction module is used for collecting historical operation data of the target unit and constructing a unit abnormality detection model; the unit abnormality information acquisition module is used for carrying out unit abnormality detection according to the unit abnormality detection model and combining the unit operation characteristic set to obtain unit abnormality information; and the fault early warning module is used for carrying out unit operation risk assessment aiming at the unit abnormal information and carrying out unit potential fault early warning according to an operation risk assessment result.
One or more technical schemes provided by the application have at least the following technical effects or advantages:
The application provides a method and a system for detecting abnormal operation performance of a fan unit, and relates to the technical field of abnormality detection, wherein the method comprises the following steps: collecting real-time operation data, preprocessing to obtain trusted operation data, extracting unit operation characteristics to obtain a unit operation characteristic data set, collecting historical operation data, constructing a unit abnormality detection model to obtain unit abnormality information, performing unit operation risk assessment on the unit abnormality information, and performing unit potential fault early warning according to an operation risk assessment result.
The application mainly solves the problems that the lack of real-time monitoring and analysis of the operation parameters can not find abnormal conditions in time and can not adapt to the change under various working conditions, thereby causing the inaccuracy of detection. The operation risk type judgment and the risk grade evaluation are carried out by presetting an operation risk evaluation constraint interval and combining abnormal information to carry out early warning, so that the operation efficiency and reliability of the unit are improved, and the occurrence and influence of potential faults are reduced.
The foregoing description is only an overview of the present application, and is intended to be implemented in accordance with the teachings of the present application in order that the same may be more clearly understood and to make the same and other objects, features and advantages of the present application more readily apparent.
Drawings
In order to more clearly illustrate the application or the technical solutions of the prior art, the following brief description will be given of the drawings used in the description of the embodiments or the prior art, it being obvious that the drawings in the description below are only exemplary and that other drawings can be obtained from the drawings provided without the inventive effort for a person skilled in the art.
FIG. 1 is a schematic flow chart of a method for detecting abnormal operation performance of a fan unit according to an embodiment of the present application;
FIG. 2 is a schematic flow chart of a method for feature extraction in a method for detecting abnormal operation performance of a fan unit according to an embodiment of the present application;
FIG. 3 is a schematic flow chart of a method for performing unit potential fault early warning in a method for detecting abnormal operation performance of a fan unit according to an embodiment of the present application;
Fig. 4 is a schematic structural diagram of an abnormal operation performance detection system for a fan unit according to an embodiment of the present application.
Reference numerals illustrate: the system comprises a real-time operation data acquisition module 10, a preprocessing module 20, a characteristic data set acquisition module 30, an abnormality detection model construction module 40, a unit abnormality information acquisition module 50 and a fault early warning module 60.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
The application mainly solves the problems that the lack of real-time monitoring and analysis of the operation parameters can not find abnormal conditions in time and can not adapt to the change under various working conditions, thereby causing the inaccuracy of detection. The operation risk type judgment and the risk grade evaluation are carried out by presetting an operation risk evaluation constraint interval and combining abnormal information to carry out early warning, so that the operation efficiency and reliability of the unit are improved, and the occurrence and influence of potential faults are reduced.
For a better understanding of the foregoing technical solutions, the following detailed description will be given with reference to the accompanying drawings and specific embodiments of the present invention:
Example 1
The method for detecting abnormal operation performance of the fan unit as shown in fig. 1 comprises the following steps:
based on the sensor module, collecting real-time operation data of the target unit in a preset period, wherein the real-time operation data comprise the rotating speed, power, flow and wind pressure data of the target unit;
Specifically, by collecting real-time operation data of the target unit in a preset period, key performance parameters including rotation speed, power, flow, wind pressure data and the like of the target unit can be obtained. These data can be used for subsequent analysis and anomaly detection. When real-time operation data is acquired, a high-precision sensor module is selected to obtain complete operation data, and then a reasonable preset period is set, for example, the operation period of a unit or 1h of a using peak period. By monitoring these performance parameters in real time, anomalies can be found in time. For example, if the rotational speed suddenly drops or the flow rate abnormally fluctuates, a malfunction or problem of the unit may be indicated. Similarly, abnormal changes in power and wind pressure may also indicate the occurrence of performance anomalies.
