Relay Fault Detection Method

Abstract

The present application relates to the technical field of electrical variable measurement, and specifically to a relay fault detection method. The method includes: obtaining multiple parameter sequences; dividing the contact resistance sequence into multiple segments, and obtaining the coefficient of change of the contact resistance according to the fitting straight lines of the multiple segments; obtaining the change trend measurement according to the starting time sequence, the closing time sequence and the closing voltage sequence, and obtaining the wear coefficient of the contact point according to the trend and slope of the overtravel time series, the variance of the starting voltage sequence, the consistency of the closing voltage sequence and the closing time sequence, and the change trend measurement; adjusting the range of the wear coefficient and the change coefficient, adjusting the threshold interval according to the adjusted wear coefficient and the change coefficient to obtain the adjusted threshold interval; comparing the contact resistance with the upper and lower limits of the adjusted threshold interval to complete the fault detection of the relay. The present application improves the accuracy of relay fault detection and ensures the effect and performance of relay fault detection.

Description

Technical Field

[0001] The present application relates to the technical field of electrical variable measurement, and in particular to a relay fault detection method. Background Art

[0002] Electromagnetic relays are widely used in power systems, industrial control, aerospace and other fields. They play an important role in various industries due to their simple structure, strong anti-electromagnetic interference ability and good switching characteristics. Therefore, the stability of the relay becomes a crucial issue. Once the relay fails, it may cause problems in the entire system or even a devastating failure. Therefore, it is necessary to measure the contact resistance to detect relay failure.

[0003] Normal contact resistance is usually only tens of milliohms, which is difficult to measure accurately. At the same time, factors such as contact wear and temperature can also cause changes in contact resistance. In engineering, a detection threshold is usually given. This threshold is generally high, so there is almost no false detection, but at the same time the false alarm rate is high, making it difficult to detect the hidden faults of the relay; in order to overcome the defects of a single threshold, many existing solutions use the method of expanding a single threshold into a threshold interval. Although the false alarm rate is reduced to a certain extent, this method is often based on the interval estimation of the measured contact resistance, and the contact resistance is greatly affected by the operating conditions of the relay. When the accuracy of the contact resistance is poor or insufficient for accurate judgment, it is almost inevitable that errors will also occur in the follow-up, which increases the false alarm rate to a certain extent. Therefore, how to improve the accuracy of relay fault detection and reduce the false alarm rate and false alarm rate has become an urgent problem to be solved in relay detection. Summary of the invention

[0004] In order to solve the technical problem of low accuracy of relay fault detection, the present application provides a relay fault detection method, and the technical solution adopted is as follows:

[0005] The present application proposes a relay fault detection method, which comprises the following steps:

[0006] The normalized parameter sequence of each parameter collected by different sensors includes contact resistance sequence, overtravel time sequence, starting voltage sequence, closing voltage sequence, starting time sequence and closing time sequence;

[0007] The contact resistance sequence is divided into multiple segments, and the fitting straight line of each segment is obtained; the coefficient of variation of the contact resistance is obtained according to the determination coefficient of all the fitting straight lines, the slope of the fitting straight lines, and the difference between the contact resistance of the last part of the fitting straight line and the minimum contact resistance;

[0008] Perform a straight line fitting on the overtravel time series to obtain the slope of the corresponding sequence; divide the starting time sequence, closing time sequence and closing voltage sequence into left and right sequences and then decompose them to obtain the trend sequence, and perform differential calculation on the trend sequence to obtain the starting left differential sequence, starting right differential sequence, closing left differential sequence, closing right differential sequence, voltage left differential sequence and voltage right differential sequence; obtain the change trend measurement according to the proportion of positive and negative numbers in all differential sequences; obtain the wear coefficient of the contact point according to the proportion of positive numbers in the first-order difference sequence of the overtravel time series, the slope of the overtravel time series, the variance of the starting voltage sequence, the consistency of the closing voltage sequence and the closing time sequence, and the change trend measurement;

[0009] Adjust the range of the wear coefficient and the variation coefficient, set the range of the wear coefficient to [0, 1], and the range of the variation coefficient to [-1, 1]; set an empirical threshold, obtain the threshold interval through the empirical threshold, and adjust the threshold interval according to the adjusted wear coefficient and variation coefficient to obtain the adjusted threshold interval; when the variation coefficient is small, a threshold interval with a smaller interval range is required; when the variation coefficient is large and the wear coefficient is small, the upper limit of the threshold interval is slightly increased, and the lower limit is slightly decreased, and a threshold interval with a slightly increased threshold interval is required; when the variation coefficient is large and the wear coefficient is large, the upper limit of the threshold interval is significantly increased, and the lower limit is significantly decreased, and a threshold interval with a significantly increased threshold interval is required;

[0010] The contact resistance at the current moment is compared with the upper and lower limits of the adjusted threshold range to complete the fault detection of the relay.

