A method for detecting faults in distribution automation terminals

Abstract

The present invention discloses a method for detecting faults in distribution automation terminals, and relates to the technical field of fault detection. The present invention realizes real-time monitoring and fault prediction of the status of distribution automation terminal equipment through multi-level data collection, processing, weighting, early warning and data fusion technology. The present invention constructs a membership function using triangular fuzzy numbers, quantitatively describes the normal operating range of each status indicator, and calculates the initial weight of each indicator through a normalization method, so that the degree of deviation between each indicator and the normal operating status of the equipment can be objectively reflected, thereby avoiding the imbalance of the overall status evaluation caused by the abnormality of a single indicator; at the same time, the present invention introduces a staged early warning system, integrates and calculates the abnormal indicators within the monitoring period, and dynamically corrects the initial weight only after the abnormality continues to reach a preset threshold; this staged and hierarchical adjustment mechanism effectively distinguishes short-term occasional abnormalities from continuous abnormalities, and prevents the system from misjudging the equipment status due to instantaneous fluctuations.

Description

Technical Field

[0001] The present invention relates to the technical field of fault detection, and in particular to a method for detecting faults in a distribution automation terminal. Background Art

[0002] Traditional distribution automation terminal fault detection methods rely on preset thresholds and rules to make individual judgments on various equipment indicators.

[0003] In actual applications, field data is often affected by electromagnetic interference, equipment aging, and environmental changes, causing some indicators to occasionally deviate from the normal range. Since each indicator operates independently, when a sudden change occurs in one indicator, the system will generally immediately trigger a fault alarm, causing the overall status assessment to decline significantly, while other indicators remain in good condition. This phenomenon causes the fault detection response to be overly sensitive and easily leads to false alarms or unnecessary maintenance dispatches, placing an additional burden on power grid operation and maintenance.

[0004] In this context, although some traditional solutions can smooth judgments by increasing data redundancy, adopting multi-level warnings or fuzzy logic, these methods have difficulty balancing real-time performance and accuracy. When a device experiences a slight abnormality at a certain moment, traditional methods cannot effectively distinguish whether it is a temporary fluctuation or a potential fault hazard, resulting in the overall evaluation index being greatly affected by a single abnormal data. In long-term operation, the true root cause of the fault will be concealed and the risk of misjudgment will increase. Therefore, a distribution automation terminal fault detection method is urgently needed to solve this problem. Summary of the Invention

[0005] In view of the above existing problems, the present invention is proposed.

[0006] The present invention provides a distribution automation terminal fault detection method to solve the problem that during the status evaluation process of distribution equipment, a certain fault indicator, such as no communication response or instantaneous sensor abnormality, may drop suddenly, causing the entire equipment status to slide directly from normal to serious, resulting in misjudgment or even unnecessary maintenance actions.

[0007] In order to solve the above technical problems, the present invention provides the following technical solutions:

[0008] An embodiment of the present invention provides a method for detecting faults in a distribution automation terminal, which includes:

[0009] Step S1, using a sensor device to collect status data of the power distribution terminal equipment in real time and pre-process the status data;

[0010] Step S2: Using a weight allocation method based on triangular fuzzy numbers, the state data collected and pre-processed in step S1 are calculated and weighted to obtain various state indicators; real-time monitoring is performed based on the state indicators, and a phased early warning system is established based on the monitoring feedback of the real-time monitoring;

[0011] Step S3, setting up two or more sets of redundant sensors for key status indicators, where the key status indicators are defined as indicators that play a decisive role in device communication or response;

[0012] In step S3, a distributed data aggregation algorithm is used to perform distributed aggregation and comparison on the same key indicator data from multiple sensors;

[0013] In step S4, a real-time evaluation of the equipment status is completed based on the status data pre-processed in step S1, the dynamic weight smoothing in step S2, and the redundant data aggregation results in step S3. A fault warning is triggered for equipment whose overall status evaluation is lower than the set safety threshold.

[0014] As a preferred solution of the method for detecting faults in a distribution automation terminal according to the present invention, wherein:

[0015] The status data includes communication status and sensor signals;

[0016] The preprocessing includes adaptive filtering, elimination of abnormal data and multi-source data redundancy checking.

[0017] As a preferred solution of the distribution automation terminal fault detection method described in the present invention, in which: in step S2, the step of calculating and weighting based on the state data collected and pre-processed in step S1 is:

[0018] Defining Metrics The membership function is , expressed as:

[0019] ,

[0020] in, Indicates the index after step S1 preprocessing The actual observed value of Indicator The normal working lower limit, Indicator The normal working median value, Indicator The normal working limit of

[0021] Further calculation of indicators The initial weight of is calculated as follows:

[0022] ,

[0023] in, Indicator The initial weight of Indicates the The membership value of an indicator, is the index when summing all indices, Indicates the total number of status indicators.