Preprocessing the real-time operation data to obtain trusted operation data;
specifically, preprocessing typically includes data cleansing, format conversion, outlier handling, etc., in order to extract trusted operating data from the raw data. The purpose of data cleansing is to remove or correct erroneous, abnormal or inaccurate data. For example, for data collected by a sensor, there may be some outliers, such as maxima, minima, or data that are clearly not practical. These outliers may be due to sensor failures, environmental disturbances, or other causes. Format conversion is the conversion of data into a format suitable for analysis as required. For example, some of the data collected by the sensors may be in the form of analog signals, which need to be converted to digital signals and appropriately quantified. Outlier handling is special handling for some outlier cases to ensure data integrity. For example, in some cases, the sensor may fail or malfunction suddenly, resulting in anomalies in the data collected.
Extracting unit operation characteristics based on the trusted operation data to obtain a unit operation characteristic data set;
Specifically, by extracting the unit operation characteristics, a simplified data representation reflecting the unit operation state can be obtained, which is helpful for subsequent anomaly detection and analysis. Meaningful features such as mean, variance, trend, frequency, etc. can be extracted from the raw data by statistical analysis, time-series analysis, frequency-domain analysis, for example, for rotational speed data, average rotational speed, rotational speed fluctuation rate, etc. features can be extracted; for power data, characteristics such as average power and power factor can be extracted, and for flow data, characteristics such as average flow and flow fluctuation rate can be extracted. These features can reflect the operating state and performance of the unit, select features associated with anomaly detection tasks when extracting unit operating features, and ensure that these features have adequate discrimination and stability.
Collecting historical operation data of a target unit, and constructing a unit anomaly detection model;
Specifically, by analyzing the historical data, the performance and the behavior mode of the unit in normal operation can be known, and a reference and comparison basis is provided for subsequent abnormality detection. When historical operation data are collected, reliable data sources are ensured, no missing or abnormal values exist, and the operation states of the machine set under different working conditions and time periods can be covered. Enough historical data is collected to better capture the operating characteristics and changes of the unit. The method for constructing the unit anomaly detection model can be selected according to specific application scenes and requirements. For example, a classifier model may be constructed using Support Vector Machines (SVMs), random forests, neural networks, and the like algorithms for identifying outlier data and predicting potential faults. These models can set reasonable thresholds and rules to distinguish between normal and abnormal behavior based on the normal behavior patterns of the historical data. By constructing the unit abnormality detection model, abnormality detection can be automatically performed, and manual intervention and erroneous judgment are reduced. The model can be continuously updated and adjusted according to actual conditions so as to adapt to the change of the unit performance and a new abnormal mode.
Performing unit abnormality detection according to the unit abnormality detection model and the unit operation feature set to obtain unit abnormality information;
Specifically, by comparing and analyzing the current unit operation feature data set with the abnormality detection model, abnormality information reflecting the unit operation state can be obtained. First, a current set of unit operation feature data is input into an anomaly detection model. The method comprises the steps of comparing and analyzing the unit operation characteristic data acquired in real time with the historical operation data. By comparing these data, a characteristic pattern that is inconsistent with the normal behavior pattern or that is out of the normal range can be identified. Then, based on the output of the abnormality detection model, it can be determined whether or not an abnormality exists. If the output of the model indicates that the current operating characteristic dataset deviates from the normal behavior pattern, then the unit can be considered to be abnormal. After the unit abnormal information is obtained, corresponding processing measures can be further adopted. For example, sending out an early warning signal, recording abnormal data, performing fault diagnosis, and the like. The information has important reference value for maintenance and management personnel, is helpful for timely finding and solving potential problems, and improves the operation efficiency and reliability of the fan unit.
And aiming at the unit abnormal information, performing unit operation risk assessment, and performing unit potential fault early warning according to an operation risk assessment result.