[0011] In the above scheme, according to the time series change characteristics of the relay contact resistance, the change coefficient of the contact resistance is calculated. This coefficient quantifies the change characteristics of the contact resistance. Combined with the change of the contact resistance when the relay contact is worn, the diagnosis ability of the abnormal relay contact resistance is improved; then according to the relationship and change trend between the contact wear condition and each voltage and the occurrence time, the wear coefficient is calculated. This coefficient enhances the measurement of the wear condition of the relay contact and helps to determine the subsequent threshold interval; finally, the threshold interval is determined and used for relay fault detection, which enhances the performance and effect of relay fault detection. In this way, when performing relay fault detection, the change of contact resistance is utilized, and the wear condition of the contact is taken into account. The threshold interval is adaptively adjusted according to the contact wear condition, and the excessively high false alarm rate is avoided while reducing the missed alarm rate, thereby improving the accuracy of relay fault detection and ensuring the effect and performance of relay fault detection.

[0012] In one embodiment, after the contact resistance sequence is divided into multiple sections, for each section, a straight line fitting is performed using the contact resistance as the ordinate and the position number thereof as the abscissa to obtain a fitting straight line for each section.

[0013] In one embodiment, the method for obtaining the coefficient of variation of the contact resistance according to the determination coefficient of all the fitting straight lines, the slope of the fitting straight line, and the difference between the contact resistance of the last part of the fitting straight line and the minimum contact resistance is:

[0014] For each segment of the fitting straight line, obtain the elements in the upper quartile, and take the average of the difference between the contact resistance of all elements greater than the upper quartile and the minimum contact resistance in each segment of the fitting straight line as the range measure of each segment of the fitting straight line;

[0015] The coefficient of determination of the fitted straight line is recorded as the fitting effect of the fitted straight line;

[0016] Obtain the coefficient of variation of the contact resistance according to the range measurement, slope and fitting effect of all segments of the fitting straight line;

[0017] The coefficient of variation of contact resistance is positively correlated with the range measure and slope of the fitting line, and negatively correlated with the fitting effect of the fitting line.

[0018] In one embodiment, the expression of the coefficient of variation is: ; is the coefficient of variation of contact resistance, is the number of segments of the contact resistance series, It is The slope of the segment fitting line, It is The fitting effect of the segment fitting line, It is The range measure in the segment, is the tuning coefficient.

[0019] In one embodiment, the maximum value of the elements in the starting time sequence, the closing time sequence and the closing voltage sequence is obtained, and a sequence is obtained on the left side and the right side of the maximum value respectively. A decomposition algorithm is used to obtain the trend sequence of all left and right sequences, and the first-order difference is calculated for each trend sequence to obtain the starting left differential sequence, the starting right differential sequence, the closing left differential sequence, the closing right differential sequence, the voltage left differential sequence and the voltage right differential sequence respectively.

[0020] In one embodiment, the method for obtaining the change trend measurement according to the proportion of positive and negative numbers in all difference sequences is:

[0021] Count the proportion of positive numbers in the starting left differential sequence, closing left differential sequence, and voltage left differential sequence;

[0022] Count the proportion of negative numbers in the starting right differential sequence, closing right differential sequence, and voltage right differential sequence;

[0023] The expression of the change trend measurement is: , where It represents the proportion of positive numbers in the first-order difference sequence of the left sequence of sequence n. It represents the proportion of negative numbers in the first-order difference sequence of the right sequence of sequence n. Sequence n takes different values ​​to represent the starting time sequence, closing time sequence, and closing voltage sequence respectively. M represents the change trend measure.

[0024] In one embodiment, the method for obtaining the wear coefficient of the contact point according to the proportion of positive numbers of the first-order difference sequence of the over-travel time series, the slope of the over-travel time series, the variance of the starting voltage sequence, the consistency of the closing voltage sequence and the closing time sequence, and the change trend measurement is:

[0025] The correlation coefficient between the closing voltage sequence and the closing time sequence is taken as the consistency of the two sequences;

[0026] The proportion of positive numbers in the first-order difference sequence of the over-range time series is recorded as the trend of the over-range time series;

[0027] The wear coefficient of the contact point is positively correlated with the negative number of the slope of the overtravel time series, the change trend measure, and the consistency of the closing voltage series and the closing moment series, and is negatively correlated with the variance of the starting voltage series and the trend of the overtravel time series.

[0028] In one embodiment, the method for adjusting the range of the wear coefficient and the variation coefficient is: , , where A is the coefficient of change of contact resistance, B is the wear coefficient of the contact point, is the adjusted coefficient of variation, is the adjusted wear coefficient.

[0029] In one embodiment, the threshold interval is adjusted according to the adjusted wear coefficient and the variation coefficient to obtain the adjusted threshold interval: : , are the lower and upper limits of the adjusted threshold interval, , are the lower and upper limits of the original threshold interval, , are the adjusted coefficient of variation and the adjusted coefficient of wear, respectively.

