A fault diagnosis method and system for an electrical control cabinet

Abstract

The present invention relates to the technical field of electrical fault diagnosis, and specifically to a fault diagnosis method and system for an electrical control cabinet. The technical solution generates a current curve based on the output time corresponding to the output current, and performs characteristic analysis on each current value of the current curve, so as to obtain the fluctuation degree of each current value, as well as the smoothness characteristics and wave characteristics presented in the period on the current curve, and respectively obtain the steady-state load coefficient and oscillation load coefficient of each current value in the corresponding period, and obtain the abnormal confidence of each current value according to the fluctuation degree, steady-state load coefficient and oscillation load coefficient of each current value; output the fault diagnosis result of the electrical control cabinet according to all the abnormal confidences and preset confidence thresholds. The technical solution performs data analysis based on the current value on the current curve, and obtains the abnormal confidence that characterizes the abnormal characteristics of the current value after comprehensive analysis of multiple dimensions, thereby accurately obtaining the fault diagnosis result of the electrical control cabinet.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical fault diagnosis, and in particular to a fault diagnosis method and system for an electrical control cabinet. Background Art

[0002] An electrical control cabinet is a closed cabinet used for centralized control and protection of electrical equipment, which usually contains switches, relays, circuit breakers and other electrical components. The main function of an electrical control cabinet is to provide power to various devices, such as motors, sensors, etc., or to start, stop or adjust the operating status of the controlled equipment. Due to the complex internal structure of the electrical control cabinet, if an unexpected connection fault occurs in its internal circuit, the output current will increase abnormally; if the relay fails, the control signal will fail, thus affecting the operation of the equipment. In order to prevent the electrical control cabinet from affecting the controlled equipment, it is necessary to perform fault diagnosis on it.

[0003] In order to diagnose faults of electrical control cabinets, fault diagnosis is performed by collecting the output current of the electrical control cabinet. However, since the electrical control cabinet is often connected to more than one load, the current states of different loads when running are sinusoidal waves, square waves, etc., and the currents between different loads are mixed together, making it difficult to perform fault diagnosis based on the output current of the electrical control cabinet. Summary of the invention

[0004] In order to solve the technical problem of how to accurately perform fault diagnosis based on the output current of the electrical control cabinet, the purpose of the present invention is to provide a fault diagnosis method and system for the electrical control cabinet. The technical solution adopted is as follows:

[0005] In a first aspect, an embodiment of the present invention provides a fault diagnosis method for an electrical control cabinet, the method comprising:

[0006] Obtaining the output current of the electrical control cabinet, and generating a current curve based on the output time corresponding to the output current, wherein the current curve is a curve of the output current of the electrical control cabinet changing with the output time;

[0007] Performing a fluctuation characteristic analysis on each current value of the current curve to obtain the fluctuation degree of each current value;

[0008] According to the smoothness and ripple characteristics of each current value in the period on the current curve, the steady-state load factor and the oscillation load factor of each current value in the corresponding period are obtained respectively;

[0009] According to the fluctuation degree, steady-state load factor and oscillation load factor of each current value, the abnormal confidence of each current value is obtained;

[0010] According to all abnormal confidence levels and preset confidence thresholds, the fault diagnosis results of the electrical control cabinet are output.

[0011] In an optional embodiment, performing a fluctuation characteristic analysis on each current value of the current curve to obtain the fluctuation degree of each current value includes:

[0012] According to the difference between the current value on the current curve and the current value in the positive direction of time, the forward difference of the current value is obtained;

[0013] Obtaining a backward difference of the current value according to the difference between the current value and the current value in the negative direction of time;

[0014] According to the forward difference and the backward difference, the fluctuation degree of the current value is obtained.

[0015] In an optional embodiment, according to the smoothness characteristics of each current value in the period on the current curve, the steady-state load factor of each current value in the corresponding period is obtained, including:

[0016] According to the autocorrelation coefficient between the current value and the adjacent current values ​​on the current curve, the current curve is divided into a plurality of periodically changing curve segments;

[0017] According to the extreme value variation characteristics of the current values ​​in the multiple curve segments, a sine wave superposition coefficient of the current value in the cycle is obtained, wherein the sine wave superposition coefficient is a coefficient representing the number of sine waves superimposed on the current value in the cycle;

[0018] The steady-state load factor of the current value is obtained according to the sine wave superposition coefficient and the average value of the fluctuation of the adjacent current values.

[0019] In an optional embodiment, according to the extreme value variation characteristics of the current values ​​in the multiple curve segments, the sine wave superposition coefficient of the current value in the cycle is obtained, including:

[0020] Performing data extraction according to the current value represented by each curve segment to obtain a periodic data set of each curve segment;

[0021] Perform extreme value extraction on the current value in each cycle data set to obtain the maximum current value and the minimum current value of each cycle data set;

[0022] According to the maximum current value and the minimum current value in the periodic data set of the current curve segment and its surrounding curve segments, the sine wave superposition coefficient of the current current value of the current curve segment is obtained.

[0023] In an optional embodiment, the sine wave superposition coefficient of the current current value of the current curve segment is obtained according to the maximum current value and the minimum current value in the periodic data set of the current curve segment and its surrounding curve segments, including:

[0024] According to the formula: , get the first Sine wave superposition coefficient of current value , where N is the maximum current value of the current curve segment and its surrounding curve segments, The first The maximum current value, The current curve segment and its surrounding curve segments are The current value closest to the maximum current value, is the minimum current value of the current curve segment, The maximum current value and It is an exponential function with the natural constant e as its base.