Specifically, through evaluating the operation risk of the unit, the current state and potential problems of the unit can be better known, and a basis is provided for preventive maintenance and fault early warning. The operation risk assessment is a comprehensive analysis and assessment process of unit abnormal information, and is used for determining the risk level and the potential fault possibility of unit operation. Including factors such as severity of abnormality, influence range, development tendency, etc. By comparing the current abnormal information with the historical data and other related information, the running risk of the unit can be estimated more accurately. In assessing the risk of operation, a combination of qualitative and quantitative methods may be employed. The qualitative method can evaluate the severity of abnormal information and the possibility of potential faults according to experience, expert judgment and unit operation characteristics. The quantitative method can utilize technologies such as data analysis, machine learning and the like to establish a mathematical model or algorithm, and carry out quantification and risk assessment on abnormal information. And according to the running risk assessment result, potential fault early warning of the unit can be performed. The early warning modes can be diversified, such as sending short messages or mails, displaying warning information on a monitoring system and the like through signals such as sound, light and electricity. The early warning content comprises information such as description of abnormal information, risk level, prediction of potential faults and the like, so that maintenance and management staff can know the state of the unit in time and take corresponding treatment measures.
Further, the method of the present application preprocesses the real-time operation data to obtain trusted operation data, and the method includes:
respectively cleaning the data of the rotating speed, the power, the flow and the wind pressure of the target unit, and removing abnormal values, missing values and repeated values to obtain complete operation data;
And carrying out smoothing treatment and normalization treatment on the complete operation data, and carrying out serialization treatment according to a time sequence to obtain a unit rotating speed sequence, a unit power sequence, a unit flow sequence and a unit wind pressure sequence to form the trusted operation data.
Specifically, data cleaning: removing abnormal values: and identifying and removing abnormal data points according to the normal range and rule of the rotating speed, the power, the flow and the wind pressure. The outliers may be due to sensor failures, environmental disturbances, or unit anomalies. Removing the missing value: it is checked whether missing values exist in the data, which may be due to sensor failures, data transmission problems, or other reasons. Depending on the particular situation, it may be selected to fill in missing values or to directly remove data points containing missing values. Removing the duplicate values: in a dataset there may be duplicate or redundant data points that have no practical significance for anomaly detection and need to be removed. Smoothing: and smoothing the cleaned data to eliminate random fluctuation and noise. The smoothing method includes moving average filtering, median filtering, etc., and may be selected according to the characteristics and needs of the data. The smoothing process helps to reduce noise interference and improve data quality. Normalization: and carrying out normalization processing on the data subjected to the smoothing processing, and converting the data into the same dimension or range so as to enable the data to have comparability. Methods of normalization include min-max normalization, Z-score normalization, and the like. By normalization processing, the influence of different physical dimensions on anomaly detection can be eliminated. And (3) serializing: and carrying out serialization processing on the normalized data according to a time sequence to form a unit rotating speed sequence, a unit power sequence, a unit flow sequence and a unit wind pressure sequence. The sequences can reflect the running states of the machine set at different time points and provide time sequence data for subsequent abnormality detection. Through the preprocessing steps, complete, accurate and comparable operation data can be obtained. The data can be used for subsequent unit operation feature extraction and abnormality detection model construction, and helps accurately identify abnormal conditions in unit operation.
Furthermore, the method of the application extracts the unit operation characteristic based on the trusted operation data to obtain the unit operation characteristic data set, and the method further comprises the following steps:
the unit operation characteristics comprise time domain characteristics, frequency domain characteristics and time-frequency domain characteristics;
Extracting a unit rotating speed sequence, a unit power sequence, a unit flow sequence and a unit wind pressure sequence based on the trusted operation data;
and respectively extracting time domain features, frequency domain features and time-frequency domain features according to the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence to obtain a unit operation feature data set.
Specifically, when the unit operation features are extracted, feature extraction can be performed from three aspects of time domain, frequency domain and time-frequency domain so as to comprehensively reflect the unit operation state. Extracting time domain features: and extracting time domain features based on the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence. The time domain features include statistics of mean, variance, standard deviation, maximum, minimum, etc. These features can reflect the stability, volatility, and periodicity of the unit operating conditions. Extracting frequency domain features: and carrying out frequency domain analysis such as Fourier transform on the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence, and converting the frequency domain analysis from a time domain to a frequency domain. In the frequency domain, the characteristics of frequency spectrum, power spectral density and the like can be extracted to know the operation characteristics of the unit under different frequencies. The frequency domain features help to identify whether the unit has vibration, unbalance, asymmetry and other problems. And (3) extracting time-frequency domain features: for time-frequency domain feature extraction, the method such as wavelet transformation, empirical mode decomposition and the like can be used for processing the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence. The time-frequency domain features can capture the characteristics of the signal at different times and frequencies, and are very useful for analyzing non-stationary signals and identifying emergencies. Through time-frequency analysis, abrupt change and transient behavior in unit operation can be detected. The extraction of the time domain features, the frequency domain features and the time-frequency domain features can comprehensively reflect the running state of the unit, and provide rich feature data for anomaly detection. The abnormal detection model constructed based on the characteristics can accurately identify abnormal behaviors, and the accuracy and reliability of abnormal detection are improved.