[0030] In one embodiment, the method for completing relay fault detection by comparing the contact resistance at the current moment with the upper and lower limits of the adjusted threshold interval is:

[0031] The area with a value less than the lower limit of the adjusted threshold interval is identified as the normal operating condition area, the area with a value greater than the upper limit of the adjusted threshold interval is identified as the serious fault area, and the area within the adjusted threshold interval is identified as the hidden fault area;

[0032] If the contact resistance is in the severe fault region, there is a relay fault.

[0033] The beneficial effects of this application are:

[0034] According to the time series change characteristics of the relay contact resistance, the change coefficient of the contact resistance is calculated. This coefficient quantifies the change characteristics of the contact resistance. Combined with the change of the contact resistance when the relay contact is worn, the diagnosis ability of the abnormal relay contact resistance is improved; then according to the relationship and change trend between the contact wear condition and each voltage and the occurrence time, the wear coefficient is calculated. This coefficient enhances the measurement of the wear condition of the relay contact and helps to determine the subsequent threshold interval; finally, the threshold interval is determined and used for relay fault detection, which enhances the performance and effect of relay fault detection. In this way, when performing relay fault detection, the change of contact resistance is utilized, and the wear condition of the contact is considered. The threshold interval is adaptively adjusted according to the contact wear condition, which avoids excessive false alarm rate while reducing the missed alarm rate, improves the accuracy of relay fault detection, and ensures the effect and performance of relay fault detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] 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.

[0036] Figure 1 A flow chart of a relay fault detection method provided by an embodiment of the present application;

[0037] Figure 2 This is a schematic diagram of the relay housing;

[0038] Figure 3 This is the internal schematic diagram of the relay;

[0039] Figure 4 It is a schematic diagram of the fitting straight line of each segment of the contact resistance series;

[0040] Figure 5 It is a schematic diagram of closing time and closing voltage;

[0041] Figure 6 Schematic diagram of starting time and starting voltage. DETAILED DESCRIPTION

[0042] In order to further explain the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the following is a detailed description of the relay fault detection method proposed in the present application, its specific implementation, structure, features and effects, in conjunction 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.

[0043] 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 application belongs.

[0044] Relay fault detection method embodiment:

[0045] The specific scheme of the relay fault detection method provided by the present application is described in detail below with reference to the accompanying drawings.

[0046] See also Figure 1 , which shows a flow chart of a relay fault detection method provided by an embodiment of the present application, the method comprising the following steps:

[0047] Step S001, collecting the normalized parameter sequence of each historical parameter through different sensors, including contact resistance sequence, overtravel time sequence, starting voltage sequence, closing voltage sequence, starting time sequence and closing time sequence.

[0048] An electromagnetic relay is a switch that uses an electromagnet to control the working current. It includes a housing, an electromagnetic module, and a contact module. The electromagnetic module consists of an iron core, a coil yoke, and an armature. The iron core is fixed to the yoke, the coil is sleeved on the iron core, the armature is located at the other end of the iron core, and the two ends of the coil are welded to two lead pins; the contact module consists of a base plate, a static spring, and a moving spring. The lead wires, static spring, and moving spring all pass through the base plate. The housing is fixed to the base plate, and one end of the moving spring is welded to the armature. The outer shell and internal schematic diagram of the electromagnetic relay are shown in the figure below. Figure 2 and Figure 3 shown.

[0049] The two probes of the milliohmmeter are connected to the static contact and the moving contact of the relay respectively. When the protected element in the power system fails, the static contact and the moving contact will touch each other. The milliohmmeter is used to collect the contact resistance at this time. After the static contact and the moving contact touch each other, the contact continues to move. At this time, the moving contact and its corresponding armature continue to move until the armature is completely closed. The displacement between the first contact of the two contacts and the complete closure of the armature is called overtravel, and the corresponding movement time is called overtravel time, which is collected and obtained by the displacement sensor and the timer. During the displacement of the armature, the movement of the armature causes the conductor in the coil to cut the magnetic flux lines, thereby generating a motional electromotive force, forcing the voltage to drop. After the armature is closed, the voltage rises. By installing an oscilloscope on the coil power supply circuit of the relay to monitor the voltage data, the voltage at the moment when the armature starts to move is recorded as the starting voltage, the voltage when the armature is fully closed is recorded as the closing voltage, and the time corresponding to the starting voltage and the closing voltage is recorded as the starting time and the closing time respectively.

[0050] Each time the relay contacts and the armature closes is recorded as a cycle. The contact resistance, overtravel time, starting voltage, closing voltage, starting time and closing time of all cycles in the relay history are counted, and each is recorded as a parameter of a cycle. Each parameter is sorted in sequence according to the sequence of the cycles to obtain the parameter sequence of each parameter. Each parameter sequence is normalized to the maximum value to obtain the contact resistance sequence, overtravel time sequence, starting voltage sequence, closing voltage sequence, starting time sequence and closing time sequence.

