Method and device for on-line diagnosis of faults in static converters

Abstract

The present application relates to the technical field of electric digital data processing, in particular to a kind of fault online diagnosis method and device for static converter, method includes: obtaining the temperature of multiple monitoring points in static converter at current time and its before adjacent time period of associated time period;Respectively determine the temperature variation trend of each monitoring point in two time periods, according to the temperature variation trend difference between each two monitoring points in associated time period, and the temperature variation trend difference between each monitoring point in two time periods, the heat transfer relationship of each two monitoring points is corrected, and the corrected heat transfer relationship is obtained, according to the temperature variation trend of monitoring point and the corrected heat transfer relationship of all other monitoring points in associated time period, the temperature early warning threshold of monitoring point is corrected, according to the temperature early warning threshold after correction, the accuracy and reliability of static converter fault online diagnosis can be improved, and false alarm rate is reduced.

Description

Technical Field

[0001] This invention relates to the field of electrical digital data processing technology, and specifically to a method and apparatus for online fault diagnosis of static converters. Background Technology

[0002] A static converter (SDC) is a power electronic device that uses semiconductor power devices as switching elements to convert between different forms of electrical energy (AC and DC) or change AC parameters (voltage, frequency, and phase) in a static state, without the involvement of rotating mechanical parts. With the widespread adoption of power plants, industrial manufacturing, and large-scale electrified equipment, higher demands are placed on the efficiency, safety, and stability of charging equipment. In the process of charging industrial batteries with high-performance chargers, it is often necessary to use an internal SDC to convert the grid AC power into cleaner and more stable DC power to ensure the safe operation of precision loads and batteries. Therefore, ensuring the normal and stable operation of the SDC is crucial.

[0003] During operation, the internal electronic components of a static converter are highly sensitive to temperature. Excessive heat can cause thermal expansion and volatilization of materials such as silicon chips, copper substrates, and electrolytes, easily leading to component failure and affecting the normal operation of the static converter. Current technology often uses temperature sensors deployed around the electronic components inside the static converter to monitor temperature changes in real time and assigns corresponding temperature warning thresholds to each component. When the actual monitored temperature exceeds the corresponding warning threshold, an alert is issued, enabling online fault diagnosis of the static converter. However, the temperature warning thresholds are preset fixed values. These fixed values ​​do not account for the heat transfer interactions between the internal electronic components of the static converter. Therefore, fault diagnosis based on fixed temperature warning thresholds can affect the accuracy and reliability of fault diagnosis, easily leading to misdiagnosis. Summary of the Invention

[0004] To address the technical problem that using a fixed temperature warning threshold affects the accuracy of online fault diagnosis for static converters, this invention provides a method and apparatus for online fault diagnosis of static converters.

[0005] An online fault diagnosis method for static converters includes: Obtain the associated time period of multiple monitoring points within the static converter at the current moment, as well as the temperature of adjacent time periods before the associated time period; Determine the temperature change trend of each monitoring point in the associated time period and the adjacent time period; The heat transfer relationship between each pair of monitoring points is determined based on the temperature relationship and distance between each pair of monitoring points within the associated time period. Based on the difference in temperature change trends between each pair of monitoring points during the associated time period, and the difference in temperature change trends between each monitoring point during the associated time period and adjacent time periods, the heat transfer relationship is corrected to obtain the corrected heat transfer relationship. Based on the corrected heat transfer relationship between the monitoring point and other monitoring points, and the temperature change trend of the monitoring point during the associated period, the temperature warning threshold of the monitoring point is corrected.

[0006] In one specific embodiment, the process of obtaining the temperature change trend includes: The temperature deviation between the temperature and the ambient temperature at each moment within the candidate time period is determined, and the temperature stability of the candidate time period is obtained by combining the time interval between each moment and the current moment; the candidate time period is any one of the associated time period and the adjacent time period; the temperature stability is negatively correlated with both the temperature deviation and the time interval. An adjustment coefficient is generated based on the temperature stability; the adjustment coefficient is negatively correlated with the temperature stability. The overall temperature difference of the candidate time period is determined, and the temperature change trend of the candidate time period is obtained by combining the adjustment coefficient.

[0007] In one specific embodiment, the process of obtaining the heat transfer relationship includes: The average temperature difference between the first monitoring point and the second monitoring point during the associated time period is calculated. Combined with the distance between the first and second monitoring points, the heat transfer relationship between them is obtained. The first and second monitoring points can be any two monitoring points. The heat transfer relationship is positively correlated with the average temperature difference and negatively correlated with the distance. If the value of the heat transfer relationship is positive, it indicates that the heat transfer direction between the first and second monitoring points is from the first monitoring point to the second monitoring point. If the value of the heat transfer relationship is negative, it indicates that the heat transfer direction between the first and second monitoring points is from the second monitoring point to the first monitoring point.

[0008] In one specific embodiment, the process of obtaining the corrected heat transfer relationship includes: Calculate the first temperature change trend difference between the reference monitoring point and the candidate monitoring point during the associated time period, and the second temperature change trend difference between the candidate monitoring point during the adjacent time period and the associated time period; the candidate monitoring point is any monitoring point, the reference monitoring point is any other monitoring point, and the value of the heat transfer relationship between the candidate monitoring point and the reference monitoring point is positive. The first correction coefficient is obtained based on the difference between the first temperature change trend and the difference between the second temperature change trend; The heat transfer relationship is corrected based on the first correction coefficient to obtain the corrected heat transfer relationship.