Further, as shown in fig. 2, the method of the present application performs time domain feature, frequency domain feature and time-frequency domain feature extraction according to the unit rotation speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence, and further includes:
Aiming at the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence, a time domain characteristic data set is obtained through statistical analysis;
Aiming at the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence, obtaining a frequency domain characteristic data set through Fourier transformation and spectrum analysis;
and acquiring a time-frequency domain characteristic data set by wavelet transformation aiming at the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence.
Specifically, for a unit rotating speed sequence, a power sequence, a flow sequence and an air pressure sequence, feature extraction is performed from three aspects of a time domain, a frequency domain and a time-frequency domain, and the time domain feature extraction is performed: for each sequence, a series of statistics, such as mean, variance, standard deviation, maximum, minimum, etc., are calculated to describe the distribution characteristics, volatility, and periodicity of the sequence. These statistics may reflect the operational stability and state changes of the unit. And (3) sorting the calculated statistic into a time domain feature data set through statistical analysis, and providing time domain feature data for subsequent anomaly detection. Extracting frequency domain features: each sequence is fourier transformed and transformed from the time domain to the frequency domain. In the frequency domain, parameters such as frequency spectrum, power spectral density and the like are analyzed to know the operating characteristics and energy distribution of the unit at different frequencies. The frequency domain characteristic data set can reveal whether the unit has the problems of vibration, unbalance, asymmetry and the like, and is helpful for abnormality detection and fault diagnosis. And (3) extracting time-frequency domain features: wavelet transformation is performed on each sequence, which is an effective method of processing non-stationary signals. The wavelet transform is capable of analyzing the signal in both the time and frequency domains, capturing local characteristics and variations of the signal. Through wavelet transformation, abrupt change and transient behavior in unit operation can be extracted, and a time-frequency domain characteristic data set is obtained. In summary, by combining the feature extraction methods of the time domain, the frequency domain and the time-frequency domain, a richer and comprehensive unit operation feature data set can be obtained. The feature data sets provide powerful support for the subsequent construction of the abnormality detection model, and are helpful for improving the accuracy and reliability of abnormality detection.
Furthermore, the method of the application collects the historical operation data of the target unit, builds the unit abnormality detection model, and also comprises the following steps:
According to historical operation data of a target unit, extracting operation characteristics of a sample unit and corresponding sample unit states, wherein the sample unit states comprise a normal state and an abnormal state;
and performing supervised training by combining a machine learning principle based on the operation characteristics of the sample unit and the state of the sample unit to obtain the unit abnormality detection model.
Specifically, according to historical operation data of a target unit, sample unit operation characteristics and corresponding sample unit states are extracted, and the sample unit operation characteristics are extracted: and extracting the operation characteristics of the sample unit from the historical operation data of the target unit. The features can comprise time domain features, frequency domain features, time-frequency domain features and the like, so that the extracted features can comprehensively reflect the running state of the unit, including normal state and abnormal state. Determining a sample unit state: and dividing the samples into a normal state and an abnormal state according to the actual performance and the state of the unit in the historical operation data. The normal state refers to a state when the unit is operating normally, and the abnormal state refers to a state when a fault, performance degradation or other abnormal condition occurs. And (3) supervised training: based on the extracted operation characteristics of the sample unit and the corresponding state of the sample unit, the machine learning principle is combined to perform supervised training. And selecting a proper machine learning algorithm, such as a Support Vector Machine (SVM), a random forest, a neural network and the like, for constructing an anomaly detection model. And adjusting model parameters by using the marked training data, and training an anomaly detection model.
