Power station auxiliary machine fault diagnosis method and system for smart factory

Abstract

The invention relates to the technical field of electrical performance testing, in particular to a power station auxiliary machine fault diagnosis method and system oriented to a smart factory, and the method comprises the steps: obtaining a current sequence, a power sequence and a vibration sequence of a boiler feed pump; equally dividing the collection duration into a plurality of time periods, and determining the abnormal vibration value of the boiler feed pump in each time period; determining trend noise significance of the power sequence and the current sequence in each time period; determining a current power influence value of each time period based on the trend noise significance, the correlation between the current sequence and the power sequence in each time period and the distribution condition of power data; and based on the dispersion degree of the difference between the vibration abnormal value and the current power influence value and the distribution range of the vibration abnormal value, determining the load vibration difference of the boiler feed pump, and performing fault diagnosis on the boiler feed pump. The invention aims to improve the accuracy of fault diagnosis of the boiler feed pump.

Description

Technical Field

[0001] The present application relates to the field of electrical performance testing technology, and specifically to a power plant auxiliary equipment fault diagnosis method and system for smart factories. Background Art

[0002] Power plant auxiliary equipment is an important part of power plant production equipment and is crucial to the normal operation of power plant equipment. There are many types of power plant auxiliary equipment. In a narrow sense, power plant auxiliary equipment specifically refers to coal mills, blowers, induced draft fans, boiler feed pumps and high-pressure heaters. These equipment are called the five major auxiliary equipment due to their high reliability requirements, difficulty in manufacturing and direct impact on unit efficiency.

[0003] Among them, the boiler feed pump plays a vital role in the intelligent power distribution system, mainly including water delivery, pressurization, stabilizing the boiler water level, etc., to ensure the normal operation and thermal efficiency of the boiler. Therefore, it is necessary to monitor the operating status of the boiler feed pump in real time. Existing methods mainly perform fault diagnosis based on the abnormal vibration characteristics of the boiler feed pump, which usually ignores the influence of the dynamic changes in the operating load of the power station unit. The vibration characteristic form of the boiler feed pump is different during normal operation, resulting in low accuracy in fault diagnosis of the boiler feed pump. Summary of the invention

[0004] In view of the above, it is necessary to provide a power plant auxiliary machine fault diagnosis method and system for smart factories. Compared with the traditional power plant auxiliary machine fault diagnosis method, the accuracy of fault diagnosis of boiler feed water pump is improved:

[0005] In a first aspect, an embodiment of the present application provides a power plant auxiliary machine fault diagnosis method for a smart factory, the method comprising the following steps:

[0006] Collect the current data, power data and vibration data of the boiler feed pump at each collection time, and form a current sequence, a power sequence and a vibration sequence respectively according to the time sequence;

[0007] The acquisition time is divided into multiple time periods, and the modal components of the vibration sequence in each time period are obtained by using a signal decomposition algorithm. Based on the difference in the maximum amplitude corresponding frequency in the spectrum diagram between the vibration sequence in each time period and its modal components, the basic signal of the vibration sequence in each time period is determined;

[0008] Based on the correlation between the vibration signal and its basic signal in each time period, and the distribution of the neighboring amplitudes of all peak points in the spectrum diagram of all modal components except the basic signal, the vibration abnormal value of the boiler feed water pump in each time period is determined;

[0009] Based on the changing trend of the trend components of the power series and the current series in each time period and the autocorrelation of the residual components, the trend noise significance of the power series and the current series in each time period is determined;

[0010] Determine the current power impact value of each time period based on the trend noise significance, the correlation between the current sequence and the power sequence in each time period, and the distribution of power data;

[0011] Based on the discrete degree of the difference between the abnormal vibration value and the current power influence value, and the distribution range of the abnormal vibration value, the load vibration difference of the boiler feed water pump is determined, and fault diagnosis of the boiler feed water pump is performed.

[0012] In one embodiment, the basic signal is determined as follows:

[0013] The frequencies corresponding to the maximum amplitude values ​​in the spectrum diagrams of the vibration sequence and each modal component in each time period are respectively taken as the main frequencies of the vibration sequence and each modal component in each time period;

[0014] Among all the modal components of the vibration sequence within any time period, the modal component with the smallest main frequency difference with the vibration sequence within the any time period is used as the basic signal of the vibration sequence within the any time period.

[0015] In one embodiment, the process of determining the vibration abnormal value is as follows:

[0016] Calculate the absolute value of the correlation between the vibration sequence and its basic signal in each time period;

[0017] For the remaining modal components except the basic signal in each time period, calculate the mean value of all amplitudes in a local area of ​​a preset size centered on each peak point in the spectrum diagram of each modal component;

[0018] Calculate the cumulative sum of all the mean values ​​corresponding to all modal components except the basic signal in each time period;

[0019] Based on the absolute value of the correlation between the vibration sequence and its basic signal in each time period, and the cumulative sum corresponding to each time period, the vibration abnormal value of the boiler feed water pump in each time period is determined.

[0020] In one embodiment, the expression of the vibration abnormal value is:

[0021] Where D i represents the abnormal vibration value of the boiler feed water pump in the i-th time period; C i A represents the cumulative sum corresponding to the i-th time period; iIt represents the absolute value of the correlation between the vibration sequence and its basic signal in the i-th time period, and α represents a preset value greater than 0.

[0022] In one embodiment, the process of determining the trend noise significance is:

[0023] Obtain the first-order difference sequence of the trend component of the power sequence in each time period, and take the accumulated value of the absolute values ​​of all data in the first-order difference sequence as the trend change of the power sequence in each time period;

[0024] The autocorrelation function is used to calculate the autocorrelation coefficient of the residual component of the power sequence in each time period at each preset time delay; the absolute value of the mean of the autocorrelation function of the power sequence in each time period corresponding to all preset time delays is calculated, which is recorded as the autocorrelation mean;

[0025] Based on the trend change amount and the autocorrelation mean, the trend noise significance of the power sequence in each time period is obtained;

[0026] For the current sequence in each time period, the trend noise significance of the current sequence in each time period is calculated according to the same calculation steps as the trend noise significance of the power sequence in each time period.

[0027] In one embodiment, the trend noise significance of the power sequence in each time period is the ratio of the trend change to the autocorrelation mean.

[0028] In one embodiment, the expression of the current power impact value is:

[0029] Where M i represents the current power impact value of the i-th time period; G i Indicates the trend noise significance of the power series in the i-th time period; H i represents the trend noise significance of the current series in the i-th time period; μ i represents the mean of the normalized values ​​of all data in the power sequence in the i-th time period; L i represents the Spearman correlation coefficient between the current sequence and the power sequence in the i-th time period; β represents a value preset to be greater than 0.

[0030] In one embodiment, the process of determining the load vibration difference of the boiler feed water pump is as follows:

[0031] Calculate the difference between the abnormal vibration value and the current power impact value corresponding to each time period and record it as the comprehensive difference;

[0032] The discrete degree of the comprehensive difference in all time periods and the extreme difference of the vibration abnormal values ​​corresponding to all time periods are fused as the load vibration difference of the boiler feed water pump.

[0033] In one embodiment, the method for diagnosing a fault of a boiler feed water pump is:

[0034] When the normalized value of the load vibration difference of the boiler feed water pump is greater than the preset fault abnormality threshold, it is determined that there is a fault in the boiler feed water pump; otherwise, there is no fault.

[0035] In the second aspect, an embodiment of the present application also provides a power plant auxiliary equipment fault diagnosis system for a smart factory, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein when the processor executes the computer program, the steps of any one of the above-mentioned power plant auxiliary equipment fault diagnosis methods for a smart factory are implemented.

[0036] This application has at least the following beneficial effects:

[0037] The present application determines the abnormal vibration value of the boiler feed pump in each time period by analyzing the content of abnormal vibration frequency in the vibration sequence and the degree of deviation of the vibration sequence relative to the basic signal, thereby improving the accuracy of analyzing the possibility of the existence of abnormal vibration characteristics in each time period;

[0038] Furthermore, by analyzing the trend changes and noise characteristics of the boiler feed water pump current and power data, combined with the correlation between the current series and the power series, the current power impact value of each time period is determined, and then based on the difference characteristics between the vibration abnormal value and the current power impact value, combined with the distribution range of the vibration abnormal value, the load vibration difference is determined. Combined with the correlation between the vibration change and the load change, the accuracy of extracting the abnormal operation characteristics of the boiler feed water pump is improved, thereby improving the accuracy of fault diagnosis of the boiler feed water pump. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0040] Figure 1 A flowchart of a method for diagnosing power plant auxiliary equipment faults in a smart factory provided in accordance with an embodiment of the present application;

[0041] Figure 2 The comparison diagram of vibration signal spectrum between normal state and abnormal state;

[0042] Figure 3 It is a schematic diagram of the process of obtaining vibration abnormal values;

[0043] Figure 4 Schematic diagram of the process of obtaining the trend noise significance of the power series;

[0044] Figure 5 This is a schematic diagram of current power in normal state;

[0045] Figure 6 Schematic diagram of current power in abnormal state. DETAILED DESCRIPTION

[0046] In the description of the embodiments of the present application, words such as "exemplary", "or", "for example" and the like are used to indicate examples, illustrations or descriptions. Any embodiment or design described as "exemplary" or "for example" in the embodiments of the present application should not be interpreted as being more preferred or more advantageous than other embodiments or designs. Specifically, the use of words such as "exemplary", "or", "for example" and the like is intended to present related concepts in a concrete manner.

[0047] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art in the present application. The terms used in the present application are only for the purpose of describing specific embodiments and are not intended to limit the present application. It should be understood that, unless otherwise specified, " / " means or.

[0048] It should also be noted that the terms "first" and "second" in the present application are used to distinguish similar objects rather than to describe a specific order or sequence.

[0049] The specific scheme of the power plant auxiliary equipment fault diagnosis method and system for smart factories provided by the present application is described in detail below with reference to the accompanying drawings.

[0050] See also Figure 1 , which shows a flowchart of a method for diagnosing power plant auxiliary equipment faults for a smart factory provided by an embodiment of the present application, the method comprising the following steps:

[0051] Step 1: Collect the current data, power data and vibration data of the boiler feed pump at each collection time, and form a current sequence, a power sequence and a vibration sequence according to the time sequence.

[0052] The auxiliary equipment of the power plant includes many equipment in the production process of the power plant. This embodiment performs fault diagnosis on the boiler feed pump. When an abnormality occurs inside the boiler feed pump, it is usually accompanied by increased vibration and changes in working performance. Therefore, the current data and power data of the boiler feed pump motor are collected through a digital power meter. The displacement sensor is used to collect the vibration data of the pump body position. The collected current data, power data and vibration data are arranged in time sequence to form a current sequence, a power sequence and a vibration sequence respectively.

[0053] In this embodiment, when using a digital power meter and a displacement sensor to collect data, the collection time interval is 0.01 seconds and the collection duration is 1 minute. The values ​​of the time interval and the collection duration are preset manually and can be set by the implementer. This application does not impose any special restrictions.

[0054] Step 2: Determine the abnormal vibration value of the boiler feed pump based on the content of abnormal vibration frequency in the vibration sequence of the boiler feed pump and the degree of deviation of the vibration sequence relative to the basic signal; determine the current power impact value of each time period by analyzing the trend changes and noise characteristics of the boiler feed pump current and power data, combined with the correlation between the current sequence and the power sequence; and further determine the load vibration difference.

[0055] In the boiler system, the water in the boiler is heated and discharged as steam to provide power for the steam turbine. In order to meet the requirements of the water level and steam pressure in the boiler, the boiler feed pump needs to continuously supply water to the boiler. The boiler feed pump operates in a high temperature and high pressure environment for a long time, and it has a very high speed during operation. Therefore, the boiler feed pump is prone to failure. Common types of failures include internal leakage of the feed pump, filter blockage and abnormal vibration of the feed pump. These failures will affect various parameters of the boiler feed pump. Among them, the vibration parameter is relatively easy to measure and is an important feature to characterize the operating status of the boiler feed pump. Many mechanical failures of the boiler feed pump can be reflected in the form of abnormal vibration, such as rotor imbalance, impeller corrosion and bearing damage.

[0056] The boiler feed pump will generate a uniform and fixed-frequency vibration signal during normal operation, and the vibration frequency under normal operating conditions is used as the base frequency. When the rotor is unbalanced or the bearing is damaged, the vibration sequence will show significant amplitude changes, and the number of abnormal signals detected in the low-frequency range will increase. When an impeller corrosion failure occurs, it may cause uneven distribution of impeller mass. This mass shift will cause additional vibration when the pump rotates at high speed, resulting in increased vibration intensity and easy introduction of high-frequency vibration. Based on the above characteristics, a preliminary assessment of the operating vibration state of the boiler feed pump in the power plant auxiliary equipment can be made. The comparison of the vibration signal spectrum between normal and abnormal conditions is shown in the figure below. Figure 2As shown in the figure, the horizontal axis is frequency and the vertical axis is amplitude. When the boiler feed water pump is operating normally, the data in the spectrum diagram of its vibration sequence is mainly concentrated in a certain frequency domain. Taking 50HZ as an example, in the normal signal spectrum diagram, the amplitude of the 50HZ adjacent frequency is large, and the amplitude of other frequencies is small; the abnormal state signal spectrum diagram (a) is a vibration signal spectrum diagram with an abnormal increase in low frequencies when the rotor is unbalanced or the bearing is damaged. Compared with the normal signal spectrum diagram, the amplitudes of the 100HZ, 150HZ and 200HZ adjacent frequencies are larger; the abnormal state signal spectrum diagram (b) is a vibration signal spectrum diagram of the introduction of high-frequency vibration signals when the impeller fails. Compared with the normal signal spectrum diagram, the amplitudes of the 400HZ and 600HZ adjacent frequencies in the figure are larger.

[0057] Step 2.1, divide the acquisition time into multiple time periods, use the signal decomposition algorithm to obtain the modal components of the vibration sequence in each time period, and determine the basic signal of the vibration sequence in each time period based on the difference in the corresponding frequency of the maximum amplitude in the spectrum diagram between the vibration sequence in each time period and its modal components.

[0058] Since the abnormal operation of boiler feed water pumps is highly random, the abnormal characteristics in different time periods may vary significantly. In order to accurately analyze the operation status in the short term, the collection time is divided into a preset number of time periods.

[0059] In this embodiment, the value of the preset number is 10. The value of the preset number is preset manually and can be set by the implementer. This application does not impose any special restrictions.

[0060] The present application uses a signal decomposition algorithm to extract vibration fault features. The signal decomposition algorithm is used to perform modal decomposition on the vibration sequence in each time period to obtain each modal component of the vibration sequence in each time period, wherein each modal component is also a discrete sequence, representing each frequency component in the vibration sequence in each time period.

[0061] In this embodiment, the ensemble empirical mode decomposition (EEMD) method is used for modal decomposition. As other implementation methods, on the basis of being able to perform modal decomposition on the vibration sequence, the implementer may use other existing technologies for modal decomposition, such as the empirical mode decomposition method, the complementary ensemble empirical mode decomposition method, etc., and this application does not impose any special restrictions.

[0062] When the boiler feed pump is operating normally, the data in the spectrum diagram of its vibration sequence is mainly concentrated in a certain frequency domain. Taking 50HZ as an example, 50HZ is used as the basic frequency. When the boiler feed pump fails, the vibration sequence contains data of other abnormal frequencies in addition to the 50HZ basic frequency data with the most content, such as low-frequency 100HZ and 150HZ data, and high-frequency 350HZ data. The higher the amplitude of these abnormal frequencies, the greater the possibility of failure of the boiler feed pump.

[0063] The vibration sequence in each time period is subjected to discrete Fourier transform to obtain the frequency spectrum of the vibration sequence in each time period, and the frequency corresponding to the maximum amplitude in the frequency spectrum of the vibration sequence in each time period is taken as the main frequency of the vibration sequence in each time period; for each modal component, the main frequency of each modal component is obtained according to the same acquisition method as the main frequency of the vibration sequence in each time period.

[0064] Among all the modal components of the vibration sequence within any time period, the modal component with the smallest main frequency difference with the vibration sequence within the any time period is used as the basic signal of the vibration sequence within the any time period.

[0065] Step 2.2, based on the correlation between the vibration signal and its basic signal in each time period, and the distribution of the neighboring amplitudes of all peak points in the spectrum diagram of all modal components except the basic signal, determine the vibration abnormal value of the boiler feed water pump in each time period.

[0066] Taking the i-th time period as an example, the absolute value of the correlation between the vibration sequence and its basic signal in the i-th time period is calculated. The smaller the absolute value of the correlation is, the greater the difference between the vibration sequence and the decomposed basic signal in the i-th time period is, that is, the greater the deviation of the actual vibration signal relative to the uniform and fixed-frequency vibration signal under normal operating conditions.

[0067] In this embodiment, the correlation between the vibration sequence and its underlying signal is the Pearson correlation coefficient. As other implementation methods, on the basis of being able to measure the correlation between the vibration sequence and its underlying signal, the implementer may adopt other existing technologies to measure the correlation between the vibration sequence and its underlying signal, such as the Spearman correlation coefficient, the Kendall rank correlation coefficient, etc., and this application does not impose any special restrictions.

[0068] The absolute value of the correlation analyzes the possibility of abnormality in the vibration sequence from a time domain perspective, and it is necessary to further analyze the possibility of faults in combination with the amplitude of the abnormal frequency signal. For the spectrum diagrams of other modal components except the basic signal, the present application uses the findpeaks function in matalb to extract peak values, calculates the mean of all amplitudes in the local area with a preset radius centered on each peak point, and calculates the cumulative sum of the mean values ​​of all peak points in the spectrum diagrams of all modal components except the basic signal. The larger the value of the cumulative sum, the more abnormal frequencies are contained in the vibration sequence in the i-th time period. Among them, when the local area constructed with a peak point in a certain spectrum diagram as the center exceeds the range of the spectrum diagram, the excess part is filled with data according to the spectrum diagram.

[0069] In this embodiment, the value of the preset radius is 10. The value of the preset radius is preset manually and can be set by the implementer. This application does not impose any special restrictions.

[0070] In this embodiment, the mean filling method is used for data filling. As other implementation methods, on the basis of achievable data filling, the implementer may use other existing technologies for data filling, such as the median filling method, interpolation method, etc., and this application does not impose any special restrictions.

[0071] Based on the above analysis, based on the absolute value of the correlation between the vibration sequence and its basic signal in each time period, and the cumulative sum corresponding to each time period, the vibration abnormal value of the boiler feed water pump in each time period is determined, and the expression is:

[0072] Where D i represents the abnormal vibration value of the boiler feed water pump in the i-th time period; C i A represents the cumulative sum corresponding to the i-th time period; i It represents the absolute value of the correlation between the vibration sequence and its basic signal in the i-th time period. α represents a preset value greater than 0. The purpose is to avoid the denominator being 0. The value of α is preset manually and can be set by the implementer. In this embodiment, the value of α is 0.01.

[0073] It should be noted that: the vibration abnormal value D i The larger the value of is, the greater the possibility that the boiler feed water pump has a vibration fault in the i-th time period. The schematic diagram of the process of obtaining vibration abnormal values ​​is as follows: Figure 3 shown.

[0074] Step 2.3 determines the trend noise significance of the power sequence and the current sequence in each time period based on the changing trend of the trend components of the power sequence and the current sequence in each time period and the autocorrelation of the residual components.

[0075] In the actual operation of the power plant unit, the operating load is dynamic, which causes the current and power required by the boiler feed pump to change accordingly. The changes in current and power will affect the rotation rate of the internal components of the boiler feed pump, which will in turn cause changes in the vibration characteristics. In addition, under a certain load state, the current and power of the boiler feed pump are not smooth and stable, and noise or fluctuations may occur due to factors such as poor heat dissipation. Further analyze the overall change trend of current and power and the noise content. Taking power data as an example, the STL (Seasonal and Trend decomposition using Loess) algorithm is used to decompose the power series in each time period to obtain trend components, seasonal components, and residual components.

[0076] Among them, the trend component represents the trend state of the power data changing with the load in each time period, and the first-order difference sequence of the trend component of the power sequence in each time period is obtained, and the cumulative value of the absolute value of all data in the first-order difference sequence is taken as the trend change of the power sequence in each time period; if the value of the trend change is larger, it means that the trend characteristics of the power change in each time period are more significant. If the collected power sequence has more noise, the more random fluctuations the residual component has, and the weaker the autocorrelation. Therefore, the autocorrelation function is used to calculate the autocorrelation coefficient of the residual component of the power sequence in each time period at each preset time delay, and the absolute value of the mean of the autocorrelation function of the power sequence in each time period corresponding to all preset time delays is calculated, which is recorded as the autocorrelation mean. If the autocorrelation mean is smaller, it means that the autocorrelation of the residual component is weaker and the power sequence contains more noise.

[0077] In this embodiment, the value range of the preset time delay is an integer in [1, N-1], where N represents the length of the residual component. The value range of the preset time delay is preset manually and can be set by the implementer. This application does not impose any special restrictions.

[0078] The ratio of the trend change of the power sequence in each time period to the corresponding autocorrelation mean is taken as the trend noise significance of the power sequence in each time period. The larger the value of the trend noise significance is, the more obvious the trend change of the power in each time period is and the more noise content is. The schematic diagram of the process of obtaining the trend noise significance of the power sequence is as follows: Figure 4 shown.

[0079] For the current data, the trend noise significance of the current series in each time period is calculated according to the same calculation method as the trend noise significance of the power series in each time period. If the value of the trend noise significance is larger, it indicates that the trend change of the current in each time period is more obvious and the noise content is more.

[0080] Step 2.4, based on the trend noise significance, the correlation between the current sequence and the power sequence in each time period, and the distribution of power data, determine the current power impact value of each time period.

[0081] Since the current and power of the boiler feed pump are positively correlated in a good state, the power will increase as the current increases during load changes. However, if the motor terminals are loose, the correlation between current and power will decrease, which will affect the speed of the boiler feed pump and easily cause failure. The normal current power diagram is as follows Figure 5 As shown, the current and power are positively correlated in normal state, and the current and power diagram in abnormal state is as follows Figure 6 As shown, under abnormal conditions, the correlation between current and power decreases.

[0082] The Spearman correlation coefficient between the current sequence and the power sequence in each time period is calculated. If the value of the Spearman correlation coefficient is smaller, it indicates that the positive correlation between current and power is lower, and the power supply state of the boiler feed water pump is more unstable.

[0083] Based on the trend noise significance of the power sequence in each time period, the trend noise significance of the current sequence, and the Spearman correlation coefficient between the current sequence and the power sequence, combined with the distribution of power data in each time period, the current power impact value of each time period is determined, and the expression is:

[0084] Where M i represents the current power impact value of the i-th time period; G i Indicates the trend noise significance of the power series in the i-th time period; H i represents the trend noise significance of the current series in the i-th time period; μ i represents the mean of the normalized values ​​of all data in the power sequence in the i-th time period; L i represents the Spearman correlation coefficient between the current sequence and the power sequence in the i-th time period; β represents a preset value greater than 0, the purpose of which is to avoid the denominator being 0. The value of β is preset manually and can be set by the implementer. In this embodiment, the value of β is 0.01.

[0085] It should be noted that: the current power influence value M i The larger it is, the greater the operating load of the boiler feed water pump in the i-th time period, and the greater the degree to which the boiler feed water pump is affected by the changes in current and power in the i-th time period.

[0086] Step 2.5, based on the discrete degree of the difference between the abnormal vibration value and the current power impact value, and the distribution range of the abnormal vibration value, determine the load vibration difference of the boiler feed water pump, and perform fault diagnosis on the boiler feed water pump.

[0087] Under different power plant unit loads, the vibration characteristics of the boiler feed pump are different. For example, when the boiler feed pump is running at high load, the operating conditions of the boiler feed pump are more stringent under high load, resulting in more mechanical stress or fluid dynamic problems, which increases vibration, so the calculated vibration abnormality value is relatively large; when the boiler feed pump is running at low load, the operation of the boiler feed pump is more stable, the vibration source is reduced, and the calculated vibration abnormality value is relatively small. Therefore, the diagnostic results of fault diagnosis based only on the size of the vibration abnormality value are not accurate enough. Combined with the above characteristics, the load vibration difference of the boiler feed pump is calculated.

[0088] The difference between the vibration abnormality value and the current power impact value corresponding to each time period is calculated respectively, recorded as the comprehensive difference, and the discrete degree of the comprehensive difference of all time periods is calculated. The greater the discrete degree, the greater the difference between the vibration change and the load change of the boiler feed water pump. Under good operating conditions, the vibration abnormality value of the boiler feed water pump usually does not change significantly despite load fluctuations. Therefore, the range of the vibration abnormality value of the boiler feed water pump in all time periods is calculated. The greater the range, the greater the degree of change in the vibration characteristics.

[0089] In this embodiment, the difference between the vibration abnormality value and the current power impact value is the absolute value of the difference. As other implementation methods, on the basis of the measurable difference between the vibration abnormality value and the current power impact value, the implementer may use other calculation methods to measure the difference between the vibration abnormality value and the current power impact value, such as ratio, square of the difference, etc. This application does not impose any special restrictions.

[0090] In this embodiment, the degree of dispersion is the coefficient of variation. As other implementation methods, on the basis of being able to measure the unevenness of the comprehensive difference distribution, the implementer may use other statistics to measure the unevenness of the comprehensive difference distribution, such as standard deviation, variance, etc., and this application does not impose any special restrictions.

[0091] The discrete degree of the comprehensive difference in all time periods and the extreme difference of the vibration abnormal values ​​corresponding to all time periods are fused as the load vibration difference of the boiler feed water pump.

[0092] It should be understood that fusion refers to combining multiple independent variables in a way that enhances the overall effect, such as multiplication relationship, addition relationship, etc., and implementers can limit it according to actual conditions.

[0093] In this embodiment, the product of the discrete degree of the comprehensive difference in all time periods and the range of the corresponding vibration abnormal values ​​in all time periods is taken as the load vibration difference of the boiler feed water pump.

[0094] In another embodiment, the sum of the discreteness of the comprehensive differences in all time periods and the range of the corresponding vibration abnormal values ​​in all time periods is used as the load vibration variability of the boiler feed water pump.

[0095] It should be noted that the greater the load vibration difference R of the boiler feed pump, the greater the possibility that the boiler feed pump is faulty.

[0096] The load vibration variability of the boiler feed pump is normalized, and the expression is: S=1-exp(-R); where S represents the normalized value of the load vibration variability of the boiler feed pump, exp() is an exponential function with a natural constant as the base, and R represents the load vibration variability of the boiler feed pump.

[0097] Furthermore, a fault abnormality threshold is set. When the normalized value of the load vibration difference of the boiler feed water pump is greater than the fault abnormality threshold, it is determined that there is a fault in the boiler feed water pump; otherwise, it is determined that the boiler feed water pump is in good operating condition and there is no fault.

[0098] In this embodiment, the value of the fault abnormality threshold is 0.7. The value of the fault abnormality threshold is preset manually and can be set by the implementer. This application does not impose any special restrictions.

[0099] Based on the same inventive concept as the above method, an embodiment of the present application also provides a power plant auxiliary equipment fault diagnosis system for smart factories, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, the steps of any one of the above-mentioned power plant auxiliary equipment fault diagnosis methods for smart factories are implemented.

[0100] In summary, the present application determines the vibration abnormality value of the boiler feed pump in each time period by analyzing the content of abnormal vibration frequency in the vibration sequence and the degree of deviation of the vibration sequence relative to the basic signal, thereby improving the accuracy of analyzing the possibility of the existence of abnormal vibration characteristics in each time period;

[0101] Furthermore, by analyzing the trend changes and noise characteristics of the boiler feed water pump current and power data, combined with the correlation between the current series and the power series, the current power impact value of each time period is determined, and then based on the difference characteristics between the vibration abnormal value and the current power impact value, combined with the distribution range of the vibration abnormal value, the load vibration difference is determined. Combined with the correlation between the vibration change and the load change, the accuracy of extracting the abnormal operation characteristics of the boiler feed water pump is improved, thereby improving the accuracy of fault diagnosis of the boiler feed water pump.

[0102] The flowchart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to the embodiment of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. In some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. In the description corresponding to the flowchart and the block diagram in the accompanying drawings, the operations or steps corresponding to different boxes can also occur in a different order from the order disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. Each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented by a dedicated hardware-based system that performs the specified functions or actions, or may be implemented by a combination of dedicated hardware and computer instructions.

[0103] It is obvious to those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and that the present application can be implemented in other specific forms without departing from the basic features of the present application. Therefore, no matter from which point of view, the above embodiments of the present application should be regarded as exemplary and non-restrictive.

Claims

1. A power plant auxiliary equipment fault diagnosis method for smart factories, characterized in that: The method comprises the following steps: Collect the current data, power data and vibration data of the boiler feed pump at each collection time, and form a current sequence, a power sequence and a vibration sequence respectively according to the time sequence; The acquisition time is divided into multiple time periods, and the modal components of the vibration sequence in each time period are obtained by using a signal decomposition algorithm. Based on the difference in the maximum amplitude corresponding frequency in the spectrum diagram between the vibration sequence in each time period and its modal components, the basic signal of the vibration sequence in each time period is determined; Based on the correlation between the vibration signal and its basic signal in each time period, and the distribution of the neighboring amplitudes of all peak points in the spectrum diagram of all modal components except the basic signal, the vibration abnormal value of the boiler feed water pump in each time period is determined; Based on the changing trend of the trend components of the power series and the current series in each time period and the autocorrelation of the residual components, the trend noise significance of the power series and the current series in each time period is determined; Determine the current power impact value of each time period based on the trend noise significance, the correlation between the current sequence and the power sequence in each time period, and the distribution of power data; Based on the discrete degree of the difference between the abnormal vibration value and the current power influence value, and the distribution range of the abnormal vibration value, the load vibration difference of the boiler feed water pump is determined, and fault diagnosis of the boiler feed water pump is performed.

2. The power plant auxiliary machine fault diagnosis method for a smart factory according to claim 1, characterized in that: The process of determining the basic signal is as follows: The frequencies corresponding to the maximum amplitude values ​​in the spectrum diagrams of the vibration sequence and each modal component in each time period are respectively taken as the main frequencies of the vibration sequence and each modal component in each time period; Among all the modal components of the vibration sequence within any time period, the modal component with the smallest main frequency difference with the vibration sequence within the any time period is used as the basic signal of the vibration sequence within the any time period.

3. The power plant auxiliary machine fault diagnosis method for smart factories according to claim 1, characterized in that: The process of determining the vibration abnormal value is as follows: Calculate the absolute value of the correlation between the vibration sequence and its basic signal in each time period; For the remaining modal components except the basic signal in each time period, calculate the mean value of all amplitudes in a local area of ​​a preset size centered on each peak point in the spectrum diagram of each modal component; Calculate the cumulative sum of all the mean values ​​corresponding to all modal components except the basic signal in each time period; Based on the absolute value of the correlation between the vibration sequence and its basic signal in each time period, and the cumulative sum corresponding to each time period, the vibration abnormal value of the boiler feed water pump in each time period is determined.

4. The power plant auxiliary machine fault diagnosis method for a smart factory according to claim 3, characterized in that: The expression of the vibration abnormal value is: Where D i represents the abnormal vibration value of the boiler feed water pump in the i-th time period; C i A represents the cumulative sum corresponding to the i-th time period; i It represents the absolute value of the correlation between the vibration sequence and its basic signal in the i-th time period, and α represents a preset value greater than 0.

5. The power plant auxiliary machine fault diagnosis method for smart factories according to claim 1, characterized in that: The process of determining the significance of the trend noise is as follows: Obtain the first-order difference sequence of the trend component of the power sequence in each time period, and take the accumulated value of the absolute values ​​of all data in the first-order difference sequence as the trend change of the power sequence in each time period; The autocorrelation function is used to calculate the autocorrelation coefficient of the residual component of the power sequence in each time period at each preset time delay; the absolute value of the mean of the autocorrelation function of the power sequence in each time period corresponding to all preset time delays is calculated, which is recorded as the autocorrelation mean; Based on the trend change amount and the autocorrelation mean, the trend noise significance of the power sequence in each time period is obtained; For the current sequence in each time period, the trend noise significance of the current sequence in each time period is calculated according to the same calculation steps as the trend noise significance of the power sequence in each time period.

6. The power plant auxiliary machine fault diagnosis method for smart factories according to claim 5, characterized in that: The trend noise significance of the power sequence in each time period is the ratio of the trend change amount to the autocorrelation mean.

7. The power plant auxiliary machine fault diagnosis method for a smart factory according to claim 1, characterized in that: The expression of the current power influence value is: Where M i represents the current power impact value of the i-th time period; G i Indicates the trend noise significance of the power series in the i-th time period; H i Indicates the trend noise significance of the current series in the i-th time period; μ i represents the mean of the normalized values ​​of all data in the power sequence in the i-th time period; L i represents the Spearman correlation coefficient between the current sequence and the power sequence in the i-th time period; β represents a value preset to be greater than 0.

8. The power plant auxiliary machine fault diagnosis method for smart factories according to claim 1, characterized in that: The process of determining the load vibration difference of the boiler feed water pump is as follows: Calculate the difference between the abnormal vibration value and the current power impact value corresponding to each time period and record it as the comprehensive difference; The discrete degree of the comprehensive difference in all time periods and the extreme difference of the vibration abnormal values ​​corresponding to all time periods are fused as the load vibration difference of the boiler feed water pump.

9. The power plant auxiliary machine fault diagnosis method for smart factories according to claim 1, characterized in that: The method for diagnosing a fault of a boiler feed pump is as follows: When the normalized value of the load vibration difference of the boiler feed water pump is greater than the preset fault abnormality threshold, it is determined that there is a fault in the boiler feed water pump; otherwise, there is no fault.

10. A power plant auxiliary machine fault diagnosis system for a smart factory, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the power plant auxiliary equipment fault diagnosis method for a smart factory as described in any one of claims 1 to 9 are implemented.

Images