[0009] In one specific embodiment, the process of obtaining the first correction coefficient includes: When both the first temperature change trend difference and the second temperature change trend difference are positive, the first correction coefficient is obtained by multiplying the first temperature change trend difference and the second temperature change trend difference. When the first temperature change trend difference and / or the second temperature change trend difference are negative, the negative of the absolute value of the product of the first temperature change trend difference and the second temperature change trend difference is taken to obtain the first correction coefficient.

[0010] In one specific embodiment, after obtaining the corrected heat transfer relationship, the online fault diagnosis method for static converters further includes: Target monitoring points with a positive temperature change trend during the associated time period are selected, and the target monitoring points are monitoring points that receive heat. By integrating the corrected heat transfer relationships between other monitoring points and the target monitoring point, the fusion result of the corrected heat transfer relationship corresponding to the target monitoring point is determined; Accordingly, the step of correcting the temperature warning threshold of the monitoring point based on the modified heat transfer relationship between the monitoring point and other monitoring points, and the temperature change trend of the monitoring point during the associated time period, includes: Based on the fusion results of the corrected heat transfer relationship corresponding to the target monitoring point, and the temperature change trend of the target monitoring point during the associated period, the temperature warning threshold of the target monitoring point is corrected.

[0011] In one specific embodiment, the fusion result of the corrected heat transfer relationship is the average value of the corrected heat transfer relationships between other monitoring points and the target monitoring point.

[0012] In one specific embodiment, the correction of the temperature warning threshold for the target monitoring point includes: Based on the fusion result of the corrected heat transfer relationship corresponding to the target monitoring point, and the temperature change trend of the target monitoring point during the associated period, a second correction coefficient for the target monitoring point is obtained; the second correction coefficient is negatively correlated with both the fusion result and the temperature change trend. Based on the second correction coefficient, the temperature warning threshold of the target monitoring point is corrected to obtain the corrected temperature warning threshold of the target monitoring point.

[0013] In one specific embodiment, the corrected temperature warning threshold is the product of the second correction coefficient and the temperature warning threshold.

[0014] The present invention also provides an online fault diagnosis device for a static converter, comprising: a memory and a processor; the memory is connected to the processor; the memory is used to store program instructions; the processor is used to implement the above-described online fault diagnosis method for a static converter when the program instructions are executed.

[0015] This invention offers the following technical advantages: It acquires temperature data for the current time period within a related time frame, and also acquires temperature data for adjacent time periods preceding the related time frame. This provides ample data support for subsequently capturing temperature trends over time, improving the accuracy and reliability of online fault diagnosis for static converters. Based on the temperature magnitude relationship and distance between every two monitoring points within the related time frame, the heat transfer relationship between each pair of monitoring points is determined. Introducing physical spatial information effectively eliminates the bias caused by isolated analysis of the temperature of a single monitoring point. If the temperature change trends of two monitoring points differ significantly, it indicates a greater disturbance to the heat transfer relationship, necessitating correction of the heat transfer relationship. Furthermore, if the temperature change trends of the same monitoring point differ significantly between two time periods,… This indicates a drastic change in the heating state of the monitoring point. By jointly correcting the differences in these two temperature change trends, the heat transfer relationship can be transformed from a static characteristic to a dynamic characteristic that adapts to real-time operating conditions, thereby accurately removing noise interference. Finally, by combining the corrected heat transfer relationship with the temperature change trend of the monitoring point during the associated period, the temperature warning threshold of the monitoring point is corrected. This achieves differentiated and adaptive dynamic adjustment of the temperature warning threshold, so that the corrected temperature warning threshold is no longer an isolated and fixed value, but a comprehensive judgment that integrates the actual thermal impact of all surrounding heat sources and its own current temperature change trend. Therefore, based on the corrected temperature warning threshold, the accuracy and reliability of online fault diagnosis of static converters can be improved, and the false alarm rate can be reduced. Attached Figure Description

[0016] Figure 1 This is a flowchart of an online fault diagnosis method for static converters provided by the present invention. Detailed Implementation

[0017] When performing online temperature fault diagnosis on a static converter, heat transfer exists between the various electronic components inside the converter, affecting subsequent temperature changes. Therefore, relying solely on a fixed temperature warning threshold for online fault diagnosis yields low accuracy. This embodiment provides a method for online fault diagnosis of static converters. During the online fault diagnosis process, a fixed temperature warning threshold is corrected to obtain a revised temperature warning threshold for the monitoring points. Finally, accurate online fault diagnosis of the static converter is achieved based on this revised temperature warning threshold.

[0018] This embodiment sets up multiple monitoring points (specifically temperature monitoring points) within the static converter. Typically, a static converter contains various power semiconductor devices (such as power diodes, thyristors, etc.), DC bus capacitors, control chips, fan heat sinks, and connecting circuits, among other electronic components. This embodiment selects all electronic components, or selects key core components (key core components can be user-defined, such as capacitors, various semiconductor switching elements, control chips, and heat sinks). Monitoring points are set at each selected electronic component. The number of monitoring points is set according to monitoring needs, and usually a relatively large number are set. Temperature sensors are deployed at each monitoring point; high-temperature resistant, wide-range temperature sensors can be selected. After determining each monitoring point, the distance between every two monitoring points is determined; the distance is the straight-line distance in three-dimensional space within the static converter. Additionally, this embodiment deploys a temperature sensor in the environment where the static converter is located to detect the temperature of the surrounding environment. It should be noted that the temperature in this embodiment is Kelvin temperature, with the dimension corresponding to Kelvin temperature.

[0019] Each temperature sensor collects temperature data according to a preset sampling frequency, which can be set as needed for monitoring; in this embodiment, it is set to once per second. Furthermore, all temperature sensors sample synchronously, ensuring that the temperatures from all sensors can be acquired simultaneously at any given moment.

[0020] like Figure 1 As shown in the figure, the online fault diagnosis method for static converters provided in this embodiment includes the following steps: Step S100: Obtain the associated time period of multiple monitoring points in the static converter at the current moment, as well as the temperature of the adjacent time period before the associated time period.

[0021] This embodiment uses the current moment as an example to correct the temperature warning threshold of the monitoring point at the current moment. For the current moment, the associated time period is determined. The associated time period is a time period that is sequentially related to the current moment. In a specific embodiment, the current moment is used as the end time of the associated time period. The duration of the associated time period is set according to actual monitoring needs. If the duration of the associated time period is too long, moments with weak correlation to the current moment may be included; if the duration of the associated time period is too short, it may not fully include moments with some correlation to the current moment. As a specific example, based on a temperature sampling frequency of once per second, the duration of the associated time period is assumed to include 30 moments. Simultaneously, adjacent time periods before the associated time period are determined. The end time of the adjacent time period is the moment before the start time of the associated time period, and the duration of the adjacent time period is the same as the duration of the associated time period, i.e., it includes the same number of moments. ,but It equals 30.

[0022] During operation, the electronic components within a static converter inevitably experience various losses due to their non-ideal nature. These losses are converted into heat, leading to temperature increases. Subsequently, heat is transferred and diffused into the air via heat pipes and fans, allowing the temperature of each component to gradually stabilize within a certain range after reaching a certain level. However, even in a stable state, the temperature does not remain static but continues to change dynamically. Under conditions such as excessive load or insufficient heat dissipation, overheating can still occur. Therefore, this embodiment can assess temperature trends based on temperature changes monitored over a relevant time period for subsequent fault diagnosis. Furthermore, heat dissipation is a slow process. As the electronic components continue to operate, the temperature rises and heat is dissipated. Essentially, the heat dissipation behavior is a process of heat exchange between the static converter and its environment, gradually bringing their temperatures towards a dynamic equilibrium.

[0023] For any monitoring point, with the first Taking the first monitoring point as an example, determine the first... The temperature of each monitoring point during the relevant time period, and the temperature during adjacent time periods. Among them, the temperature of the [number]th monitoring point... The temperature at each monitoring point during the associated time period is essentially the [missing information]. The temperature time series of a monitoring point in a related time period, the temperature in adjacent time periods is essentially the first The temperature time series of a monitoring point in adjacent time periods, where the two temperature time series contain the same number of temperature values. The temperature time series is composed of the temperatures at each moment within the corresponding time period, sorted chronologically.

[0024] Step S200: Determine the temperature change trend of each monitoring point in the associated time period and adjacent time periods respectively.

[0025] Determine the first The temperature change trend of each monitoring point in the associated time period, as well as the temperature change trend in adjacent time periods. The temperature change trend characterizes the overall temperature change trend in the corresponding time period, including the direction of the temperature change trend (such as an increasing temperature trend, a decreasing temperature trend, and a stable temperature trend), and the magnitude of the temperature change trend (i.e., the quantitative value of the temperature change trend).

[0026] In one specific embodiment, for ease of explanation, any time period among the associated time period and adjacent time periods is defined as a candidate time period. The following gives the... A specific process for obtaining the temperature change trend of a monitoring point during a candidate time period: Due to the The number of temperatures at each monitoring point in the associated time period is the same as the number of temperatures in adjacent time periods. , then, the first The moment is the first At any time during the candidate time period, each monitoring point Indicates the first The time sequence number of each moment. The numerical range is The larger the time sequence number, the closer it is to the current time, that is, the shorter the time interval between the current time and the current time.

[0027] Determine the first The monitoring point during the candidate time period The temperature at time and the The temperature deviation of the ambient temperature at time t, specifically the absolute value of the temperature difference, is used to obtain the t-th time. The temperature deviation at each monitoring point within the candidate time period. A larger temperature deviation indicates a higher temperature deviation at the first monitoring point. The greater the difference between the temperature at each monitoring point and the ambient temperature, the more... The more unstable the temperature at each monitoring point during the candidate time period, the more... The lower the temperature stability of a monitoring point during the candidate period, the more negatively correlated the temperature stability is with the temperature deviation.

[0028] Determine the first candidate period The time interval between the i-th moment and the current moment, the longer the time interval, the more... The weaker the influence of the temperature at time t on the temperature at the current time, the less significant the effect of the temperature at time t. The weaker the influence of each monitoring point on the temperature stability during the candidate time period, the lower its reference value in the temperature stability calculation. Therefore, temperature stability is negatively correlated with the time interval. Because the first... The time sequence number of each moment The larger the value, the closer it is to the current time, meaning the shorter the time interval between it and the current time. Therefore, the first value can be used... The time sequence number of each moment To characterize the first The time interval between each time point and the current time point is positively correlated with the time interval, since the degree of temperature stability is negatively correlated with the time interval.

[0029] In one specific embodiment, based on the time sequence number of each moment within the candidate time period, the reference weight of each moment for the degree of temperature stability is obtained in the following manner: in, Indicates the first candidate within the candidate time period The reference weight of temperature stability at each moment. Indicates from 1 to The summation of the reference weights for temperature stability at each moment is achieved in this way.

[0030] Based on the logical analysis above, the following is given: A specific method for calculating the temperature stability of a monitoring point during a candidate time period: ; in, Indicates the first Temperature stability at each monitoring point during the candidate time period Represented by natural constant An exponential function with base 0. Indicates the first The monitoring point during the candidate time period Temperature deviation at any given moment This represents the preset attenuation smoothing constant, with a value ranging from 0 to 1. The specific value is determined by the user; in this embodiment, it is set to 0.1. The dimension of is the reciprocal of the dimension of temperature, that is Thus making It is a dimensionless numerical value.

[0031] No. The higher the temperature stability of each monitoring point during the candidate time period, the better the temperature stability of the first monitoring point. The higher the degree of dynamic thermal balance between the temperature at each monitoring point and the environment, the lower the temperature change trend shown by the temperature sensor, and the more it has a reverse influence on the first monitoring point. The less obvious the temperature change trend at each monitoring point during the candidate time period, the better. Therefore, according to the... The temperature stability of each monitoring point during the candidate time period is obtained. The adjustment coefficient of each monitoring point during the candidate time period, wherein the adjustment coefficient is negatively correlated with the degree of temperature stability. In a specific embodiment, the adjustment coefficient is quantified as follows: .

[0032] Determine the first The overall temperature difference of each monitoring point during the candidate time period represents the overall temperature difference of the first monitoring point. In a specific embodiment, the overall temperature change of each monitoring point during the candidate time period is determined. Temperature at the last moment of the candidate time period at each monitoring point and the Temperature at each monitoring point at the first moment of the candidate time period , No. The overall temperature difference among the monitoring points during the candidate time period is as follows: The overall temperature difference represents the first... The average temperature difference of each monitoring point during the candidate time period can reflect the first... The overall temperature change trend of each monitoring point during the candidate time period.

[0033] Calculate the first The adjustment coefficient of the first monitoring point in the candidate time period and the first monitoring point in the candidate time period The product of the overall temperature differences of the monitoring points during the candidate time period yields the result of the first monitoring point. The temperature change trend of each monitoring point during the candidate time period. Therefore, the first... The temperature change trend of each monitoring point during the candidate time period has three possibilities: positive, negative, or 0. If it is positive, it indicates that the temperature change trend of the first monitoring point during the candidate time period is positive. The temperature change trend of each monitoring point during the candidate time period is specifically an increasing temperature trend, and the larger the value, the more obvious the increasing temperature trend; if it is negative, it indicates that the temperature is decreasing. The temperature change trend of each monitoring point during the candidate time period is specifically a decreasing trend, and the larger the absolute value of the value, the more obvious the decreasing trend; if it is 0, it indicates that the temperature is decreasing. The temperature change trend of each monitoring point during the candidate time period is specifically a stable temperature trend. Using the above process, the temperature change trends of each monitoring point during the associated time period and adjacent time periods are obtained respectively.

[0034] Step S300: Determine the heat transfer relationship between each pair of monitoring points based on the temperature relationship and distance between them within the associated time period.

[0035] Temperature changes at each monitoring point are affected not only by the heat generated by the operation of electronic devices, but also by the temperature differences between different monitoring points. These temperature differences, combined with the operation of the heat dissipation system, cause heat to diffuse from the higher-heating monitoring point to the relatively lower-heating surrounding monitoring points, thus influencing the temperature changes at each monitoring point and maintaining dynamic stability. The heat transfer relationship between each pair of monitoring points is determined based on the temperature relationship and distance between them within the associated time period.

[0036] For ease of explanation, this embodiment sets the first monitoring point and the second monitoring point as any two monitoring points, where the first monitoring point is denoted as the... The first monitoring point, and the second monitoring point is denoted as the th monitoring point. One monitoring point.

[0037] Calculate the first The average temperature of each monitoring point at various times within the associated time period is used as the first... Average temperature at each monitoring point during the relevant time period Similarly, calculate the first... The average temperature of each monitoring point at various times within the associated time period is used as the first... Average temperature at each monitoring point during the relevant time period Calculate the first... Average temperature at each monitoring point during the relevant time period With the Average temperature at each monitoring point during the relevant time period Average temperature difference: . The larger the absolute value, the greater the difference in average temperature between the two monitoring points, and the more likely there is a heat transfer relationship, that is, the more significant the heat transfer relationship is. Therefore, the significance of the heat transfer relationship is positively correlated with the average temperature difference.

[0038] Determine the first The monitoring point and the first Distance between monitoring points ,distance The smaller the value, the closer the distance between the two monitoring points, and the more likely there is a heat transfer relationship between the two monitoring points. The heat transfer relationship is negatively correlated with distance.

[0039] Therefore, according to the first The monitoring point and the first The average temperature difference at each monitoring point, and the... The monitoring point and the first The distance to the monitoring point is obtained. The monitoring point and the first The heat transfer relationship at each monitoring point is quantified as follows: ; in, Indicates the first The monitoring point and the first Heat transfer relationship at each monitoring point This represents the maximum distance between all two monitoring points. To Normalization, This represents the function that takes the minimum value. express Compared to The degree of change. right Make corrections.

[0040] When it is positive, it indicates the first... Average temperature at each monitoring point during the relevant time period Higher than Average temperature at each monitoring point during the relevant time period Then the first The monitoring point and the first The heat transfer relationship values ​​at the first monitoring point are positive. The monitoring point and the first The heat transfer direction at the monitoring point is: from the first... The monitoring point to the first The monitoring point, the first The monitoring point may subsequently send to the... The heat is transferred at the monitoring point, the first The first monitoring point is for heat transfer monitoring, and the second... Each monitoring point is a monitoring point for receiving heat; When it is negative, it indicates that the first... Average temperature at each monitoring point during the relevant time period lower than the Average temperature at each monitoring point during the relevant time period Then the first The monitoring point and the first The heat transfer relationship value at the first monitoring point is negative. The monitoring point and the first The heat transfer direction at the monitoring point is: from the first... The monitoring point to the first The monitoring point, the first The monitoring point may subsequently send to the... Each monitoring point transmits heat; When it is 0, it means the first... Average temperature at each monitoring point during the relevant time period equal to the Average temperature at each monitoring point during the relevant time period Then the first The monitoring point and the first The value of the heat transfer relationship at the first monitoring point is 0, and the value of the second monitoring point is 0. The monitoring point and the first There is no direction of heat transfer at any of the monitoring points. Furthermore, The larger the absolute value of , the stronger the subsequent . The monitoring point and the first The higher the degree of heat transfer between monitoring points, the better. Using the above process, the heat transfer relationship between each pair of monitoring points in the associated time period can be obtained.

[0041] Step S400: Based on the difference in temperature change trends between each pair of monitoring points in the associated time period, and the difference in temperature change trends between each monitoring point in the associated time period and adjacent time periods, the heat transfer relationship is corrected to obtain the corrected heat transfer relationship.

[0042] Because heat transfer has a certain delay, it typically takes some time after the temperature rise in high-heat areas begins to complete. Therefore, determining the heat transfer relationship solely based on temperature changes between two monitoring points during a related time period is not accurate enough. It is necessary to further verify and correct this relationship by incorporating earlier historical temperature data, specifically temperature data from adjacent time periods.

[0043] The following analysis focuses on cases where the heat transfer relationship is positive. Specifically, it selects heat transfer relationships with positive values ​​from the heat transfer relationships between every two monitoring points over the associated time period. The analysis continues for heat transfer relationships with positive values; negative or zero values ​​are no longer analyzed.

[0044] For any heat transfer relationship with a positive value, this embodiment defines the two monitoring points corresponding to the heat transfer relationship as a candidate monitoring point and a reference monitoring point, respectively. The average temperature of the candidate monitoring point during the associated time period is higher than the average temperature of the reference monitoring point during the associated time period. The heat transfer direction between the candidate and reference monitoring points is from the candidate monitoring point to the reference monitoring point, and the candidate monitoring point may subsequently transfer heat to the reference monitoring point. Therefore, a candidate monitoring point is any monitoring point that satisfies the above conditions, and a reference monitoring point is any other monitoring point that satisfies the above conditions.

[0045] For ease of explanation, the following will be referred to as the first... The monitoring point and the first Taking a positive value as an example, the candidate monitoring point is represented as the first... The monitoring point is denoted as the th monitoring point, and the reference monitoring point is denoted as the th monitoring point. One monitoring point.

[0046] Determine the first The monitoring point and the first The temperature change trend of the monitoring point during the associated time period, and the first monitoring point The temperature change trend of each monitoring point during the relevant time period minus the first... The temperature change trends of the two monitoring points over the associated time period are analyzed, and the resulting difference is defined as the first temperature change trend difference. Therefore, the first temperature change trend difference characterizes the difference in temperature change trends between the two monitoring points in the spatial dimension. If the first temperature change trend difference is positive, it indicates that the monitoring point that received heat after heat transfer (i.e., the first...) The temperature rise at the monitoring point (i.e., the monitoring point transferring heat) is relatively greater than that at the monitoring point (i.e., the first monitoring point). The higher the temperature rise at each monitoring point, and the greater the difference in the first temperature change trend, the higher the possibility of heat transfer between them and the better the transfer effect.

[0047] Determine the first The temperature change trend of the monitoring point in adjacent time periods, and the first monitoring point... The temperature change trend of each monitoring point in adjacent time periods minus the first... The difference in temperature change trends at each monitoring point over a related time period is defined as the second temperature change trend difference. Therefore, the second temperature change trend difference characterizes the difference in temperature change trends at the same monitoring point over time. The second temperature change trend difference represents the difference in heat transfer between the monitoring points before and after heat transfer (i.e., the first...). The difference in temperature change trends at the monitoring points (number of monitoring points) is considered. If the difference in the second temperature change trend is positive, and the larger the value, the more significant the difference, indicating that the heat transfer has occurred at the second monitoring point. The greater the degree of relief in the temperature rise trend at each monitoring point, the better. The better the heat transfer effect at each monitoring point.

[0048] Based on the difference between the first and second temperature change trends, the first... The monitoring point and the first Heat transfer relationship at each monitoring point After making corrections, we obtain the first... The monitoring point and the first The corrected heat transfer relationship for each monitoring point. Specifically: First, based on the difference between the first and second temperature change trends, a first correction coefficient for the heat transfer relationship is obtained. Wherein, when both the difference between the first and second temperature change trends are positive, it indicates that the first... The monitoring point and the first The more likely there is to be a better heat transfer between monitoring points, the positive value of the first correction coefficient is used to positively correct the heat transfer relationship, that is, to increase the heat transfer relationship. The first correction coefficient is obtained by multiplying the difference between the first temperature change trend and the difference between the second temperature change trend.

[0049] When the difference in the first temperature trend and / or the difference in the second temperature trend are negative, that is, at least one of the difference in the first temperature trend and the difference in the second temperature trend is negative, the negative of the absolute value of the product of the difference in the first temperature trend and the difference in the second temperature trend is taken to obtain the first correction coefficient.

[0050] In one specific embodiment, when both the first temperature change trend difference and the second temperature change trend difference are positive, the first correction coefficient is calculated as follows: in, Indicates the first The monitoring point and the first The first correction coefficient corresponding to each monitoring point; This represents the product of the difference between the first and second temperature change trends. This represents the function that takes the absolute value. This represents the normalization function. The normalization method here is maximum-minimum normalization. The maximum and minimum values ​​involved in this method can be the theoretical upper and lower limits of the absolute values ​​of the products of the first and second temperature trend differences, as preset. Alternatively, they can be the maximum and minimum values ​​determined from the absolute values ​​of the products of all the first and second temperature trend differences, assuming all heat transfer relationship values ​​are positive. When the normalized object is greater than the maximum value, the normalization result is set to 1; when the normalized object is less than the minimum value, the normalization result is set to 0. Furthermore, in extreme cases where the maximum and minimum values ​​involved in the maximum-minimum normalization method are equal, the first correction coefficient is directly set to 1, meaning no correction is made to the heat transfer relationship.

[0051] In one specific embodiment, when the difference in the first temperature change trend and / or the difference in the second temperature change trend are negative, the first correction coefficient is calculated as follows: The calculation of the first correction coefficient ensures that when both the first and second temperature trend differences are positive, the value of the first correction coefficient ranges from 1 to 2; when the first and / or second temperature trend differences are negative, the value of the first correction coefficient ranges from 0 to 1. It should be understood that if at least one of the first and second temperature trend differences is 0, the value of the first correction coefficient is set to 1, meaning no correction is made to the heat transfer relationship.

[0052] Calculate the first The monitoring point and the first Heat transfer relationship at each monitoring point With the first correction factor The product of: The resulting product is the first The monitoring point and the first Corrected heat transfer relationship at each monitoring point This is used to correct for each heat transfer relationship with a positive value. It should be noted that if there is no heat transfer relationship with a positive value, no further data processing will be performed, and the temperature warning thresholds for each monitoring point will not be corrected.

[0053] Through the above process, the heat transfer relationships with positive values ​​are corrected. It should be understood that the corrected heat transfer relationships are also positive. Based on this, this embodiment also needs to determine the monitoring points where the temperature change trend is positive during the associated time period, in order to perform threshold correction on these monitoring points.

[0054] For any modified heat transfer relation, the modified heat transfer relation For example, determine the first One monitoring point (due to correction of heat transfer relationship) If it is a positive value, then the first... If the temperature change trend of the first monitoring point (the one receiving heat) during the associated time period is positive, then the first monitoring point will be... The first monitoring point is determined as the target monitoring point; if it is not a positive value, then the second monitoring point is determined as the target monitoring point. The monitoring point is not the target monitoring point. This process is repeated across all corrected heat transfer relationships to filter for target monitoring points. Essentially, this selection process involves determining whether each monitoring point receiving heat is a target monitoring point, based on a positive corrected heat transfer relationship. It should be noted that if no target monitoring point is found through this selection, no further data processing is performed, and the temperature warning thresholds for each monitoring point are not corrected.

[0055] For any target monitoring point, set as the first... Based on the corrected heat transfer relationships calculated above, the target monitoring points are determined to be related to the first target monitoring point. There are other monitoring points with corrected heat transfer relationships for each target monitoring point, thus obtaining the relationship between each of the other monitoring points and the first target monitoring point. The corrected heat transfer relationship for the first target monitoring point should be understood to be related to the relationship between the other monitoring points and the first target monitoring point. The corrected heat transfer relationships for all target monitoring points are positive. This is achieved by integrating the data from other monitoring points with the first... The corrected heat transfer relationship of the target monitoring point is obtained. The fusion result of the corrected heat transfer relationship corresponding to each target monitoring point. In one specific embodiment, the relationship between each of the other monitoring points and the first... The average value of the corrected heat transfer relationship at each of the target monitoring points is used as the first... The fusion results of the corrected heat transfer relationships corresponding to each target monitoring point.

[0056] Step S500: Based on the corrected heat transfer relationship between the monitoring point and other monitoring points, and the temperature change trend of the monitoring point during the associated time period, the temperature warning threshold of the monitoring point is corrected.

[0057] For the The target monitoring point, according to the first The fusion results of the corrected heat transfer relationships corresponding to the target monitoring points, and the first The temperature change trend of the target monitoring point during the associated time period, for the first The temperature warning threshold for each target monitoring point is adjusted. Specifically: based on the... The fusion results of the corrected heat transfer relationships corresponding to the target monitoring points, and the first The temperature change trend of the target monitoring points during the associated time period is obtained. The second correction coefficient for each target monitoring point. Wherein, the... The larger the value of the fusion result of the corrected heat transfer relationship corresponding to each target monitoring point, the greater the value of the first... The more heat a target monitoring point receives from other monitoring points, the lower the temperature warning threshold needs to be to avoid missed alarms. In other words, the smaller the second correction coefficient, the more negatively correlated the second correction coefficient is with the fusion result of correcting the heat transfer relationship. The larger the value of the temperature change trend of the target monitoring points over the associated time period, the more significant the temperature change trend of the first monitoring point. The more pronounced the temperature rise trend at each target monitoring point within the associated time period, the more heat it receives from other monitoring points. To avoid missed alarms, the temperature warning threshold needs to be lowered, meaning the smaller the second correction coefficient, the more negatively correlated the second correction coefficient is with the temperature change trend during the associated time period. The following gives the... One specific method for calculating the second correction coefficient for each target monitoring point: in, Indicates the first The second correction factor for each target monitoring point; Indicates the first Temperature change trends at individual monitoring points over a related time period; Indicates the first The fusion results of the corrected heat transfer relationships corresponding to each target monitoring point; This represents a preset scaling factor, ranging from 0.1 to 0.2. The specific value is set according to the scaling requirements. This embodiment uses 0.1 as an example to limit the range of the second correction factor, ensuring that the second correction factor has a lower limit, thereby avoiding the final corrected temperature warning threshold being too low. The normalization method is a maximum-minimum normalization method. The maximum and minimum values ​​involved can be artificially set theoretical upper and lower limits, or they can be obtained from historical data. Specifically, the maximum value can be calculated by the static converter under historical extreme full-load conditions. The minimum value can be calculated from the static converter under historical stable operation or standby conditions. .

[0058] According to the The second correction coefficient for each target monitoring point , for the The temperature warning threshold for each target monitoring point is corrected; specifically, the threshold for the first monitoring point is calculated. The second correction coefficient for each target monitoring point With the The product of the temperature warning thresholds of each target monitoring point is used as the first... The corrected temperature warning thresholds for each target monitoring point. In this embodiment, the pre-corrected temperature warning threshold is a pre-set fixed value. This yields the corrected temperature warning thresholds for each target monitoring point at the current time. For monitoring points other than the target monitoring points, no temperature warning threshold correction is performed on these monitoring points at the current time.

[0059] In subsequent practical applications, online fault diagnosis of the static converter can be achieved based on the temperature warning thresholds of each monitoring point determined at the current moment. The overall process is as follows: compare the temperature collected by the monitoring point with the temperature warning threshold determined at the current moment for that monitoring point. If the collected temperature is greater than the temperature warning threshold determined at the current moment, a temperature fault alarm signal for the monitoring point is output; otherwise, no temperature fault alarm signal is output. In a specific embodiment, a monitoring period can be set after the current moment. Within the monitoring period, the temperature warning thresholds of each monitoring point are the temperature warning thresholds of each monitoring point determined at the current moment (including the corrected temperature warning thresholds of each target monitoring point and the uncorrected temperature warning thresholds of other monitoring points). Temperature anomalies at each monitoring point are identified according to the temperature warning thresholds determined at the current moment. The duration of the monitoring period is set according to the actual monitoring needs, such as 10 seconds. For any monitoring point, within the monitoring period, the temperature collected by the temperature sensor at that monitoring point is compared with the temperature warning threshold determined at the current moment. If the temperature collected by the temperature sensor is greater than the temperature warning threshold determined at the current moment, a temperature fault alarm signal for that monitoring point is output to realize online fault diagnosis of the static converter; otherwise, no temperature fault alarm signal for that monitoring point is output.

[0060] This embodiment also provides an online fault diagnosis device for a static converter, including: a memory and a processor; the memory is connected to the processor, and the memory is used to store program instructions; the processor is used to implement the steps in the above-described embodiment of the online fault diagnosis method for a static converter when the program instructions are executed.

Claims

1. A method for online fault diagnosis of a static converter, characterized in that, include: Obtain the associated time period of multiple monitoring points within the static converter at the current moment, as well as the temperature of adjacent time periods before the associated time period; Determine the temperature change trend of each monitoring point in the associated time period and the adjacent time period; The heat transfer relationship between each pair of monitoring points is determined based on the temperature relationship and distance between each pair of monitoring points within the associated time period. Based on the difference in temperature change trends between any two monitoring points during the associated time period, and the difference in temperature change trends between each monitoring point during the associated time period and adjacent time periods, the heat transfer relationship is corrected to obtain the corrected heat transfer relationship. Based on the corrected heat transfer relationship between the monitoring point and other monitoring points, and the temperature change trend of the monitoring point during the associated period, the temperature warning threshold of the monitoring point is corrected. The process of obtaining the temperature change trend includes: The temperature deviation between the temperature and the ambient temperature at each moment within the candidate time period is determined, and the temperature stability of the candidate time period is obtained by combining the time interval between each moment and the current moment; the candidate time period is any one of the associated time period and the adjacent time period; the temperature stability is negatively correlated with both the temperature deviation and the time interval. An adjustment coefficient is generated based on the temperature stability; the adjustment coefficient is negatively correlated with the temperature stability. Determine the overall temperature difference of the candidate time period, and combine it with the adjustment coefficient to obtain the temperature change trend of the candidate time period; The process of obtaining the corrected heat transfer relationship includes: Calculate the first temperature change trend difference between the reference monitoring point and the candidate monitoring point during the associated time period, and the second temperature change trend difference between the candidate monitoring point during the adjacent time period and the associated time period; the candidate monitoring point is any monitoring point, the reference monitoring point is any other monitoring point, and the value of the heat transfer relationship between the candidate monitoring point and the reference monitoring point is positive. The first correction coefficient is obtained based on the difference between the first temperature change trend and the difference between the second temperature change trend; The heat transfer relationship is corrected according to the first correction coefficient to obtain the corrected heat transfer relationship; The process of obtaining the first correction coefficient includes: When both the difference between the first temperature change trend and the difference between the second temperature change trend are positive... ; When the difference in the first temperature change trend and / or the difference in the second temperature change trend are negative... ; In the formula, This represents the first correction coefficient corresponding to the i-th monitoring point and the j-th monitoring point; This represents the product of the difference between the first and second temperature change trends. This represents the function that takes the absolute value. This represents the normalization function.

2. The online fault diagnosis method for static converters according to claim 1, characterized in that, The process of obtaining the heat transfer relationship includes: The average temperature difference between the first monitoring point and the second monitoring point during the associated time period is calculated. Combined with the distance between the first and second monitoring points, the heat transfer relationship between them is obtained. The first and second monitoring points can be any two monitoring points. The heat transfer relationship is positively correlated with the average temperature difference and negatively correlated with the distance. If the value of the heat transfer relationship is positive, it indicates that the heat transfer direction between the first and second monitoring points is from the first monitoring point to the second monitoring point. If the value of the heat transfer relationship is negative, it indicates that the heat transfer direction between the first and second monitoring points is from the second monitoring point to the first monitoring point.

3. The online fault diagnosis method for static converters according to claim 1, characterized in that, After obtaining the corrected heat transfer relationship, the online fault diagnosis method for static converters further includes: Target monitoring points with a positive temperature change trend during the associated time period are selected, and the target monitoring points are monitoring points that receive heat. By integrating the corrected heat transfer relationships between other monitoring points and the target monitoring point, the fusion result of the corrected heat transfer relationship corresponding to the target monitoring point is determined; Accordingly, the step of correcting the temperature warning threshold of the monitoring point based on the modified heat transfer relationship between the monitoring point and other monitoring points, and the temperature change trend of the monitoring point during the associated time period, includes: Based on the fusion results of the corrected heat transfer relationship corresponding to the target monitoring point, and the temperature change trend of the target monitoring point during the associated period, the temperature warning threshold of the target monitoring point is corrected.

4. The online fault diagnosis method for a static converter according to claim 3, characterized in that, The fusion result of the corrected heat transfer relationship is the average value of the corrected heat transfer relationships between other monitoring points and the target monitoring point.

5. The online fault diagnosis method for a static converter according to claim 3, characterized in that, The correction of the temperature warning threshold for the target monitoring point includes: Based on the fusion result of the corrected heat transfer relationship corresponding to the target monitoring point, and the temperature change trend of the target monitoring point during the associated period, a second correction coefficient for the target monitoring point is obtained; the second correction coefficient is negatively correlated with both the fusion result and the temperature change trend. Based on the second correction coefficient, the temperature warning threshold of the target monitoring point is corrected to obtain the corrected temperature warning threshold of the target monitoring point.

6. The online fault diagnosis method for a static converter according to claim 5, characterized in that, The corrected temperature warning threshold is the product of the second correction coefficient and the temperature warning threshold.

7. An online fault diagnosis device for a static converter, characterized in that, include: Memory and processor; The memory is connected to the processor; The memory is used to store program instructions; The processor is configured to implement, when program instructions are executed, the online fault diagnosis method for a static converter as described in any one of claims 1-6.

Images