Furthermore, the method of the application carries out unit abnormality detection according to the unit abnormality detection model and in combination with the unit operation feature set to obtain unit abnormality information, and further comprises the following steps:
Based on the group of operation feature sets, respectively acquiring time domain feature data, frequency domain feature data and time-frequency domain feature data of rotating speed, power, flow and wind pressure;
And according to the unit abnormality detection model, unit abnormality detection is respectively carried out on the time domain feature data, the frequency domain feature data and the time-frequency domain feature data of the rotating speed, the power, the flow and the wind pressure, so as to obtain unit rotating speed abnormality information, unit power abnormality information, unit flow abnormality information and unit wind pressure abnormality information.
Specifically, based on a unit operation feature set, time domain features, frequency domain features and time-frequency domain features of rotation speed, power, flow and wind pressure are respectively extracted, and then anomaly detection is carried out on each item of feature data by using an anomaly detection model, so that more specific and detailed anomaly information can be obtained. Extracting time domain feature data: and calculating statistics such as mean value, variance, standard deviation, maximum value, minimum value and the like of time domain characteristic data of the rotating speed, the power, the flow and the wind pressure. These time domain features can reflect the stability, volatility, and periodicity of the unit operating conditions. Extracting frequency domain characteristic data: and carrying out Fourier transformation and spectrum analysis on the frequency domain characteristic data of the rotating speed, the power, the flow and the wind pressure. Parameters such as frequency spectrum, power spectrum density and the like are extracted, and the operation characteristics and energy distribution of the unit under different frequencies are known. These frequency domain features help identify whether the unit has vibration, imbalance, asymmetry, etc. Extracting time-frequency domain feature data: and carrying out wavelet transformation on the time-frequency domain characteristic data of the rotating speed, the power, the flow and the wind pressure. Wavelet transformation enables simultaneous analysis of the characteristics of the signal in both the time and frequency domains, capturing local characteristics and variations of the signal. Through wavelet transformation, mutation and transient behavior in unit operation can be extracted, and powerful support is provided for abnormality detection. Abnormality detection is performed: based on the abnormality detection model, abnormality detection is respectively carried out on time domain characteristic data, frequency domain characteristic data and time-frequency domain characteristic data of the rotating speed, the power, the flow and the wind pressure. And the model judges and classifies the new operation data according to the sample characteristic learning of the normal state and the abnormal state. And judging whether an abnormal condition exists or not according to the result output by the model for each characteristic data. And setting a reasonable threshold or rule according to the actual situation so as to accurately judge the occurrence of the abnormality. Outputting abnormal information: and outputting unit rotating speed abnormality information, unit power abnormality information, unit flow abnormality information and unit wind pressure abnormality information according to the abnormality detection result. Such information has important reference value for maintenance and management personnel. The anomaly information may include the time at which the anomaly occurred, a characterization of the anomaly, and possibly the cause, etc., to facilitate timely discovery and resolution of potential problems.
Further, as shown in fig. 3, the method of the present application performs unit operation risk assessment for the unit abnormal information, and performs unit potential fault early warning according to the operation risk assessment result, and further includes:
Presetting an operation risk assessment constraint interval;
Based on the operation risk assessment constraint interval, combining the unit rotating speed abnormality information, the unit power abnormality information, the unit flow abnormality information and the unit wind pressure abnormality information, judging an operation risk type and evaluating a risk level to obtain an operation risk assessment result;
And carrying out potential fault early warning on the unit according to the running risk assessment result.