[0025] In an optional embodiment, according to the wave characteristics of each current value in the period on the current curve, the oscillation load factor of each current value in the corresponding period is obtained, including:

[0026] Determine whether the fluctuation of the current value adjacent to the current value on the current curve is greater than a preset value;

[0027] If yes, the corresponding current value is marked as a data turning point;

[0028] According to the distribution characteristics of the data turning point on the current curve, a period factor representing the timing correlation between the data turning point and the current current value is obtained;

[0029] The oscillation load factor of the current current value is obtained according to the period factor and the data interval, wherein the data interval is the data amount between the current current value and the data turning point.

[0030] In an optional embodiment, obtaining the oscillation load factor of the current current value according to the period factor and the data interval includes:

[0031] According to the formula , obtain the Oscillation load factor for each data point ,in, is the period factor, For the The stationary coefficient of the data points, For the The amount of data between the data point and the turning point of the backward data, For the The amount of data between the data point and the turning point of the forward data, is a hyperbolic function.

[0032] In an optional embodiment, the abnormal confidence of each current value is obtained according to the fluctuation degree, steady-state load factor and oscillation load factor of each current value, including:

[0033] According to the ratio of the steady-state load factor to the oscillating load factor of each current value, a load stability ratio of each current value is obtained;

[0034] Based on the load stability ratio and fluctuation degree of each current value, the abnormality confidence level is obtained.

[0035] In an optional embodiment, according to all abnormal confidence levels and a preset confidence threshold, a fault diagnosis result of the electrical control cabinet is output, including:

[0036] When the abnormal confidence is greater than the confidence threshold, the output current corresponding to the abnormal confidence is marked as an abnormal current;

[0037] When the density of the abnormal current is greater than a preset value, the fault diagnosis result is determined to be an operation fault.

[0038] In a second aspect, an embodiment of the present invention further provides a fault diagnosis system for an electrical control cabinet, the system comprising:

[0039] The acquisition terminal is used to obtain the output current of the electrical control cabinet and output it to the outside;

[0040] A data processing terminal is connected to the acquisition terminal, and the data processing terminal is used to generate a current curve based on the output time corresponding to the output current, wherein the current curve is a curve of the output current of the electrical control cabinet changing with the output time;

[0041] The data processing terminal is also used to perform fluctuation characteristic analysis on each current value of the current curve to obtain the fluctuation degree of each current value; obtain the steady-state load factor and oscillation load factor of each current value in the corresponding period according to the smoothness and wave characteristics of each current value in the period on the current curve; obtain the abnormal confidence of each current value according to the fluctuation degree, steady-state load factor and oscillation load factor of each current value; output the fault diagnosis result of the electrical control cabinet according to all abnormal confidence degrees and preset confidence thresholds.

[0042] The present invention has the following beneficial effects:

[0043] The technical solution of the present invention obtains the output current of the electrical control cabinet, generates a current curve based on the output time corresponding to the output current, and performs fluctuation characteristic analysis on each current value of the current curve, so as to obtain the fluctuation degree of each current value, and characterize the fluctuation magnitude between each current value and the adjacent current value by the fluctuation degree; according to the smoothness and tortuosity characteristics presented by each current value in the period on the current curve, respectively obtain the steady-state load coefficient and the oscillation load coefficient of each current value in the corresponding period, and respectively characterize the steady-state effect magnitude and the oscillation effect magnitude generated by each current value by the steady-state load coefficient and the oscillation load coefficient; according to the fluctuation degree, the steady-state load coefficient and the oscillation load coefficient of each current value, obtain the abnormal confidence of each current value; according to all abnormal confidences and the preset confidence threshold, output the fault diagnosis result of the electrical control cabinet. The technical solution performs data analysis based on the current value on the current curve, obtains the fluctuation magnitude, the steady-state effect magnitude and the oscillation effect magnitude caused by each current value, and obtains the abnormal confidence degree characterizing the abnormal characteristics of the current value after comprehensive analysis of multiple dimensions, and then accurately obtains the fault diagnosis result of the electrical control cabinet, so as to achieve the purpose of accurately implementing fault diagnosis based on the output current of the electrical control cabinet. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention 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 invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0045] Figure 1 A flow chart of a fault diagnosis method for an electrical control cabinet provided by one embodiment of the present invention;

[0046] Figure 2 A schematic diagram of a current curve provided by an embodiment of the present invention;

[0047] Figure 3 A schematic diagram of a current curve after a plurality of sine waves are superimposed provided by an embodiment of the present invention;

[0048] Figure 4 A schematic structural diagram of a fault diagnosis system for an electrical control cabinet provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the fault diagnosis method and system of an electrical control cabinet proposed by the present invention, its specific implementation method, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0050] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0051] The specific scheme of a fault diagnosis method and system for an electrical control cabinet provided by the present invention is described in detail below with reference to the accompanying drawings.

[0052] See also Figure 1 , Figure 1 A flowchart of a fault diagnosis method for an electrical control cabinet provided by an embodiment of the present invention. The method can be applied to a fault diagnosis terminal of an electrical control cabinet to perform fault diagnosis on it. The fault diagnosis terminal can be a control board with a processor, or other equipment capable of running the method, and is not specifically limited here. The fault diagnosis method includes:

[0053] S11. Obtain the output current of the electrical control cabinet, and generate a current curve based on the output time corresponding to the output current, wherein the current curve is a curve of the output current of the electrical control cabinet changing with the output time.