[0051] At this point, the normalized parameter sequence of each parameter is obtained.

[0052] Step S002, divide the contact resistance sequence into multiple segments, obtain the fitting straight line of each segment; obtain the contact resistance variation coefficient according to the determination coefficient of all fitting straight lines, the slope of the fitting straight line and the difference between the contact resistance of the last part of the fitting straight line and the minimum contact resistance.

[0053] In relays, the detection of contact resistance is one of the main means of detecting relay failures. However, since the contact resistance in relays is often relatively small and difficult to measure accurately, traditional methods expand the single threshold to a threshold interval for judgment in order to improve the accuracy of relay fault judgment using contact resistance. However, these methods often only consider the contact resistance. When the contact resistance accuracy is poor or insufficient for accurate judgment, the expanded threshold interval will inevitably have corresponding errors. In the operation of the relay, in addition to the contact resistance, there are other characteristics that can also synchronously reflect the operating status of the relay. When the relay fails, various monitoring data will also undergo certain changes. This application analyzes multiple features and adjusts the interval.

[0054] During the operation of the relay, the change trend of the contact resistance is closely related to the condition of the contacts in the relay. When there is no fault in the relay, the contact surface is relatively smooth and there is no severe wear, so the contact area is relatively large, so the contact resistance will not change significantly and is relatively stable; when the contact is significantly worn, the roughness of the contact surface increases and the contact area decreases, which increases the contact resistance. At the same time, due to the long-term wear of the contact, the anti-oxidation measures on the contact surface fail, and attachments such as but not limited to oxide films are generated, which increases the contact resistance. As the attachments fall off, the resistance decreases, so the change of the contact resistance is unstable; and with long-term wear, the contact surface may be severely worn or ablated, at which time the contact resistance may even increase by more than ten times, and even fail to form a reliable electrical contact.

[0055] In the present application, the contact resistance sequence is divided into n segments. If the contact resistance sequence cannot be evenly divided, the difference in quantity between different segments is made as small as possible. In the present embodiment, the remainder is added to the previous segment in sequence until the remainder is added. In the present embodiment, n is set to 5. In each segment, a straight line is fitted using the contact resistance as the ordinate and the position number as the abscissa to obtain a fitted straight line for each segment. The method of straight line fitting is a well-known technique and will not be elaborated in the present application. The fitted straight line of the contact resistance of each segment is as follows: Figure 4 shown.

[0056] First, the coefficient of determination of each segment of the fitting straight line is calculated. The calculation method of the coefficient of determination is a well-known method and will not be described in detail here. The coefficient of determination is used to represent the fitting effect of each segment, and then the slope of each segment of the fitting straight line is calculated. In order to avoid the abnormal increase of a certain data or a few data in the segment, which leads to excessive range and affects the accuracy of the judgment, for each segment of the fitting straight line, the mean of the difference between the contact resistance of the last part of the fitting straight line and the minimum contact resistance is selected as the range measurement. In this embodiment, the elements of the upper quartile are obtained, and the mean of the difference between the contact resistance of all elements greater than the upper quartile in each segment of the fitting straight line and the minimum contact resistance is used as the range measurement of each segment of the fitting straight line.

[0057] The coefficient of variation of the contact resistance is obtained based on the range measurement, slope and fitting effect of all segments of the fitting straight line.

[0058] The coefficient of variation of contact resistance is positively correlated with the range measure and slope of the fitting line, and negatively correlated with the fitting effect of the fitting line.

[0059] It should be noted that positive correlation means that when one variable increases, the other variable also increases, and the two variables change in the same direction. When one variable changes from large to small or from small to large, the other variable also changes from large to small or from small to large. The specific relationship is determined by actual application and this application does not impose any special restrictions.

[0060] It should be noted that negative correlation means that when one variable increases, the other variable decreases accordingly, and the two variables change in opposite directions. When one variable changes from large to small or from small to large, the other variable also changes from small to large or from large to small. The specific relationship is determined by actual application and this application does not impose any special restrictions.

[0061] Preferably, in this embodiment, the expression of the variation coefficient of the contact resistance is: ; is the coefficient of variation of contact resistance, is the number of segments of the contact resistance series, It is The slope of the segment fitting line, It is The fitting effect of the segment fitting line, It is The range measure in the segment, is the parameter adjustment coefficient, and its empirical value is 1. The purpose of the parameter adjustment coefficient is to prevent the situation where the coefficient of variation is 0 when the slope is 0. The purpose of the exponential function is to increase the impact of the range.