Specifically, a running risk assessment constraint interval is preset: and presetting a constraint interval for running risk assessment according to the actual condition and the historical data. These constraint intervals may be normal ranges and abnormal thresholds based on parameters such as rotational speed, power, flow, wind pressure, etc. The constraint interval is set by comprehensively considering the performance parameters, the safety requirements and the experience data of the unit so as to ensure the validity and the accuracy of the evaluation. And carrying out operation risk type judgment by combining the abnormal information: based on a preset constraint interval, the operation risk type is judged by combining the acquired unit rotating speed abnormality information, unit power abnormality information, unit flow abnormality information and unit wind pressure abnormality information. The operational risk is classified into different types, such as performance degradation, unstable operation, potential failure, etc., according to the severity and characteristics of the anomaly. And (3) performing risk grade evaluation: and further evaluating the risk level of each risk type according to the determined risk type. The risk level may be based on a number of factors, such as anomaly duration, degree of parameter deviation, potential impact, and the like. The risk level can be divided into different levels of low risk, medium risk, high risk and the like, and the basis is provided for subsequent decisions. And (3) outputting an operation risk assessment result: and sorting the determined risk type and the estimated risk level into an operation risk estimation result. And (3) performing unit potential fault early warning: and carrying out potential fault early warning on the unit for the abnormal conditions of high risk and medium risk according to the running risk assessment result. The pre-warning information may include anomaly type, risk level, possible cause and countermeasure, etc. Through an early warning mechanism, maintenance and management personnel are informed in time to take corresponding measures, such as checking equipment, adjusting parameters or arranging maintenance and the like, so that the occurrence probability and the influence degree of potential faults are reduced.
Example two
Based on the same inventive concept as the operation performance abnormality detection method for a fan unit of the foregoing embodiment, as shown in fig. 4, the present application provides an operation performance abnormality detection system for a fan unit, the system comprising:
the real-time operation data acquisition module 10 is based on a sensor module, and the real-time operation data acquisition module 10 acquires real-time operation data of a target unit in a preset period, including data of the rotating speed, the power, the flow and the wind pressure of the target unit;
the preprocessing module 20 is used for preprocessing the real-time operation data to obtain trusted operation data;
The feature data set acquisition module 30 is used for extracting unit operation features based on the trusted operation data to obtain a unit operation feature data set;
The abnormality detection model construction module 40 is used for collecting historical operation data of the target unit and constructing a unit abnormality detection model;
The unit abnormality information acquisition module 50 is used for carrying out unit abnormality detection according to the unit abnormality detection model and the unit operation characteristic set to obtain unit abnormality information;
The fault early warning module 60 is configured to perform a unit operation risk assessment for the unit abnormality information, and perform a unit potential fault early warning according to an operation risk assessment result.
Further, the system further comprises:
The trusted data acquisition module is used for respectively cleaning the rotating speed, the power, the flow and the wind pressure data of the target unit, removing abnormal values, missing values and repeated values and obtaining complete operation data; and carrying out smoothing treatment and normalization treatment on the complete operation data, and carrying out serialization treatment according to a time sequence to obtain a unit rotating speed sequence, a unit power sequence, a unit flow sequence and a unit wind pressure sequence to form the trusted operation data.
Further, the system further comprises:
The characteristic data set acquisition module is characterized in that the unit operation characteristics comprise time domain characteristics, frequency domain characteristics and time-frequency domain characteristics; extracting a unit rotating speed sequence, a unit power sequence, a unit flow sequence and a unit wind pressure sequence based on the trusted operation data; and respectively extracting time domain features, frequency domain features and time-frequency domain features according to the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence to obtain a unit operation feature data set.
Further, the system further comprises:
the feature extraction module is used for obtaining a time domain feature data set through statistical analysis aiming at the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence; aiming at the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence, obtaining a frequency domain characteristic data set through Fourier transformation and spectrum analysis; and acquiring a time-frequency domain characteristic data set by wavelet transformation aiming at the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence.
Further, the system further comprises:
The abnormal detection model acquisition module is used for extracting the operation characteristics of the sample unit and the corresponding sample unit states according to the historical operation data of the target unit, wherein the sample unit states comprise a normal state and an abnormal state; and performing supervised training by combining a machine learning principle based on the operation characteristics of the sample unit and the state of the sample unit to obtain the unit abnormality detection model.
Further, the system further comprises:
The abnormal information acquisition module is used for respectively acquiring time domain characteristic data, frequency domain characteristic data and time-frequency domain characteristic data of rotating speed, power, flow and wind pressure based on the unit operation characteristic set; and according to the unit abnormality detection model, unit abnormality detection is respectively carried out on the time domain feature data, the frequency domain feature data and the time-frequency domain feature data of the rotating speed, the power, the flow and the wind pressure, so as to obtain unit rotating speed abnormality information, unit power abnormality information, unit flow abnormality information and unit wind pressure abnormality information.