[0024] As a preferred solution of the distribution automation terminal fault detection method of the present invention, in step S2, the step of building a phased early warning system is as follows:

[0025] An exception indication function is introduced to monitor real-time data, which is defined as:

[0026] like ,but ,otherwise ,

[0027] in, Indicates time Time Indicator The numerical value of and Respectively represent indicators The normal working lower and upper limits, Indicates time When the indicator If it is outside the normal range, the value is 1, otherwise it is 0;

[0028] During the monitoring period Internally, calculate indicators The cumulative abnormal effect is:

[0029] ，

[0030] in, Indicates that in the cycle Internal indicators The cumulative abnormal time, Indicates the length of the early warning monitoring cycle;

[0031] Build a phased warning system. Reaching the preset threshold When , the initial weight is corrected and the corrected weight is calculated , the formula is:

[0032] ,

[0033] in, Indicates the weight after staged warning correction, represents the initial weight, represents the adjustment coefficient used to amplify the abnormal impact, Indicator The abnormal duration threshold.

[0034] As a preferred solution of the distribution automation terminal fault detection method described in the present invention, the triangular fuzzy number is determined based on a preset normal working range of the equipment.

[0035] As a preferred solution of the distribution automation terminal fault detection method of the present invention, the staged early warning system is specifically:

[0036] a) When a status indicator is detected to be abnormal and deviates from the normal range, the overall status evaluation will be slightly lowered only in the early warning period, and the duration of the abnormal indicator will be recorded;

[0037] b) When the duration of the anomaly exceeds a preset time window, the weight of the indicator is gradually increased.

[0038] As a preferred solution of the distribution automation terminal fault detection method described in the present invention, in which: in step S2, a dynamic weight smoothing function is used to perform time integration calculation on the values ​​of various abnormal indicators, and only when the cumulative abnormal effect of a certain indicator exceeds a preset threshold, the overall state degradation processing is triggered.

[0039] As a preferred solution of the distribution automation terminal fault detection method described in the present invention, in step S2, the step of using a dynamic weight smoothing function to perform time integral calculation on each abnormal index value is as follows:

[0040] Using dynamic weight smoothing function, the index The final weight is defined as :

[0041] ,

[0042] in, Indicates the final indicator after smoothing weight, represents the initial weight, represents the dynamic smoothing coefficient, Indicator The accumulated abnormal time during the monitoring period, Indicator The abnormal duration threshold, Indicates the length of the monitoring period.

[0043] As a preferred solution of the distribution automation terminal fault detection method described in the present invention, in step S3, the step of using a distributed data aggregation algorithm to distribute and compare the same key indicator data from multiple sensors is as follows:

[0044] For key status indicators , assuming that the data is Redundant sensors collect data, and each sensor data is recorded as ,in , and assign each sensor a reliability weight , using distributed data aggregation algorithm to calculate indicators Aggregate value of :

[0045] ,

[0046] in, Indicates the index after multi-sensor data aggregation Aggregate values, Indicates the indicator The number of redundant sensors, Indicates the Indicators collected by sensors data, Indicates the The reliability weight of sensor data, The index used when aggregating sensor data.

[0047] As a preferred solution of the distribution automation terminal fault detection method described in the present invention, in which: in step S4, the step of completing the real-time evaluation of the equipment status and triggering the fault warning for the equipment whose overall status evaluation is lower than the set safety threshold is as follows:

[0048] The weight after the pre-processed data in step S1 and the dynamic smoothing in step S2 And the index value after aggregation in step S3 , construct the overall equipment status evaluation function :

[0049] ,

[0050] in, Indicates the comprehensive evaluation value of the equipment status, Indicates the indicator after dynamic smoothing weight, Indicator The aggregate data value of Indicates the total number of status indicators,

[0051] When satisfied When the system triggers a fault warning, Indicates the preset minimum evaluation threshold for the device's security status.

[0052] The beneficial effects of the present invention are as follows: the present invention realizes real-time monitoring and fault prediction of the status of distribution automation terminal equipment through multi-level data collection, processing, empowerment, early warning and data fusion technology. The present invention uses triangular fuzzy numbers to construct a membership function, quantitatively describes the normal working range of each status indicator, and calculates the initial weight of each indicator through a normalization method, so that the degree of deviation between each indicator and the normal operating status of the equipment can be objectively reflected, thereby avoiding the imbalance of the overall status evaluation caused by the abnormality of a single indicator.

[0053] At the same time, the present invention introduces a phased early warning system, which integrates and calculates abnormal indicators within the monitoring period, and dynamically corrects the initial weight only after the abnormality persists to reach a preset threshold. This phased and graded adjustment mechanism effectively distinguishes short-term sporadic abnormalities from continuous abnormalities, preventing the system from misjudging the equipment status due to instantaneous fluctuations.

[0054] The present invention sets up multiple groups of redundant sensors on key status indicators and adopts a distributed data aggregation algorithm to perform weighted fusion on multi-sensor data, thereby reducing the interference of a single sensor failure on the overall judgment and further improving the robustness of the data.

[0055] The present invention forms a comprehensive evaluation function for the equipment status by performing weighted summation on the smoothed weights and aggregated data, and sets a safety threshold. When the overall evaluation value is lower than the threshold, a fault warning is immediately triggered, providing an accurate and reliable basis for operation and maintenance decisions.

[0056] In summary, the present invention effectively improves the deficiency of the traditional method that a single indicator abnormality causes overall imbalance, improves the sensitivity and stability of fault detection, and reduces the risk of false alarms and missed alarms. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0058] Figure 1 It is a flow chart of the distribution automation terminal fault detection method of the present invention. DETAILED DESCRIPTION

[0059] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0060] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0061] Secondly, the term "one embodiment" or "embodiment" herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in various places throughout this specification does not necessarily refer to the same embodiment, nor does it refer to a separate or selective embodiment that is mutually exclusive of other embodiments.

[0062] Example 1, reference Figure 1 This embodiment provides a method for detecting a fault in a distribution automation terminal, comprising the following steps:

[0063] Step S1, using a sensor device to collect status data of the power distribution terminal equipment in real time and pre-process the status data;

[0064] Status data includes communication status and sensor signals;

[0065] Preprocessing includes adaptive filtering, elimination of abnormal data, and multi-source data redundancy verification;

[0066] Step S2: Using a weight allocation method based on triangular fuzzy numbers, the state data collected and pre-processed in step S1 are calculated and weighted to obtain various state indicators; real-time monitoring is performed based on the state indicators, and a phased early warning system is established based on the monitoring feedback of the real-time monitoring;

[0067] In step S2, the steps of calculating and weighting based on the state data collected and pre-processed in step S1 are:

[0068] Defining Metrics The membership function is , expressed as:

[0069] ,

[0070] in, Indicates the index after step S1 preprocessing The actual observed value of Indicator The normal working lower limit, Indicator The normal working median value, Indicator The normal working limit of

[0071] Further calculation of indicators The initial weight of is calculated as follows:

[0072] ,

[0073] in, Indicator The initial weight of Indicates the The membership value of an indicator, is the index when summing all indices, Indicates the total number of status indicators;

[0074] Specifically, triangular fuzzy numbers are used here to define the normal interval of each state indicator, the degree of deviation of the indicator from the normal range is quantified by the membership function, and the initial weight of each indicator is calculated using the normalization method. This process can objectively reflect the degree of match between each indicator and the normal working state;

[0075] In step S2, the steps of constructing a phased early warning system are:

[0076] An exception indication function is introduced to monitor real-time data, which is defined as:

[0077] like ,but ,otherwise ,

[0078] in, Indicates time Time Indicator The numerical value of and Respectively represent indicators The normal working lower and upper limits, Indicates time When the indicator If it is outside the normal range, the value is 1, otherwise it is 0;

[0079] During the monitoring period Internally, calculate indicators The cumulative abnormal effect is:

[0080] ，

[0081] in, Indicates that in the cycle Internal indicators The cumulative abnormal time, Indicates the length of the early warning monitoring cycle;

[0082] Build a phased warning system. Reaching the preset threshold When , the initial weight is corrected and the corrected weight is calculated , the formula is:

[0083] ,

[0084] in, Indicates the weight after staged warning correction, represents the initial weight, represents the adjustment coefficient used to amplify the abnormal impact, Indicator The abnormal duration threshold;

[0085] Specifically, the abnormal indications of real-time monitoring data are integrated to obtain the cumulative abnormal time of each indicator within the monitoring period. The initial weight is then adjusted in stages according to the preset threshold. The staged warning system can distinguish between short-term abnormalities and persistent abnormalities, and only makes significant adjustments to the status after the accumulation of abnormalities reaches a certain level. This prevents occasional abnormalities from causing drastic fluctuations in the overall status and improves the accuracy of warnings.

[0086] The triangular fuzzy number is determined based on the pre-set normal operating range of the equipment;

[0087] The phased early warning system is as follows:

[0088] a) When a status indicator is detected to be abnormal and deviates from the normal range, the overall status evaluation will be slightly lowered only in the early warning period, and the duration of the abnormal indicator will be recorded;

[0089] b) when the duration of the anomaly exceeds a preset time window, gradually increasing the weight of the indicator;

[0090] In step S2, a dynamic weight smoothing function is used to perform time integral calculation on the values ​​of various abnormal indicators. Only when the cumulative abnormal effect of a certain indicator exceeds a preset threshold, the overall state is triggered to downgrade.

[0091] In step S2, the step of using the dynamic weight smoothing function to perform time integral calculation on each abnormal index value is as follows:

[0092] Using dynamic weight smoothing function, the index The final weight is defined as :

[0093] ,

[0094] in, Indicates the final indicator after smoothing weight, represents the initial weight, represents the dynamic smoothing coefficient, Indicator The accumulated abnormal time during the monitoring period, Indicator The abnormal duration threshold, Indicates the length of the monitoring period;

[0095] Specifically, a dynamic weight smoothing function is used here to normalize the cumulative effect of indicator anomalies and combine it with the smoothing coefficient to smoothly adjust the initial weight. Only when the abnormal effect exceeds the threshold will the corrected weight be gradually increased, ensuring a smooth transition of weight adjustments caused by short-term fluctuations and avoiding over-response.

[0096] Step S3, setting up two or more sets of redundant sensors for key status indicators, where the key status indicators are defined as indicators that play a decisive role in device communication or response;

[0097] In step S3, a distributed data aggregation algorithm is used to perform distributed aggregation and comparison on the same key indicator data from multiple sensors;

[0098] In step S3, the distributed data aggregation algorithm is used to perform distributed aggregation and comparison on the same key indicator data from multiple sensors.

[0099] For key status indicators , assuming that the data is Redundant sensors collect data, and each sensor data is recorded as ,in , and assign each sensor a reliability weight , using distributed data aggregation algorithm to calculate indicators Aggregate value of :

[0100] ,

[0101] in, Indicates the index after multi-sensor data aggregation Aggregate values, Indicates the indicator The number of redundant sensors, Indicates the Indicators collected by sensors data, Indicates the The reliability weight of sensor data, Index for sensor data aggregation;

[0102] Specifically, redundant sensors are set up here, and a weighted average method is used to achieve data aggregation, effectively reducing the impact of a single sensor failure on the judgment of key indicators. The data of each sensor is fused according to its reliability weight, so that the final aggregate value can truly reflect the device status and improve data robustness.

[0103] Step S4: Based on the status data pre-processed in step S1, the dynamic weight smoothing in step S2, and the redundant data aggregation results in step S3, a real-time evaluation of the device status is completed. For devices whose overall status evaluation is lower than the set safety threshold, a fault warning is triggered;

[0104] In step S4, the real-time evaluation of the equipment status is completed. For equipment whose overall status evaluation is lower than the set safety threshold, the step of triggering a fault warning is as follows:

[0105] The weight after the pre-processed data in step S1 and the dynamic smoothing in step S2 And the index value after aggregation in step S3 , construct the overall equipment status evaluation function :

[0106] ,

[0107] in, Indicates the comprehensive evaluation value of the equipment status, Indicates the indicator after dynamic smoothing weight, Indicator The aggregate data value of Indicates the total number of status indicators,

[0108] When satisfied When the system triggers a fault warning, Indicates the preset minimum evaluation threshold for the safety status of the device;

[0109] Specifically, the smoothed weight of each indicator is weighted and summed with the aggregated data value to form an overall equipment status evaluation function, which reflects the comprehensive impact of each indicator on the health status of the equipment. A safety threshold is set as the basis for judgment. When the overall status drops below the critical value, a fault warning is triggered in time, providing a more intuitive and reliable basis for real-time monitoring and maintenance decisions.

[0110] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for detecting faults in a distribution automation terminal, characterized by: include, Step S1, using a sensor device to collect status data of the power distribution terminal equipment in real time and pre-process the status data; Step S2: Using a weight allocation method based on triangular fuzzy numbers, the state data collected and pre-processed in step S1 are calculated and weighted to obtain various state indicators; real-time monitoring is performed based on the state indicators, and a phased early warning system is established based on the monitoring feedback of the real-time monitoring; The phased early warning system is specifically as follows: a) When a status indicator is detected to be abnormal and deviates from the normal range, the overall status evaluation will be slightly lowered only in the early warning period, and the duration of the abnormal indicator will be recorded; b) when the duration of the anomaly exceeds a preset time window, gradually increasing the weight of the indicator; In step S2, a dynamic weight smoothing function is used to perform time integral calculation on the values ​​of various abnormal indicators. Only when the cumulative abnormal effect of a certain indicator exceeds a preset threshold, the overall state is triggered to downgrade. In step S2, the step of using the dynamic weight smoothing function to perform time integral calculation on each abnormal index value is as follows: Using dynamic weight smoothing function, the index The final weight is defined as : ， in, Indicates the final indicator after smoothing weight, represents the initial weight, represents the dynamic smoothing coefficient, Indicator The accumulated abnormal time during the monitoring period, Indicator The abnormal duration threshold, Indicates the length of the monitoring period; In step S2, the steps of calculating and weighting based on the state data collected and pre-processed in step S1 are: Defining Metrics The membership function is , expressed as: ， in, Indicates the index after step S1 preprocessing The actual observed value of Indicator The normal working lower limit, Indicator The normal working median value, Indicator The normal working limit of Further calculation of indicators The initial weight of is calculated as follows: ， in, Indicator The initial weight of Indicates the The membership value of an indicator, is the index when summing all indices, Indicates the total number of status indicators; In step S2, the steps of constructing a phased early warning system are: An exception indication function is introduced to monitor real-time data, which is defined as: like ,but ,otherwise , in, Indicates time Time Indicator The numerical value of and Respectively represent indicators The normal working lower and upper limits, Indicates time When the indicator If it is outside the normal range, the value is 1, otherwise it is 0; During the monitoring period Internally, calculate indicators The cumulative abnormal effect is: ， in, Indicates that in the cycle Internal indicators The cumulative abnormal time, Indicates the length of the early warning monitoring cycle; Build a phased warning system. Reaching the preset threshold When , the initial weight is corrected and the corrected weight is calculated , the formula is: ， in, Indicates the weight after staged warning correction, represents the initial weight, represents the adjustment coefficient used to amplify the abnormal impact, Indicator The abnormal duration threshold; Step S3, setting up two or more sets of redundant sensors for key status indicators, where the key status indicators are defined as indicators that play a decisive role in device communication or response; In step S3, a distributed data aggregation algorithm is used to perform distributed aggregation and comparison on the same key indicator data from multiple sensors; In step S4, a real-time evaluation of the equipment status is completed based on the status data pre-processed in step S1, the dynamic weight smoothing in step S2, and the redundant data aggregation results in step S3. A fault warning is triggered for equipment whose overall status evaluation is lower than the set safety threshold.

2. A distribution automation terminal fault detection method according to claim 1, characterized in that: The status data includes communication status and sensor signals; The preprocessing includes adaptive filtering, elimination of abnormal data and multi-source data redundancy checking.

3. A distribution automation terminal fault detection method according to claim 2, characterized in that: The triangular fuzzy number is determined according to a preset normal operating range of the device.

4. A distribution automation terminal fault detection method according to claim 3, characterized in that: In step S3, the steps of using a distributed data aggregation algorithm to perform distributed aggregation and comparison on the same key indicator data from multiple sensors are as follows: For key status indicators , assuming that the data is Redundant sensors collect data, and each sensor data is recorded as ,in , and assign each sensor a reliability weight , using distributed data aggregation algorithm to calculate indicators Aggregate value of : ， in, Indicates the index after multi-sensor data aggregation Aggregate values, Indicates the indicator The number of redundant sensors, Indicates the Indicators collected by sensors data, Indicates the The reliability weight of sensor data, The index used when aggregating sensor data.

5. A distribution automation terminal fault detection method according to claim 4, characterized in that: In step S4, the step of completing the real-time evaluation of the equipment status and triggering the fault warning for the equipment whose overall status evaluation is lower than the set safety threshold is as follows: The weight after the pre-processed data in step S1 and the dynamic smoothing in step S2 And the index value after aggregation in step S3 , construct the overall equipment status evaluation function : ， in, Indicates the comprehensive evaluation value of the equipment status, Indicates the indicator after dynamic smoothing weight, Indicator The aggregate data value of Indicates the total number of status indicators, When satisfied When the system triggers a fault warning, Indicates the preset minimum evaluation threshold for the device's security status.

Images