[0062] When the contact resistance of the relay changes more, according to the change trend characteristics of the contact resistance, there is no significant change at first, then there is obvious abnormal fluctuation, and finally there is obvious abnormal increase. When the contact resistance does not change significantly, the slope of the fitting line of the contact resistance is small, the fitting effect is good, and the range is small, that is, , All are smaller, When the contact resistance shows obvious abnormal fluctuation, the slope of the fitting line of the contact resistance increases relatively, and the fitting effect is relatively poor due to the increase in fluctuation, and the range also increases. , Relatively increased, Small; when the contact resistance increases abnormally, the slope and range of the fitting line are larger, and the fitting effect is worse; thus, the more drastic the change of the relay contact resistance, the greater the coefficient of change. On the contrary, when the change degree of the relay contact resistance is small, the change coefficient Smaller.

[0063] At this point, the coefficient of variation of the contact resistance is obtained.

[0064] Step S003, perform straight line fitting on the over-travel time series to obtain the slope of the corresponding sequence; divide the starting time sequence, closing time sequence and closing voltage to obtain the differential sequence of their trends; obtain the change trend measurement according to the proportion of positive and negative numbers in all differential sequences; obtain the wear coefficient of the contact point according to the proportion of positive numbers in the first-order difference sequence of the over-travel time series, the slope of the over-travel time series, the variance of the starting voltage sequence, the consistency of the closing voltage sequence and the closing time sequence, and the change trend measurement.

[0065] According to the above steps, the change coefficient of contact resistance is obtained, and the change coefficient only considers the change degree of relay contact resistance. However, since the operation of the relay will also cause changes including but not limited to temperature, these changes will also affect the contact resistance, but they are not relay failures. Therefore, it is necessary to combine other features for further judgment.

[0066] The change of contact resistance caused by relay failure is actually the wear of relay contacts, and contact wear is irreversible. At the same time, contact wear will also cause changes in the starting voltage, closing voltage, and the time when each voltage appears. In the operation of the relay, as the contact wear increases, the starting time sequence has obvious fluctuations, and the fluctuation trend is first increasing and then decreasing, while the fluctuation trend of the starting voltage sequence is relatively small; while the fluctuation trends of the closing voltage sequence and the closing time sequence are relatively consistent, both increasing first and then decreasing. The schematic diagram of the four sequences is shown in the figure below. Figure 5 , Figure 6 shown.

[0067] At the same time, as the wear of the relay contacts increases, the contact thickness gradually decreases, the contact stroke of the dynamic and static contacts increases, and the stroke of the armature closure remains unchanged. The increase in the contact stroke of the dynamic and static contacts means that the stroke from contact to armature closure decreases, that is, the overtravel gradually decreases, so the overtravel time also gradually decreases. Since the wear of the relay contacts is irreversible, the wear of the relay contacts shows an overall upward trend, and there is almost no fluctuation increase, that is, the wear at one moment must be no less than that at the previous moment, and since the wear and overtravel time are negatively correlated, the overtravel time series shows a decreasing trend as a whole, and there is no fluctuation decrease or the fluctuation decrease is less. When the relay is normal or the fault is small, there is no obvious wear on the contacts, so the overtravel time is basically stable.

[0068] For the super-range time series, its first-order difference sequence is obtained and recorded as the super-range time difference sequence, and the proportion of positive numbers in the super-range time difference sequence is recorded as the trend of the super-range time series; the time of each element in the super-range time series is used as the ordinate and the index as the abscissa to perform straight line fitting to obtain the straight line equation, and the slope of the straight line equation is recorded as the slope of the super-range time series.

[0069] Get the maximum value of the element in the starting time sequence, record the sequence composed of all elements on the left side of the maximum value as the starting left sequence, and record the sequence composed of all elements on the right side of the maximum value as the starting right sequence; similarly, get the maximum value of the element for the closing time sequence, and get the closed left sequence and closed right sequence according to the maximum value of the closing time; get the maximum value of the element for the closed voltage sequence, and get the voltage left sequence and voltage right sequence according to the maximum value of the closed voltage. Decompose the starting left sequence, the starting right sequence, the closed left sequence, the closed right sequence, the voltage left sequence and the voltage right sequence to obtain their trend sequences, and the algorithm used is the STL decomposition algorithm; and calculate the first-order difference sequence of each trend sequence and record them as the starting left difference sequence, the starting right difference sequence, the closed left difference sequence, the closed right difference sequence, the voltage left difference sequence and the voltage right difference sequence.

[0070] In the starting left differential sequence, starting right differential sequence, closing left differential sequence, closing right differential sequence, voltage left differential sequence and voltage right differential sequence, the proportion of positive numbers in the left sequence is counted, and the proportion of negative numbers in the right sequence is counted. The change trend measurement is obtained based on the proportion of the starting time, closing time and closing voltage trend difference.

[0071] Preferably, in this embodiment, the expression of the change trend measurement is: , where It represents the proportion of positive numbers in the first-order difference sequence of the left sequence of sequence n. It represents the proportion of negative numbers in the first-order difference sequence of the right sequence of sequence n. When n takes different values, sequence n represents the starting time sequence, closing time sequence, and closing voltage sequence respectively; M represents the change trend measure.