Further, the system further comprises:
the fault early warning module is used for presetting an operation risk assessment constraint interval; based on the operation risk assessment constraint interval, combining the unit rotating speed abnormality information, the unit power abnormality information, the unit flow abnormality information and the unit wind pressure abnormality information, judging an operation risk type and evaluating a risk level to obtain an operation risk assessment result; and carrying out potential fault early warning on the unit according to the running risk assessment result.
The foregoing detailed description of the method for detecting abnormal operation performance of a fan unit will be clear to those skilled in the art, and the system disclosed in this embodiment is described more simply because it corresponds to the method disclosed in the embodiment, and the relevant points refer to the description of the method section.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (8)
1. The method for detecting the abnormal running performance of the fan unit is characterized by comprising the following steps of:
based on the sensor module, collecting real-time operation data of the target unit in a preset period, wherein the real-time operation data comprise the rotating speed, power, flow and wind pressure data of the target unit;
Preprocessing the real-time operation data to obtain trusted operation data;
extracting unit operation characteristics based on the trusted operation data to obtain a unit operation characteristic data set;
Collecting historical operation data of a target unit, and constructing a unit anomaly detection model;
Performing unit abnormality detection according to the unit abnormality detection model and the unit operation feature set to obtain unit abnormality information;
and aiming at the unit abnormal information, performing unit operation risk assessment, and performing unit potential fault early warning according to an operation risk assessment result.
2. The method of claim 1, wherein preprocessing the real-time operational data to obtain trusted operational data comprises:
respectively cleaning the data of the rotating speed, the power, the flow and the wind pressure of the target unit, and removing abnormal values, missing values and repeated values to obtain complete operation data;
And carrying out smoothing treatment and normalization treatment on the complete operation data, and carrying out serialization treatment according to a time sequence to obtain a unit rotating speed sequence, a unit power sequence, a unit flow sequence and a unit wind pressure sequence to form the trusted operation data.
3. The method of claim 1, wherein performing unit operation feature extraction based on the trusted operation data to obtain a unit operation feature dataset, further comprising:
the unit operation characteristics comprise time domain characteristics, frequency domain characteristics and time-frequency domain characteristics;
Extracting a unit rotating speed sequence, a unit power sequence, a unit flow sequence and a unit wind pressure sequence based on the trusted operation data;
and respectively extracting time domain features, frequency domain features and time-frequency domain features according to the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence to obtain a unit operation feature data set.
4. The method of claim 3, wherein time domain feature, frequency domain feature and time-frequency domain feature extraction are performed according to the unit rotational speed sequence, unit power sequence, unit flow sequence and unit wind pressure sequence, respectively, further comprising:
Aiming at the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence, a time domain characteristic data set is obtained through statistical analysis;
Aiming at the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence, obtaining a frequency domain characteristic data set through Fourier transformation and spectrum analysis;
and acquiring a time-frequency domain characteristic data set by wavelet transformation aiming at the unit rotating speed sequence, the unit power sequence, the unit flow sequence and the unit wind pressure sequence.
5. The method of claim 1, wherein historical operational data of the target unit is collected, and a unit anomaly detection model is constructed, further comprising:
According to historical operation data of a target unit, extracting operation characteristics of a sample unit and corresponding sample unit states, wherein the sample unit states comprise a normal state and an abnormal state;
and performing supervised training by combining a machine learning principle based on the operation characteristics of the sample unit and the state of the sample unit to obtain the unit abnormality detection model.
6. The method of claim 1, wherein the unit anomaly detection is performed in combination with the unit operation feature set according to the unit anomaly detection model to obtain unit anomaly information, further comprising:
Based on the group of operation feature sets, respectively acquiring time domain feature data, frequency domain feature data and time-frequency domain feature data of rotating speed, power, flow and wind pressure;
And according to the unit abnormality detection model, unit abnormality detection is respectively carried out on the time domain feature data, the frequency domain feature data and the time-frequency domain feature data of the rotating speed, the power, the flow and the wind pressure, so as to obtain unit rotating speed abnormality information, unit power abnormality information, unit flow abnormality information and unit wind pressure abnormality information.