[0054] Specifically, electrical control cabinets are mostly used to power or control AC industrial equipment. Current sensors can be installed on the output terminals of the electrical control cabinet. During the operation of the electrical control cabinet, the output current of the electrical control cabinet is collected by the current sensor. There is a corresponding relationship between the output current and the output time, based on which a current curve can be generated. Figure 2 , Figure 2 It is a schematic diagram of the current curve, in which the horizontal axis is the output time t, in seconds, and the vertical axis is the output current I, in A.

[0055] After the output current is collected, data cleaning can also be performed. For example, the normal value range of the output current is determined based on the operating power of the electrical control cabinet. If the collected output current is in the normal value range, it means that the data collection is correct and the output current continues to be collected; otherwise, it means that there may be a collection error, and the output current is deleted to ensure that the output current is collected accurately. The current sensor can respond to the start signal of the electrical control cabinet to collect the output current of the electrical control cabinet in real time. After obtaining the current curve of the electrical control cabinet, enter step S12.

[0056] S12. Perform fluctuation characteristic analysis on each current value of the current curve to obtain the fluctuation degree of each current value.

[0057] Specifically, the fluctuation of each current value in the current curve is a quantitative representation of the stability of the current value. The fluctuation reflects the degree of change of the current at a specific moment relative to the surrounding current values. If the fluctuation of a current value is large, it means that the current at that moment is relatively unstable, which may be affected by external factors or changes in the internal state of the system. On the contrary, a small fluctuation indicates that the output current is relatively stable. By analyzing the fluctuation of the current value, the working status of the electrical control cabinet can be determined, and abnormal conditions such as equipment failure and load changes can be discovered in time.

[0058] Exemplarily, step S12 includes sub-steps S12-1 to S12-3, which are specifically described as follows:

[0059] S12-1. According to the difference between the current value on the current curve and the current value in the positive direction of time, obtain the forward difference of the current value. Please continue to refer to Figure 2 , any current value on the current curve is recorded as I0, the output time corresponding to the current current value I0 is t0, and the current current value I0 and the output time t0 along the positive direction of time other current values ​​I 0+1 The difference between the current value I0 and I 0+1 The difference can also be the difference between the current value and multiple current values ​​in the positive direction of time.

[0060] S12-2, according to the difference between the current value and the current value in the negative direction of time, obtain the backward difference of the current value. Similarly, the current current value I0 and other current values ​​I0 along the negative direction of time at output time t0 0-1 The difference between the current values ​​I0 and I 0-1 The difference.

[0061] S12-3. Obtain the fluctuation of the current value based on the forward difference and the backward difference. The fluctuation of the current value can be obtained based on the absolute value of the difference between the forward difference and the backward difference. The fluctuation characterizes the fluctuation amplitude of the current value compared with the adjacent current value. The greater the fluctuation, the greater the difference between the current value and the adjacent current value; conversely, the smaller the difference between the current value and the adjacent current value, the output current presents a stable characteristic.

[0062] It can be understood that since there are multiple loads at the output end of the electrical control cabinet, the power of the loads is not constant during normal operation. Therefore, under normal circumstances, the current output by the electrical control cabinet will fluctuate reasonably; the fluctuation degree of each current value can be calculated based on the following formula.

[0063]

[0064] In the formula, The current curve The fluctuation of the current value, The first Current value, adjacent to The current value in the positive direction at the time, adjacent to The current value in the negative direction at the time, is a hyperbolic function; i is a natural number, representing any data point on the current curve. It can be understood that when implementing fault detection, the current value is continuously obtained with the output time, and when the electrical control cabinet stops running, the fault diagnosis method stops running accordingly.

[0065] So far, the fluctuation characteristic analysis of each current value in the current curve has been completed, and the fluctuation degree of each current value has been obtained.

[0066] S13. According to the smoothness characteristics and the wave characteristics of each current value in the period on the current curve, the steady-state load factor and the oscillation load factor of each current value in the corresponding period are obtained respectively.

[0067] Specifically, since the output end of the electrical control cabinet is connected to various loads, and the currents required by different loads are different, the output current will continue to change, which will interfere with the fault diagnosis of the electrical control cabinet. Electrical control cabinet failures usually involve line short circuits, poor contact, and relay failure, which will cause large changes in the output current. Therefore, for the current curve obtained in the above manner, it is necessary to detect abnormal fluctuation data points in the current curve; however, different loads have different power consumption conditions, and the current of some loads is in a state of continuous fluctuation. Therefore, it is necessary to analyze the load conditions at the location according to the distribution of the current value, and perform fault detection according to the load status.

[0068] It can be understood that the current curve presents periodic fluctuations and can be divided into smooth characteristics and ripple characteristics according to the performance of each current value in the corresponding period. For the current value showing smooth characteristics, its steady-state load coefficient in the corresponding period can be obtained, which reflects the effect of the current value on the steady state. For the period with ripple characteristics, the oscillation load coefficient can be obtained based on the analysis of the current value, which reflects the effect of the current value on the ripple characteristics. By determining these two coefficients separately, we can have a more comprehensive understanding of the load characteristics of the current value in the corresponding period, providing a basis for further analysis of the fault diagnosis of the electrical control cabinet.

[0069] In practical applications, since the output end of the electrical control cabinet is connected to several loads, the load current curve of most industrial equipment is a sine wave; since the output end is connected to some control circuits, control signals and some specific equipment have square waves. Therefore, under normal circumstances, the output current of the electrical control cabinet is in a state of superposition between sine waves and square waves. In order to implement fault monitoring of the electrical control cabinet based on the output current, it is necessary to analyze the influence of sine waves and square waves on the current curve. The influence of sine waves on the current curve is reflected in smooth characteristics, and the influence of square waves on the current curve is reflected in tortuous characteristics.

[0070] The following embodiment of the present invention further specifically describes the superposition analysis of sine waves and the process of obtaining the steady-state load factor, which specifically includes:

[0071] S13-1. Divide the current curve into multiple periodic curve segments according to the autocorrelation coefficient between the current value and the adjacent current value on the current curve. Since the sine wave changes smoothly, the loads of the sine wave will also change relatively smoothly when they are superimposed on each other; and since the currents required by the loads are superimposed on each other, the superposition will affect each other, but will present a roughly periodic state. In order to obtain the periodic changes of the current curve, the autocorrelation coefficient between the current values ​​can be calculated using the autocorrelation function. For example, according to the formula:

[0072]

[0073] Get the autocorrelation coefficient ,in, is the number of current values ​​on the current curve, is the i-th current value on the current curve, i is a natural number, is the number of data points in one cycle, The current curve The current value is the value from point i to the The current value at each point.

[0074] According to the above method, we can obtain and The autocorrelation coefficient represents the periodicity between current values. The larger the autocorrelation coefficient, the better the periodicity. The value is used as the period of the current curve to divide the current curve into multiple periodic curve segments. Figure 3 , Figure 3 It is a schematic diagram of the current curve after multiple sine waves are superimposed, including two extreme points, namely point B and point D. Point A to point D in the figure is a single fluctuation cycle, that is, a single curve segment after division.

[0075] S13-2. According to the extreme value variation characteristics of the current values ​​in multiple curve segments, the sine wave superposition coefficient of the current value in the cycle is obtained, wherein the sine wave superposition coefficient is a coefficient that characterizes the number of sine waves superimposed on the current value in the cycle. The current value can be The maximum and minimum values ​​are determined from the current values ​​and used as the extreme values ​​of the curve segment. The superposition characteristics of the sine wave can be obtained through the changes in the extreme values, and then the sine wave superposition coefficient can be obtained.

[0076] Specifically, sub-step S13-2 includes:

[0077] The first step is to extract data based on the current value represented by each curve segment to obtain a periodic data set for each curve segment. A data set is established for each curve segment, and all current values ​​of the curve segment are extracted into a corresponding set to obtain a periodic data set for each curve segment.

[0078] The second step is to extract the extreme values ​​of the current values ​​in each periodic data set to obtain the maximum current value (i.e., the maximum point) and the minimum current value (i.e., the minimum point) of each periodic data set. An element (i.e., the current value) in the periodic data set is randomly selected as the initial value of the temporary maximum and minimum values. Then, each element in the periodic data set is traversed and compared with the temporary maximum value. If the element is greater than the temporary maximum value, the temporary maximum value is updated to the element. Similarly, it is compared with the temporary minimum value. If the element is less than the temporary minimum value, the temporary minimum value is updated to the element. After completing the traversal of the periodic data set, the temporary maximum and temporary minimum values ​​obtained are the true maximum current value and minimum current value. Please continue to refer to Figure 3 In the figure, point B is the minimum current value and point C is the maximum current value.

[0079] The third step is to obtain the sine wave superposition coefficient of the current current value of the current curve segment according to the maximum current value and the minimum current value in the periodic data set of the current curve segment and its surrounding curve segments. The surrounding curve segment can be a first curve segment adjacent to the current curve segment in the positive direction of time, and a second curve segment in the negative direction of time. Through the extreme value change characteristics between the first curve segment, the second curve segment and the current curve segment, the superposition state of the sine wave can be determined, and then the sine wave superposition coefficient can be obtained. The calculation method of the sine wave superposition coefficient can be implemented based on the data processing model.

[0080] For example, according to the formula: , get the first Sine wave superposition coefficient of current value , where N is the maximum current value of the current curve segment and its surrounding curve segments, The first The maximum current value, The current curve segment and its surrounding curve segments are The current value closest to the maximum current value, is the minimum current value of the current curve segment, The maximum current value and The distance between It is an exponential function with the natural constant e as its base.

[0081] It should be noted that in the above formula The smaller the calculated result is, the closer the maximum current values ​​of multiple curve segments are, and the numerical difference between the maximum current value and the minimum current value is not large. The current value fluctuation in the periodic data set is small, and there are more sine waves superimposed, so the sine wave superposition coefficient is larger; conversely, The larger the calculated result is, the smaller the sine wave superposition coefficient is.

[0082] S13-3. Obtain the steady-state load factor of the current value based on the sine wave superposition coefficient and the average value of the fluctuation of the current value. The superposition of sine waves will make the current curve smoother. The steady-state load factor can be calculated by combining the fluctuation of the current value. For example, according to the formula , calculate the steady-state load factor of the i-th current value on the current curve ,in, is the average value of the fluctuation of the current value nearby, which can be the average value of the fluctuation of 3-5 current values ​​near the current value. It is a hyperbolic function. It should be noted that during the operation of electrical control, the output current will not remain constant, so the fluctuation mean Not equal to 0. The larger the calculated value is, the larger the sine wave superposition coefficient of the current current value is, and the smaller the average fluctuation of the adjacent current values ​​is, which is consistent with the smooth waveform generated by the steady-state load superposition.

[0083] In practical applications, if the output end of the electrical control cabinet is connected to a control circuit or other equipment, its load current presents a square wave change state; due to the square wave waveform, the current curve will have large fluctuations when it changes. If the square wave is superimposed on the current of other loads, the current curve will show sharp mutation characteristics, causing the normal smooth sine wave to have a sharper curve structure, accompanied by a larger turning point. The superposition of square waves will cause interference during fault detection, so it is necessary to analyze whether there is a square wave load in the area where the current value is located. The following embodiments of the present invention further elaborate on the analysis of the twists and turns of the square wave and the process of obtaining the oscillation load coefficient, specifically including:

[0084] S13-4, determine whether the fluctuation of the neighboring current values ​​of the current value on the current curve is greater than the preset value. The neighboring current values ​​refer to the current values ​​closest to the current current value. The amount of data between the neighboring current values ​​and the current current value can be set based on actual needs, for example, set to a natural number less than 10; similarly, the preset value can be set based on actual needs, for example, 0.9.

[0085] S13-5. When the fluctuation of the adjacent current value is less than or equal to the preset value, it means that the adjacent current value is not affected by the wave characteristics of the square wave, and there may be no square wave superposition in this area; conversely, when the fluctuation of the adjacent current value is greater than the preset value, it means that the adjacent area of ​​the current current value is affected by the square wave superposition. In order to analyze the impact of the square wave superposition, the corresponding current value is marked as a data turning point.

[0086] S13-6. According to the distribution characteristics of the data turning point on the current curve, obtain the periodic factor that characterizes the timing correlation between the data turning point and the current current value. The large fluctuation of the data turning point means that it may be caused by the superposition of square waves. Normal square waves also have periodicity. The periodic factor can be determined based on the distribution characteristics of the data turning point on the current curve. For example, a sliding window of appropriate size is set. In the embodiment of the present invention, the sliding window is set to 5, and the sliding window is slid through the fluctuations of all adjacent current values. After each sliding window is recorded, the number of data turning points inside the sliding window is determined, and the distance between the data turning point inside the sliding window and the sliding window is obtained. , calculate the period factor according to the above formula.

[0087] The formula is: , where For the The period factor of the current value is is a variable representing the sliding window position, and n is a natural number. For the All neighboring current values ​​of data points sliding windows, the sliding window with the closest number of turning points in its internal data The distance between It is an exponential function with the natural constant e as its base.

[0088] It can be understood that for the analysis of square wave characteristics, the embodiment of the present invention uses a sliding window to obtain the number of turning points of the data inside each sliding window, and obtains the distance between two sliding windows with the closest numbers, and calculates the variance The smaller the value, the more consistent the sliding window distances with the closest number of data turning points are, indicating that the distribution of data turning points is periodic, so the larger the periodic factor is.

[0089] S13-7. Obtain the oscillation load coefficient of the current current value according to the periodic factor and the data interval, wherein the data interval is the amount of data between the current current value and the data turning point. Specifically, the obtained periodic factor can only indicate that the data turning point has periodicity. Since the periodicity still cannot fully determine the extent to which the area where the current value is located is affected by the square wave, it is also necessary to determine whether the data turning point affects the smoothness and consistency of the original data. The data interval can characterize the amount of data between the current current value and the data turning point, and thus the oscillation load coefficient of the current current value can be obtained based on the periodic factor and the data interval.

[0090] Furthermore, the autocorrelation function (ACF) can be used to analyze the smoothness of time series data, and determine whether there is strong seasonality or periodicity based on the surrounding data. For example, based on the location of the data turning point, two adjacent data adjacent to the data turning point are obtained, the data turning point and the two adjacent data are screened out from the current curve, and a new current data set is obtained based on the data point set separated by the screened data. The least squares method is used to obtain the fitted straight line of the obtained current data set, and the number of data points on both sides of the fitted straight line is determined for the obtained fitted straight line. The number of data points in the negative direction of time is recorded as , the number of data points in the positive direction of time is recorded as , continue to obtain its stationarity based on its autocorrelation diagram.

[0091] For example, according to the formula: , obtain the Oscillation load factor for each data point ,in, is the period factor, For the The stationary coefficient of the data points, For the The amount of data between the data point and the turning point of the backward data, For the The amount of data between the data point and the turning point of the forward data, It is a hyperbolic function. It can be understood that the current curve is continuously obtained based on the operation of the electrical control cabinet. The amount of data points , Will change accordingly with the output time; Stationary coefficient of data points By using the The data points and their neighboring data points constitute a data point sequence, and the corresponding stationary coefficient is obtained based on the ADF (Augmented Dickey-Fuller) test. The larger the stationary coefficient, the better the corresponding The better the stability of the data in the data point sequence, the better the stability of the data point sequence. I will not go into details here.

[0092] So far, the smoothness and ripple characteristics of the current curve have been analyzed, and the steady-state load factor and oscillation load factor of each current value have been obtained.

[0093] S14. Obtain the abnormal confidence of each current value according to the fluctuation degree, steady-state load factor and oscillation load factor of each current value.

[0094] Specifically, the fluctuation reflects the amplitude of the change of the current value; the steady-state load factor reflects the effect of the current value on the smooth characteristics of the curve; and the oscillation load factor reflects the effect of the current value under an unstable fluctuation state. Combining these three aspects to obtain the abnormal confidence (or abnormal factor) of each current value is to judge the possibility of abnormality of the current value by the change of the current value and its performance under different load states. The abnormal confidence can be solved based on the trained data processing model. The fluctuation, steady-state load factor and oscillation load factor of each current value are input into the trained data processing model, and the abnormal confidence is obtained based on the output result of the data processing model. The larger the abnormal confidence, the greater the probability that the current value is abnormal; otherwise, the smaller the probability of abnormality.

[0095] Exemplarily, step S14 includes sub-steps S14-1 to S14-2, which are specifically described as follows:

[0096] S14-1. Obtain the load stability ratio of each current value according to the ratio of the steady-state load coefficient to the oscillating load coefficient of each current value. The load stability ratio is calculated by the ratio of the steady-state load coefficient to the oscillating load coefficient of the current value. When the ratio is high, the steady-state load coefficient is large and the oscillating load coefficient is relatively small, indicating that the current value is in a relatively fluctuating state within the corresponding period and is subject to more interference or unstable factors; there may be some factors that cause the current to fluctuate frequently, and the higher the possibility of failure at the output time. Conversely, when the ratio is low, it means that the stability of the electrical control cabinet is relatively good under this current value.

[0097] S14-2. Obtain the abnormal confidence according to the load stability ratio and fluctuation of each current value. The load stability ratio is a stable result determined based on the dimension of the current value in the curve change period. When implementing the abnormal confidence calculation, the load stability ratio and the fluctuation can be respectively configured with corresponding weights, and the configured weight of the load stability ratio is recorded as the first weight coefficient, and the configured weight of the fluctuation is recorded as the second weight coefficient. The abnormal confidence is obtained based on the sum of the product of the first weight coefficient and the load stability ratio and the product of the second weight coefficient and the fluctuation.

[0098] Since the current of a steady-state load will not experience large fluctuations, the current of an oscillating load may experience large fluctuations, so it can be based on the formula:

[0099]

[0100] Get the current curve Anomaly confidence of data points , is the volatility, is the steady-state load factor, is the oscillation load coefficient. When the steady-state load coefficient is larger and the oscillation load coefficient is smaller, it means that the fluctuation of the current value is larger, indicating that the data point is more abnormal data.

[0101] S15. Output the fault diagnosis result of the electrical control cabinet according to all abnormal confidence levels and the preset confidence threshold.

[0102] Specifically, when the abnormality confidence is greater than the confidence threshold, the probability that the current value is abnormal is greater, the current value is determined to be abnormal data, and the fault diagnosis result of the output electrical control cabinet is an operating failure; conversely, when the abnormality confidence is less than or equal to the confidence threshold, it is determined that the current value is not abnormal, and the fault diagnosis result of the output electrical control cabinet is normal operation.

[0103] In practical applications, the electrical control cabinet is connected to multiple loads, and its operating state is complex. If the electrical control cabinet is determined to have an operating fault based on a certain current value, there may be a misjudgment. Based on this, in a specific implementation, step S15 includes sub-steps S15-1 to S15-2, which are specifically described as follows:

[0104] S15-1. When the abnormal confidence is greater than the confidence threshold, mark the output current corresponding to the abnormal confidence as abnormal current. The confidence threshold can be set to 0.8. When the abnormal confidence is greater than 0.8, it means that the output current is abnormal, and it is marked as abnormal current until all abnormal current markings are completed.

[0105] S15-2. When the density of abnormal current is greater than the preset value, the fault diagnosis result is determined to be an operation fault. The density of abnormal current can be calculated based on the time period and data volume of the abnormal current. For example, the time period of the abnormal current on the current curve is 60s, the data volume of the abnormal current is 5, and the density of abnormal current is 60 / 5=12s / piece. The greater the density of abnormal current in the area where the current curve is located, the more abnormal the output current in the area is. The fault diagnosis result is determined to be an operation fault, an alarm signal is sent to the terminal equipment, and professional maintenance personnel are arranged to troubleshoot the electrical control cabinet; otherwise, the electrical control cabinet is determined to be operating normally. This implementation strategy can further improve the accuracy of fault diagnosis of electrical control cabinets.

[0106] Based on the same technical concept as the diagnostic method, the embodiment of the present invention also provides a fault diagnosis system for an electrical control cabinet, see Figure 4 , Figure 4 40 is a schematic diagram of the structure of a fault diagnosis system. The fault diagnosis system includes a collection terminal 401 and a data processing terminal 402 .

[0107] The acquisition terminal can be configured as a current sensor, which is used to obtain the output current of the electrical control cabinet and output it to the outside.

[0108] The data processing terminal is connected to the acquisition terminal, and the data processing terminal is used to generate a current curve based on the output time corresponding to the output current, wherein the current curve is a curve of the output current of the electrical control cabinet changing with the output time; the data processing terminal is also used to perform fluctuation characteristic analysis on each current value of the current curve to obtain the fluctuation degree of each current value; according to the smoothness characteristics and tortuosity characteristics of each current value in the period on the current curve, the steady-state load coefficient and the oscillation load coefficient of each current value in the corresponding period are obtained respectively; according to the fluctuation degree, steady-state load coefficient and oscillation load coefficient of each current value, the abnormal confidence of each current value is obtained; according to all abnormal confidences and preset confidence thresholds, the fault diagnosis result of the electrical control cabinet is output.

[0109] The technical solution provided in the embodiments of the present invention has at least the following technical effects or advantages:

[0110] By obtaining the output current of the electrical control cabinet, and generating a current curve based on the output time corresponding to the output current, and analyzing the fluctuation characteristics of each current value of the current curve, the fluctuation degree of each current value can be obtained, and the fluctuation degree is used to characterize the fluctuation size between each current value and the adjacent current value; according to the smoothness and tortuosity characteristics of each current value in the period on the current curve, the steady-state load coefficient and the oscillation load coefficient of each current value in the corresponding period are obtained respectively, and the steady-state load coefficient and the oscillation load coefficient can respectively characterize the steady-state effect size and the oscillation effect size generated by each current value; according to the fluctuation degree, steady-state load coefficient and oscillation load coefficient of each current value, the abnormal confidence of each current value is obtained; according to all abnormal confidences and preset confidence thresholds, the fault diagnosis result of the electrical control cabinet is output. The technical solution performs data analysis based on the current value on the current curve, obtains the fluctuation size, steady-state effect size and oscillation effect size caused by each current value, and obtains the abnormal confidence that characterizes the abnormal characteristics of the current value after comprehensive analysis of multiple dimensions, and then accurately obtains the fault diagnosis result of the electrical control cabinet, so as to achieve the purpose of accurately implementing fault diagnosis based on the output current of the electrical control cabinet.

[0111] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0112] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A fault diagnosis method for an electrical control cabinet, characterized in that: The method comprises: Acquire the output current of the electrical control cabinet, and generate a current curve based on the output time corresponding to the output current, wherein the current curve is a curve of the output current of the electrical control cabinet changing with the output time; Performing a fluctuation characteristic analysis on each current value of the current curve to obtain a fluctuation degree of each current value; According to the smoothness characteristics and the wave characteristics presented by each current value in the period on the current curve, respectively obtain the steady-state load factor and the oscillation load factor of each current value in the corresponding period; Obtaining an abnormality confidence level of each current value according to the fluctuation level, the steady-state load factor and the oscillation load factor of each current value; Outputting the fault diagnosis result of the electrical control cabinet according to all abnormal confidences and preset confidence thresholds; The performing fluctuation characteristic analysis on each current value of the current curve to obtain the fluctuation degree of each current value includes: Obtaining a forward difference of the current value according to a difference between the current value on the current curve and the current value in the positive direction of time; Obtaining a backward difference of the current value according to a difference between the current value and the current value in the negative direction of time; According to the forward difference and the backward difference, the fluctuation degree of the current value is obtained: In the formula, The current curve The fluctuation of the current value, The first Current value, adjacent to The current value in the positive direction at the time, adjacent to The current value in the negative direction at the time, is a hyperbolic function; i is a natural number, representing any data point on the current curve; According to the smoothness characteristics of each current value in the period on the current curve, a steady-state load factor of each current value in the corresponding period is obtained, including: Dividing the current curve into a plurality of periodically changing curve segments according to an autocorrelation coefficient between a current current value and adjacent current values ​​on the current curve; According to the extreme value variation characteristics of the current values ​​in the multiple curve segments, a sine wave superposition coefficient of the current value in the cycle is obtained, wherein the sine wave superposition coefficient is a coefficient representing the number of sine waves superimposed on the current value in the cycle; According to the sine wave superposition coefficient and the average value of the fluctuation of the adjacent current value, the steady-state load factor of the current value is obtained: In the formula, Represents the steady-state load factor of the i-th current value on the current curve; is the average fluctuation of the adjacent current values; Indicates the current curve segment The sine wave superposition coefficient of each current value in the cycle; The step of obtaining the sine wave superposition coefficient of the current value in the cycle according to the extreme value variation characteristics of the current value in the multiple curve segments includes: Performing data extraction according to the current value represented by each curve segment to obtain a periodic data set of each curve segment; Performing extreme value extraction on the current value in each periodic data set to obtain the maximum current value and the minimum current value of each periodic data set; Obtaining a sine wave superposition coefficient of a current current value of the current curve segment according to a maximum current value and a minimum current value in a periodic data set of a current curve segment and its surrounding curve segments; The method of obtaining the sine wave superposition coefficient of the current current value of the current curve segment according to the maximum current value and the minimum current value in the periodic data set of the current curve segment and its surrounding curve segments includes: According to the formula: , get the first Sine wave superposition coefficient of current value , where N is the maximum current value of the current curve segment and its surrounding curve segments, The first The maximum current value, The current curve segment and its surrounding curve segments are The current value closest to the maximum current value, is the minimum current value of the current curve segment, The maximum current value and The distance between It is an exponential function with the natural constant e as the base; According to the wave characteristics of each current value in the period on the current curve, the oscillation load factor of each current value in the corresponding period is obtained, including: Determine whether the fluctuation degree of the current value adjacent to the current value on the current curve is greater than a preset value; If yes, the corresponding current value is marked as a data turning point; Obtaining a period factor representing a timing correlation between the data turning point and the current current value according to a distribution characteristic of the data turning point on the current curve; Get the period factor, including: Set a sliding window of appropriate size, slide the sliding window through the fluctuations of all adjacent current values, record each sliding window, determine the number of data turning points inside the sliding window, obtain the distance between the data turning points inside the sliding window, and calculate the period factor: , where For the The period factor of the current value is is the variable representing the sliding window position, n is a natural number; For the All neighboring current values ​​of data points sliding windows, the sliding window with the closest number of turning points in its internal data The distance between It is an exponential function with the natural constant e as the base; Indicates the calculation of variance; Obtaining an oscillation load factor of the current current value according to the period factor and the data interval, wherein the data interval is the amount of data between the current current value and the data turning point; The step of obtaining the oscillation load factor of the current current value according to the period factor and the data interval comprises: According to the formula , obtain the Oscillation load factor for each data point ,in, is the period factor, For the said The stationary coefficient of the data points, For the said The amount of data between the data point and the turning point of the backward data, For the said The amount of data between the data point and the turning point of the forward data, is a hyperbolic function; The obtaining of the abnormality confidence of each current value according to the fluctuation degree, the steady-state load factor and the oscillation load factor of each current value comprises: Obtaining a load stability ratio of each current value according to a ratio of a steady-state load factor to an oscillating load factor of each current value; According to the load stability ratio of each current value and the fluctuation, the abnormality confidence is obtained, and the calculation formula is: In the formula, The current curve The anomaly confidence of each data point; The current curve The volatility of the data points; The current curve The steady-state load factor for each data point; The current curve The oscillation load factor for each data point.

2. The fault diagnosis method of the electrical control cabinet according to claim 1, characterized in that: The method of outputting the fault diagnosis result of the electrical control cabinet according to all abnormal confidence levels and a preset confidence threshold value includes: When the abnormality confidence is greater than the confidence threshold, marking the output current corresponding to the abnormality confidence as an abnormal current; When the density of the abnormal current is greater than a preset value, the fault diagnosis result is determined to be an operation fault.

3. A fault diagnosis system for an electrical control cabinet, characterized in that: The system comprises: The acquisition terminal is used to obtain the output current of the electrical control cabinet and output it to the outside; A data processing terminal connected to the acquisition terminal, the data processing terminal is used to generate a current curve based on the output time corresponding to the output current, wherein the current curve is a curve of the output current of the electrical control cabinet changing with the output time; The data processing terminal is also used to perform fluctuation characteristic analysis on each current value of the current curve to obtain the fluctuation degree of each current value; obtain the steady-state load factor and the oscillation load factor of each current value in the corresponding period according to the smoothness and ripple characteristics of each current value in the period on the current curve; obtain the abnormality confidence of each current value according to the fluctuation degree, steady-state load factor and oscillation load factor of each current value; output the fault diagnosis result of the electrical control cabinet according to all abnormality confidences and preset confidence thresholds; The performing fluctuation characteristic analysis on each current value of the current curve to obtain the fluctuation degree of each current value includes: Obtaining a forward difference of the current value according to a difference between the current value on the current curve and the current value in the positive direction of time; Obtaining a backward difference of the current value according to a difference between the current value and the current value in the negative direction of time; According to the forward difference and the backward difference, the fluctuation degree of the current value is obtained: In the formula, The current curve The fluctuation of the current value, The first Current value, adjacent to The current value in the positive direction at the time, adjacent to The current value in the negative direction at the time, is a hyperbolic function; i is a natural number, representing any data point on the current curve; According to the smoothness characteristics of each current value in the period on the current curve, a steady-state load factor of each current value in the corresponding period is obtained, including: Dividing the current curve into a plurality of periodically changing curve segments according to an autocorrelation coefficient between a current current value and adjacent current values ​​on the current curve; According to the extreme value variation characteristics of the current values ​​in the multiple curve segments, a sine wave superposition coefficient of the current value in the cycle is obtained, wherein the sine wave superposition coefficient is a coefficient representing the number of sine waves superimposed on the current value in the cycle; According to the sine wave superposition coefficient and the average value of the fluctuation of the adjacent current value, the steady-state load factor of the current value is obtained: In the formula, Represents the steady-state load factor of the i-th current value on the current curve; is the average fluctuation of the adjacent current values; Indicates the current curve segment The sine wave superposition coefficient of each current value in the cycle; The step of obtaining the sine wave superposition coefficient of the current value in the cycle according to the extreme value variation characteristics of the current value in the multiple curve segments includes: Performing data extraction according to the current value represented by each curve segment to obtain a periodic data set of each curve segment; Performing extreme value extraction on the current value in each periodic data set to obtain the maximum current value and the minimum current value of each periodic data set; Obtaining a sine wave superposition coefficient of a current current value of the current curve segment according to a maximum current value and a minimum current value in a periodic data set of a current curve segment and its surrounding curve segments; The method of obtaining the sine wave superposition coefficient of the current current value of the current curve segment according to the maximum current value and the minimum current value in the periodic data set of the current curve segment and its surrounding curve segments includes: According to the formula: , get the first Sine wave superposition coefficient of current value , where N is the maximum current value of the current curve segment and its surrounding curve segments, The first The maximum current value, The current curve segment and its surrounding curve segments are The current value closest to the maximum current value, is the minimum current value of the current curve segment, The maximum current value and The distance between It is an exponential function with the natural constant e as the base; According to the wave characteristics of each current value in the period on the current curve, the oscillation load factor of each current value in the corresponding period is obtained, including: Determine whether the fluctuation degree of the current value adjacent to the current value on the current curve is greater than a preset value; If yes, the corresponding current value is marked as a data turning point; Obtaining a period factor representing a timing correlation between the data turning point and the current current value according to a distribution characteristic of the data turning point on the current curve; Get the period factor, including: Set a sliding window of appropriate size, slide the sliding window through the fluctuations of all adjacent current values, record each sliding window, determine the number of data turning points inside the sliding window, obtain the distance between the data turning points inside the sliding window, and calculate the period factor: , where For the The period factor of the current value is is the variable representing the sliding window position, n is a natural number; For the All neighboring current values ​​of data points sliding windows, the sliding window with the closest number of turning points in its internal data The distance between It is an exponential function with the natural constant e as the base; Indicates the calculation of variance; Obtaining an oscillation load factor of the current current value according to the period factor and the data interval, wherein the data interval is the amount of data between the current current value and the data turning point; The step of obtaining the oscillation load factor of the current current value according to the period factor and the data interval comprises: According to the formula , obtain the Oscillation load factor for each data point ,in, is the period factor, For the said The stationary coefficient of the data points, For the said The amount of data between the data point and the turning point of the backward data, For the said The amount of data between the data point and the turning point of the forward data, is a hyperbolic function; The obtaining of the abnormality confidence of each current value according to the fluctuation degree, the steady-state load factor and the oscillation load factor of each current value comprises: Obtaining a load stability ratio of each current value according to a ratio of a steady-state load factor to an oscillating load factor of each current value; According to the load stability ratio of each current value and the fluctuation, the abnormality confidence is obtained, and the calculation formula is: In the formula, The current curve The anomaly confidence of each data point; The current curve The volatility of the data points; The current curve The steady-state load factor for each data point; The current curve The oscillation load factor for each data point.

Images