[0072] Preferably, in another embodiment of the present application, the expression of the change trend metric is: , where It represents the proportion of positive numbers in the first-order difference sequence of the left sequence of sequence n. It represents the proportion of negative numbers in the first-order difference sequence of the right sequence of sequence n. Sequence n takes different values ​​to represent the starting time sequence, closing time sequence, and closing voltage sequence respectively. M represents the change trend measure.

[0073] As the wear increases, the starting time sequence, closing voltage sequence, and closing time sequence show a trend of first increasing and then decreasing, that is, M is larger.

[0074] The correlation coefficient between the closing voltage sequence and the closing time sequence is calculated as the consistency of the two sequences; in this embodiment, the correlation coefficient of the sequence is obtained by the Pearson correlation coefficient.

[0075] The wear coefficient of the contact point is obtained based on the variance of the starting voltage sequence, the slope of the overtravel time series, the change trend measure, the trend of the overtravel time series, and the correlation coefficient between the closing voltage sequence and the closing time series.

[0076] The wear coefficient of the contact point is positively correlated with the negative number of the slope of the overtravel time series, the change trend measure, and the consistency of the closing voltage series and the closing moment series, and is negatively correlated with the variance of the starting voltage series and the trend of the overtravel time series.

[0077] Preferably, in this embodiment, the wear coefficient of the contact point is: , where represents the variance of the starting voltage sequence, represents the slope of the overtravel time series, Indicates the consistency between the closing voltage sequence and the closing time sequence, Represents the trend measurement, represents the trend of the over-range time series, Indicates the wear coefficient of the contact point.

[0078] When the contacts in the relay are worn, due to the irreversibility of wear, the overtravel time series shows a monotonically decreasing trend over time, with less fluctuations, and the more severe the wear, the more the overtravel time decreases. At this time, the slope of the overtravel time series is negative, and the slope is small. Therefore, when the slope is negative, the larger its value is, the greater the wear coefficient is. The smaller the proportion of positive numbers in the first-order difference sequence of the overtravel time series is, the less fluctuations there are, and the overtravel time series is more likely to be worn, and the greater the wear coefficient is at this time.

[0079] Secondly, as the wear increases, the starting time sequence, closing voltage sequence, and closing time sequence show a trend of increasing first and then decreasing. At this time, the larger the change trend measure, the larger the wear coefficient. The starting voltage sequence is basically unchanged, and the variance of the starting voltage sequence is small. Therefore, the smaller the variance of the starting voltage sequence, the larger the wear coefficient. There is consistency in the change trends of the closing voltage sequence and the closing time sequence. The stronger the consistency, the larger the wear coefficient; therefore, when the contacts in the relay are worn, the wear coefficient is large, and the more severe the wear, the larger the wear coefficient. On the contrary, when there is no obvious wear on the relay contacts, the wear coefficient is small.

[0080] At this point, the wear coefficient of the contact point is obtained.

[0081] Step S004: adjusting the range of the wear coefficient and the variation coefficient, and adjusting the threshold interval according to the adjusted wear coefficient and the variation coefficient to obtain the adjusted threshold interval.

[0082] In the traditional fault detection method for relays, an empirical threshold needs to be set, and then the threshold interval of contact resistance is obtained by using the interval estimation method to estimate the empirical threshold through the contact resistance sequence. However, the expansion process of the threshold interval only uses the contact resistance sequence. Since the contact resistance is greatly affected by the operating conditions of the relay, the threshold interval expanded only based on the contact resistance will reduce the missed alarm rate, but also increase the false alarm rate. It is impossible to accurately distinguish whether the change in contact resistance is caused by contact wear or due to changes in normal operating conditions, and it is impossible to accurately detect relay faults.

[0083] The coefficient of variation and the coefficient of wear analyze the changes caused by the contacts from different angles. The coefficient of variation is obtained by analyzing the change in contact resistance. Since the measurement of contact resistance is relatively simple, it can improve the efficiency of relay fault detection. The wear coefficient is obtained through residual data such as overtravel time. This indicator measures the degree of contact wear and can improve the accuracy of relay fault detection. Therefore, the combination of these two indicators not only ensures the efficiency of relay fault detection, but also guarantees the accuracy of fault detection, which can effectively improve the effect and performance of relay fault detection.

[0084] The wear of relay contacts will definitely lead to changes in contact resistance, and the wear of relay contacts can be measured by the wear coefficient. In addition to contact wear, changes in contact resistance may also be caused by changes in operating conditions. This situation does not belong to relay failure or contact wear. Therefore, when the wear coefficient is large, there must be a large variation coefficient. At this time, the probability of contact wear and relay failure is high; when the variation coefficient is large and the wear coefficient is small, there is almost no contact wear, and the probability of relay failure is relatively small; when the variation coefficient is small, there is almost no abnormal change in contact resistance, and the relay is almost fault-free.

[0085] Through the above analysis, the present application adjusts the threshold interval according to the variation coefficient and the wear coefficient to improve the performance and effect of relay fault detection.

[0086] First, the empirical threshold is set to 0.2 milliohms. In this embodiment, the empirical threshold is taken as follows: the empirical threshold and the contact resistance sequence are used as inputs, and the threshold interval of the contact resistance is obtained by using the interval estimation method. The interval estimation method is to expand a single threshold to an interval. This algorithm is a well-known technology and will not be described in detail in this application.

[0087] First, the variation coefficient and wear coefficient are normalized. Since the variation coefficient has different effects in different situations, the wear coefficient is adjusted to the range of [0, 1] and the variation coefficient is adjusted to the range of [-1, 1]. If the variation coefficient and wear coefficient are in the range of [0, 1], then The value is -1. The value is 0, and the adjustment method is as follows: , , where A is the coefficient of change of contact resistance, B is the wear coefficient of the contact point, is the adjusted coefficient of variation, is the adjusted wear coefficient.

[0088] The threshold interval is adjusted according to the adjusted variation coefficient and wear coefficient. The adjustment method is: : , are the lower and upper limits of the adjusted threshold interval, , are the lower and upper limits of the original threshold interval, , are the adjusted coefficient of variation and the adjusted coefficient of wear, respectively.

[0089] When the coefficient of variation is small, the relay is almost fault-free. At this time, a threshold interval with a smaller interval range is required to avoid excessive false alarm rate. At the same time, the coefficient of variation is small. The value range of is negative at this time, so the lower limit increases and the upper limit decreases, so the threshold interval is smaller. When the coefficient of variation is large, in order to reduce the false alarm rate, the range of the threshold interval can be appropriately increased. At this time, if the failure probability of the relay is small, that is, the contact is not obviously worn, the upper limit of the threshold interval is slightly increased and the lower limit is slightly decreased, so that the relative advantage of the threshold interval range increases; if the contact is obviously worn at the same time, the upper limit of the threshold interval is significantly increased and the lower limit is significantly decreased, so that the threshold interval range is larger, thereby avoiding the false alarm rate.

[0090] At this point, the adjusted threshold range is obtained.

[0091] Step S005, comparing the current contact resistance with the upper and lower limits of the adjusted threshold value interval to complete relay fault detection.

[0092] According to the existing fault analysis methods and experience, by classifying the measurement data of contact resistance, the state of the relay contact system can be divided into three categories: normal working condition, hidden fault, and serious fault. According to the above steps, the adjusted threshold range is obtained.

[0093] The area smaller than the lower limit of the adjusted threshold interval is identified as the normal operating condition area, the area larger than the upper limit of the adjusted threshold interval is identified as the serious fault area, and the area within the adjusted threshold interval is identified as the hidden fault area.

[0094] For the contact resistance collected at the current moment, if it is in the severe fault area, it is considered that there is a relay fault, thereby completing the relay fault detection.

[0095] It should be noted that the above-described embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

[0096] 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 relay fault detection method, characterized in that: The method comprises the following steps: The normalized parameter sequence of each parameter collected by different sensors includes contact resistance sequence, overtravel time sequence, starting voltage sequence, closing voltage sequence, starting time sequence and closing time sequence; The contact resistance sequence is divided into multiple segments, and the fitting straight line of each segment is obtained; the coefficient of variation of the contact resistance is obtained according to the determination coefficient of all the fitting straight lines, the slope of the fitting straight lines, and the difference between the contact resistance of the last part of the fitting straight line and the minimum contact resistance; Perform a straight line fitting on the overtravel time series to obtain the slope of the corresponding sequence; divide the starting time sequence, closing time sequence and closing voltage sequence into left and right sequences and then decompose them to obtain the trend sequence, and perform differential calculation on the trend sequence to obtain the starting left differential sequence, starting right differential sequence, closing left differential sequence, closing right differential sequence, voltage left differential sequence and voltage right differential sequence; obtain the change trend measurement according to the proportion of positive and negative numbers in all differential sequences; obtain the wear coefficient of the contact point according to the proportion of positive numbers in the first-order difference sequence of the overtravel time series, the slope of the overtravel time series, the variance of the starting voltage sequence, the consistency of the closing voltage sequence and the closing time sequence, and the change trend measurement; Adjust the range of the wear coefficient and the variation coefficient, set the range of the wear coefficient to [0, 1], and the range of the variation coefficient to [-1, 1]; set an empirical threshold, obtain the threshold interval through the empirical threshold, and adjust the threshold interval according to the adjusted wear coefficient and variation coefficient to obtain the adjusted threshold interval; when the variation coefficient is small, a threshold interval with a smaller interval range is required; when the variation coefficient is large and the wear coefficient is small, the upper limit of the threshold interval is slightly increased, and the lower limit is slightly decreased, and a threshold interval with a slightly increased threshold interval is required; when the variation coefficient is large and the wear coefficient is large, the upper limit of the threshold interval is significantly increased, and the lower limit is significantly decreased, and a threshold interval with a significantly increased threshold interval is required; The contact resistance at the current moment is compared with the upper and lower limits of the adjusted threshold range to complete the fault detection of the relay.

2. The relay fault detection method according to claim 1, characterized in that: After the contact resistance sequence is divided into multiple segments, for each segment, a straight line fitting is performed using the contact resistance as the ordinate and the position number thereof as the abscissa to obtain a fitting straight line for each segment.

3. The relay fault detection method according to claim 1, characterized in that: The method for obtaining the coefficient of variation of the contact resistance according to the determination coefficient of all the fitting straight lines, the slope of the fitting straight line, and the difference between the contact resistance of the last part of the fitting straight line and the minimum contact resistance is: For each segment of the fitting straight line, obtain the elements in the upper quartile, and take the average of the difference between the contact resistance of all elements greater than the upper quartile and the minimum contact resistance in each segment of the fitting straight line as the range measure of each segment of the fitting straight line; The coefficient of determination of the fitted straight line is recorded as the fitting effect of the fitted straight line; Obtain the coefficient of variation of contact resistance according to the range measurement, slope and fitting effect of all segments of the fitting straight line; The coefficient of variation of contact resistance is positively correlated with the range measure and slope of the fitting line, and negatively correlated with the fitting effect of the fitting line.

4. The relay fault detection method according to claim 3, characterized in that: The expression of the coefficient of variation is: ; The coefficient of variation of contact resistance, is the number of segments of the contact resistance series, It is The slope of the segment fitting line, It is The fitting effect of the segment fitting line, It is The range measure in the segment, is the tuning coefficient.

5. The relay fault detection method according to claim 1, characterized in that: Get the maximum value of the elements in the starting time sequence, closing time sequence and closing voltage sequence, get a sequence on the left side and the right side of the maximum value respectively, use the decomposition algorithm to get the trend sequence of all left and right sequences, and calculate the first-order difference for each trend sequence to get the starting left differential sequence, starting right differential sequence, closing left differential sequence, closing right differential sequence, voltage left differential sequence and voltage right differential sequence respectively.

6. The relay fault detection method according to claim 1, characterized in that: The method for obtaining the change trend measurement based on the proportion of positive and negative numbers in all difference sequences is: Count the proportion of positive numbers in the starting left differential sequence, closing left differential sequence, and voltage left differential sequence; Count the proportion of negative numbers in the starting right differential sequence, closing right differential sequence, and voltage right differential sequence; The expression of change trend measurement is: , where It represents the proportion of positive numbers in the first-order difference sequence of the left sequence of sequence n. It represents the proportion of negative numbers in the first-order difference sequence of the right sequence of sequence n. When n takes different values, sequence n represents the starting time sequence, closing time sequence, and closing voltage sequence respectively; M represents the change trend measure.

7. The relay fault detection method according to claim 1, characterized in that: The method for obtaining the wear coefficient of the contact point according to the proportion of positive numbers of the first-order difference sequence of the overtravel time series, the slope of the overtravel time series, the variance of the starting voltage sequence, the consistency of the closing voltage sequence and the closing time sequence, and the change trend measurement is: The correlation coefficient between the closing voltage sequence and the closing time sequence is taken as the consistency of the two sequences; The proportion of positive numbers in the first-order difference sequence of the over-range time series is recorded as the trend of the over-range time series; The wear coefficient of the contact point is positively correlated with the negative number of the slope of the overtravel time series, the change trend measure, and the consistency of the closing voltage series and the closing moment series, and is negatively correlated with the variance of the starting voltage series and the trend of the overtravel time series.

8. The relay fault detection method according to claim 1, characterized in that: The method for adjusting the range of the wear coefficient and the variation coefficient is as follows: , , where A is the coefficient of change of contact resistance, B is the wear coefficient of the contact point, is the adjusted coefficient of variation, is the adjusted wear coefficient.

9. The relay fault detection method according to claim 8, characterized in that: The threshold interval is adjusted according to the adjusted wear coefficient and variation coefficient to obtain the adjusted threshold interval: : , are the lower and upper limits of the adjusted threshold interval, , are the lower and upper limits of the original threshold interval, , are the adjusted coefficient of variation and the adjusted coefficient of wear, respectively.

10. The relay fault detection method according to claim 1, characterized in that: The method for completing relay fault detection by comparing the contact resistance at the current moment with the upper and lower limits of the adjusted threshold interval is: The area with a value less than the lower limit of the adjusted threshold interval is identified as the normal operating condition area, the area with a value greater than the upper limit of the adjusted threshold interval is identified as the serious fault area, and the area within the adjusted threshold interval is identified as the hidden fault area; If the contact resistance is in the severe fault region, there is a relay fault.

Images