7. The method of claim 6, wherein, for the unit anomaly information, performing a unit operation risk assessment, and performing unit potential fault early warning according to an operation risk assessment result, further comprising:
Presetting an operation risk assessment constraint interval;
Based on the operation risk assessment constraint interval, combining the unit rotating speed abnormality information, the unit power abnormality information, the unit flow abnormality information and the unit wind pressure abnormality information, judging an operation risk type and evaluating a risk level to obtain an operation risk assessment result;
And carrying out potential fault early warning on the unit according to the running risk assessment result.
8. An abnormal operation performance detection system for a fan unit, the system comprising:
the real-time operation data acquisition module is based on the sensor module and acquires real-time operation data of the target unit in a preset period, wherein the real-time operation data comprise the rotating speed, the power, the flow and the wind pressure data of the target unit;
The preprocessing module is used for preprocessing the real-time operation data to obtain trusted operation data;
The characteristic data set acquisition module is used for extracting unit operation characteristics based on the trusted operation data to obtain a unit operation characteristic data set;
The abnormality detection model construction module is used for collecting historical operation data of the target unit and constructing a unit abnormality detection model;
The unit abnormality information acquisition module is used for carrying out unit abnormality detection according to the unit abnormality detection model and combining the unit operation characteristic set to obtain unit abnormality information;
and the fault early warning module is used for carrying out unit operation risk assessment aiming at the unit abnormal information and carrying out unit potential fault early warning according to an operation risk assessment result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410153117.7A CN118030409A (en) | 2024-02-02 | 2024-02-02 | Method and system for detecting abnormal operation performance of fan unit |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410153117.7A CN118030409A (en) | 2024-02-02 | 2024-02-02 | Method and system for detecting abnormal operation performance of fan unit |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118030409A true CN118030409A (en) | 2024-05-14 |
Family
ID=91003536
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410153117.7A Pending CN118030409A (en) | 2024-02-02 | 2024-02-02 | Method and system for detecting abnormal operation performance of fan unit |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118030409A (en) |
-
2024
- 2024-02-02 CN CN202410153117.7A patent/CN118030409A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109186813B (en) | Temperature sensor self-checking device and method | |
| US10725439B2 (en) | Apparatus and method for monitoring a device having a movable part | |
| US6088658A (en) | Statistical pattern analysis methods of partial discharge measurements in high voltage insulation | |
| CN111922095A (en) | Vibration diagnosis method for abnormal torsional vibration fault of roller of cold rolling mill | |
| Elforjani | Diagnosis and prognosis of real world wind turbine gears | |
| Xu et al. | Bayesian wavelet PCA methodology for turbomachinery damage diagnosis under uncertainty | |
| CN111538311B (en) | Flexible multi-state self-adaptive early warning method and device for mechanical equipment based on data mining | |
| CN113562562A (en) | Elevator safety early warning monitoring system and judgment method thereof | |
| KR102253230B1 (en) | Predictive diagnosis method and system of nuclear power plant equipment | |
| CN117082105B (en) | Environment-friendly intelligent hospital facility monitoring system and method | |
| CN110469461B (en) | Fracture estimation method and device for fan toothed belt and readable storage medium | |
| KR102545672B1 (en) | Method and apparatus for machine fault diagnosis | |
| CN117435908A (en) | Multi-fault feature extraction method for rotary machine | |
| McCormick et al. | Application of periodic time-varying autoregressive models to the detection of bearing faults | |
| KR20230127337A (en) | Method and device for identifying abnormalities in mechanical devices or mechanical parts | |
| CN117331790A (en) | Machine room fault detection method and device for data center | |
| CN117391443A (en) | Dust removal equipment state monitoring and early warning method and system | |
| CN118030409A (en) | Method and system for detecting abnormal operation performance of fan unit | |
| CN114689321A (en) | Bearing fault diagnosis method and device for wind generating set and electronic equipment | |
| CN115326393A (en) | Wind turbine generator bearing pair fault diagnosis method based on temperature information | |
| KR102196611B1 (en) | Predictive diagnosis method and system of nuclear power plant equipment | |
| CN112733446A (en) | Data-driven self-adaptive anomaly detection method | |
| CN117951633B (en) | Photovoltaic power generation equipment fault diagnosis method and system | |
| CN117932322B (en) | Flour equipment fault diagnosis method and system | |
| CN118015812A (en) | Equipment abnormality alarm method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |