Abnormality diagnosis method and system suitable for electric power acquisition equipment
By using multi-dimensional data and gray correlation method, the abnormal diagnosis method of power acquisition equipment can accurately identify the cause of the fault, solving the problem that the existing technology cannot identify specific fault types, and improving the fault handling efficiency and operation and maintenance efficiency.
Patent Information
- Application Number
- CN202410277961.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-03-12
- Publication Date
- 2025-03-14
AI Technical Summary
The existing abnormality monitoring technology of power collection equipment cannot identify the specific fault type and cannot provide operation and maintenance personnel with necessary fault diagnosis auxiliary functions, resulting in high pressure and low efficiency in troubleshooting.
Multi-dimensional data and the failure list of the same day are used for abnormal diagnosis, and the specific cause of the power acquisition equipment is determined by calculating the impact vector set and gray correlation method.
It realizes accurate positioning and real-time perception of power collection equipment failures, reduces the need for on-site inspection by operation and maintenance personnel, and improves fault identification and operation and maintenance efficiency.
Smart Images
Figure CN119622539A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of electric power equipment, and in particular to an abnormality diagnosis method and system applicable to electric power collection equipment. Background Art
[0002] At present, the number of metering and collection equipment (terminals, meters) connected on-site at various outlets and provincial power companies is increasing day by day. The operating environment is complex, the fault types are many, the collection equipment is abnormal, and users steal electricity, resulting in increasing pressure on collection fault handling and low work efficiency, which not only affects the fairness of users' electricity use, but also restricts the cost recovery and long-term development of the power grid. There are many reasons for the abnormality of power collection equipment. The existing abnormal monitoring of power collection equipment collects multi-dimensional data of power collection equipment and the daily failure list through the power grid system for abnormal diagnosis. However, the analysis of the data is limited to whether the power collection equipment is offline / online or whether data is collected, so as to judge whether the power collection equipment is damaged. This abnormal monitoring and analysis technology cannot identify the specific fault type of the power collection equipment, and cannot provide the necessary fault diagnosis auxiliary functions for the operation and maintenance personnel. The operation and maintenance personnel need to be dispatched to the site to troubleshoot the fault. It can no longer meet the data analysis application requirements such as power collection metering device status detection, efficient operation and maintenance of the collection system, and equipment quality evaluation. Summary of the invention
[0003] Based on this, it is necessary to provide an abnormal diagnosis method and system suitable for power collection equipment to address the problem that the existing abnormal monitoring and analysis technology cannot identify the specific fault type (i.e., the cause of the abnormality) of the power collection equipment, which expands and innovates the operation and maintenance model and builds a new customer-oriented intelligent collection and metering operation and maintenance guarantee system.
[0004] An abnormality diagnosis method applicable to power collection equipment uses collected multi-dimensional data and a daily failure list to perform abnormality diagnosis. The abnormality diagnosis method includes an equipment fault diagnosis method, and the equipment fault diagnosis method includes the following steps:
[0005] S11, calculating multiple impact vector sets according to the multi-dimensional data and the daily failure list, each of the impact vector sets corresponding to an abnormal cause;
[0006] S12, performing dimensionless processing on the failure list of the day to obtain a vector set of a parent sequence;
[0007] S13, all the influence vector sets are calculated one by one by the grey correlation method to obtain the correlation coefficient between the subsequence in each influence vector set and the parent sequence; wherein the correlation coefficient formula is:
[0008]
[0009] In the formula, i is the number of the subsequence in the influence vector set, j is the number of the dimension; ρ is the resolution coefficient, ρ∈(0,1); Y t is the vector set of the parent sequence, expressed as Y t =[Y t (1), Y t (2) Y t (3), ..., Y t (n)]; X ri is the influence vector set, expressed as X ri =[X r1 (1) X r 2(2),X r 3(3), ..., X r n (n)];
[0010] The grey correlation coefficients of the subsequences in the same influence vector set are averaged to obtain the grey correlation coefficients; the grey correlation coefficients of all influence vector sets are calculated in sequence;
[0011] S14. Determine the abnormal cause corresponding to the influence vector set with the largest grey correlation as the abnormal cause of the power collection device collection failure.
[0012] As a preferred example, the plurality of influence vector sets are respectively:
[0013] The influence vector set corresponding to the abnormal cause is the temperature environment change;
[0014] The impact vector set corresponding to the abnormal causes is clock display abnormality and clock chip damage;
[0015] The influence vector set corresponding to the abnormal cause of excessive current;
[0016] The influence vector set corresponding to the abnormal causes is voltage fluctuation and internal metering circuit failure.
[0017] As a preferred example, the multi-dimensional data is first cleaned and solidified before use; the multi-dimensional data includes environmental data, verification and detection data, automatic data collected by power collection equipment, operation and maintenance closed-loop work orders, equipment fault record data and historical processing data.
[0018] As a preferred example, one of the impact vector sets is calculated by a temperature model, the input information of the temperature model includes the temperature inside the power collection equipment, the temperature of the area where it is located, and the seasonal temperature, and the output information includes the user number, data time, the temperature inside the meter, and whether the temperature is abnormal. The temperature model calculates the impact vector set of the temperature environment change on the equipment abnormality through the output information.
[0019] As a preferred example, one of the impact vector sets is calculated by a clock model, and the operating steps of the clock model are: the first step is to obtain the clock anomaly of the online monitoring of the electric meter by default. If there is no result, the line loss metering data of the electric meter is called; the second step is to calculate the reducible power of the electric meter based on the line loss metering data; if the reducible power is greater than the set value, the corresponding abnormal result is output according to the numerical value greater than the set value, and the clock model calculates the impact vector set of the clock display abnormality and clock chip damage on the equipment abnormality according to the abnormal result.
[0020] As a preferred embodiment, one of the influence vector sets is calculated by a current overcurrent model, and the current overcurrent model is used to determine whether the current is too large. The method for determining whether the current is too large includes the following steps:
[0021] S31. Obtain the three-phase current data of the power collection device on that day, and count the number of times that the current of any phase is greater than 1*range;
[0022] S32. If the number of times that the phase current is greater than 1*range / 96>0.7 on that day, it is determined that the current is too large.
[0023] As a preferred example, one of the influence vector sets is calculated by a metering inaccuracy model, and the input information of the metering inaccuracy model is the abnormal increase or decrease information of the power collection device, and the information that the power collection device stops when there is voltage but no power load.
[0024] As a preferred example, the abnormal diagnosis method is visualized, and the visualization includes establishing an operation and maintenance knowledge base and a fault operation and maintenance knowledge graph, which is used to perform multi-dimensional statistics and presentation of multi-dimensional data, cause analysis results and processing feedback information.
[0025] As a preferred example, the abnormality diagnosis method further includes a data acquisition failure diagnosis method, and the data acquisition failure diagnosis method includes the following steps:
[0026] S21, obtaining an electric meter to be diagnosed according to the failure list of the day;
[0027] S22: Diagnose whether the concentrator is online through the power grid system. If the concentrator is online, enter S23; if the concentrator is not online, check whether there is a power outage in the local area. If there is no power outage in the local area but the concentrator is not online, or the concentrator is still not online after the power outage, it is necessary to dispatch operation and maintenance personnel to conduct on-site inspection;
[0028] S23, remotely synchronize the clocks of the electric meter and the concentrator through the power grid system; if the remote synchronization of the clock is successful, the first data supplement is performed; if the first data supplement is successful, it is determined that the clock is abnormal; if the first data supplement fails or the remote synchronization of the clock fails, enter S24;
[0029] S24, resetting the meter reading parameters of the electric meter and the concentrator through the power grid system; if the resetting is successful, performing a second data supplement; if the second data supplement is successful, determining that the meter reading parameters are abnormal; if the second data supplement fails or the resetting fails, entering S25;
[0030] S25. Perform a push process on the in-transit process through the power grid system; if the push process is normal, perform a third data supplement; if the third data supplement is successful, determine that the in-transit process is abnormal; if the third data supplement fails or the push process fails, enter the subsequent process.
[0031] The present invention also provides an abnormality diagnosis system applicable to a power collection device, which uses the abnormality diagnosis method for the power collection device as described above, and the abnormality diagnosis system comprises:
[0032] Data collection module, which is used to collect relevant multi-dimensional data from various power collection systems and various sources to support grid-based business monitoring and operation and maintenance;
[0033] A storage and statistics module is used to store the multi-dimensional data, clean and solidify the data, and compile statistics on the failure list of the day;
[0034] An intelligent diagnosis module, which is used to perform a data acquisition failure diagnosis method and an equipment fault diagnosis method;
[0035] The application layer module is used to visualize the abnormal diagnosis method.
[0036] The beneficial effects of the present invention are:
[0037] 1. The present invention uses intelligent diagnosis and monitoring to send parameters and perform clock synchronization on the electric meter and concentrator every morning to determine whether there is a communication or software failure. Send multiple times before 8 o'clock, and then focus on abnormalities such as no data on the electric meter under the concentrator and failure to collect data for three consecutive days, and combine the equipment fault diagnosis steps to determine whether it is a device or hardware failure. After determining the cause of the fault, the intelligent diagnosis work order is given to the local company, and the on-site operation and maintenance personnel will repair and verify the fault information. The present invention can accurately locate the cause of the fault, realize real-time perception of the fault, and achieve unmanned monitoring of fault anomalies without the need for operation and maintenance personnel to go to the site for troubleshooting, thereby significantly improving the efficiency of fault identification and operation and maintenance.
[0038] 2. While accurately identifying the cause of the fault, the present invention can provide targeted solutions. Based on operation and maintenance knowledge, environmental data, verification and detection data, collection data, procurement and marketing system archive (parameter) information, historical fault information, etc., it can solve simple faults online and perform qualitative analysis of the fault, thereby improving the fault handling responsiveness of the power collection equipment, improving the efficiency of operation and maintenance implementation, and ensuring the safety and standardization of the fault handling process. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] Figure 1 A flow chart of a method for diagnosing abnormalities in a power collection device;
[0040] Figure 2 A flowchart of the equipment troubleshooting steps. DETAILED DESCRIPTION
[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0042] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which the present invention belongs. The terms used herein in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. The term "or / and" used herein includes any and all combinations of one or more of the related listed items.
[0043] Due to interference from various factors, the meter often fails to collect data. The reasons for collection failure and abnormality are mainly caused by communication failure, meter failure, metering module failure, weather and other factors. The above main reasons can be attributed to two categories, one is communication and software failure; the other is equipment or hardware failure. The present invention conducts comprehensive detection and diagnosis for these two reasons.
[0044] See also Figure 1, this embodiment proposes an abnormality diagnosis method suitable for power collection equipment, which uses the collected multi-dimensional data for abnormality diagnosis. Multi-dimensional data includes environmental data, verification and detection data, automatic data collection of power collection equipment, operation and maintenance closed-loop work orders, equipment fault record data and historical processing data, etc., which are related data of power collection equipment that can be collected. Extract the data that has not been collected or shows collection failure, and collect common and regular abnormal problems, count and form a failure list for the day. The above-mentioned abnormality diagnosis includes a data collection failure diagnosis method and an equipment failure diagnosis method. The data collection failure diagnosis method includes the following steps:
[0045] S21. Obtain the meter to be diagnosed based on the daily failure list. The power grid system will call for meter data at dawn every day to collect meter reading information for the day, along with a large amount of collection failure and abnormal information. Based on the meter collection principle, we determine the target meter to be diagnosed by querying the daily failure list.
[0046] S22. Diagnose whether the concentrator is online through the power grid system. As the core device for collecting meter data, the concentrator can receive and transmit relevant data to the power grid system. If the concentrator is online, enter S23. If the concentrator is not online, check whether there is a power outage in the local area. If there is no power outage in the local area but the concentrator is not online, it is necessary to dispatch operation and maintenance personnel to conduct on-site investigation. If there is a local power outage, follow-up steps are carried out after power is supplied to ensure that the local area performs the abnormal diagnosis method when there is power. If the concentrator is still not online after power is supplied, it is also necessary to dispatch operation and maintenance personnel to conduct on-site investigation.
[0047] S23. Remotely synchronize the clocks of the meter and the concentrator through the power grid system. If the remote synchronization of the clock is successful, the first data supplement is performed. If the first data supplement is successful, the cause of the abnormality is determined, that is, the abnormality is caused by the abnormality of the clocks of the meter and the concentrator. If both the first data supplement and the remote synchronization fail, enter S24. At this time, it is unclear whether the collection abnormality is caused by the clock or other reasons, and the cause can only be further investigated by the subsequent process.
[0048] S24, reset the meter reading parameters of the electric meter and the concentrator through the power grid system. If the reset is successful, perform a second data supplement. If the second data supplement is successful, determine the cause of the abnormality, that is, the abnormality is that the meter reading parameters of the electric meter and the concentrator are abnormal. Similarly, if the second data supplement and reset both fail, and it is unclear what the cause is, enter S25.
[0049] S25. Perform a push process on the in-transit processes of the electric meter and the concentrator through the power grid system. If the push process is normal, perform a third data supplement. If the third data supplement is successful, determine the cause of the abnormality, that is, the cause of the abnormality is that the in-transit processes of the electric meter and the concentrator are abnormal. At this time, the data acquisition failure diagnosis method has been completed, and the required abnormal cause can be obtained. In another embodiment, if both the third data supplement and the push process fail, the subsequent process can be continued, and the subsequent process includes checking the items specified by the next business rule. According to the above steps, the items are continuously checked and the data is called until the abnormal cause is found or all subsequent processes fail. The above-mentioned data acquisition failure diagnosis method checks the common and most fault types on the electric meter by whether the remote data is collected or not, and is mainly used to determine the faults in the communication or parameters of the electric meter, that is, the software faults.
[0050] After the data collection failure diagnosis method, the steps of the equipment fault diagnosis method are performed, or the steps of the equipment fault diagnosis method are performed separately. If the equipment fault diagnosis method does not diagnose the cause of the abnormality, it is also necessary to dispatch operation and maintenance personnel to conduct on-site investigation. Equipment failures have the characteristics of diverse and repetitive fault types. This embodiment takes the equipment failure with high repetitiveness, that is, the mainstream equipment failure cause, as an example to perform automatic equipment fault diagnosis. Please refer to Figure 2 , the equipment fault diagnosis method includes the following steps:
[0051] S11. Input the multi-dimensional data and the daily failure list into the equipment failure prediction model to calculate and obtain multiple impact vector sets. Each impact vector set corresponds to an abnormal cause. In this embodiment, the multiple impact vector sets are:
[0052] The corresponding abnormal cause is the impact vector set of temperature environment changes.
[0053] The corresponding abnormal causes are the impact vector set of clock display abnormality and clock chip damage.
[0054] The influence vector set corresponding to the abnormal cause of excessive current.
[0055] The influence vector set corresponding to the abnormal causes is voltage fluctuation and internal metering circuit failure.
[0056] The multi-dimensional data is cleaned and solidified before use. Data cleaning is used to remove some irrelevant information and retain only data related to equipment failure for subsequent analysis and processing. Data solidification is used to avoid the overwriting of manually defined data during daily updates and improve the reliability of abnormal analysis. The equipment failure prediction models obtained by combining previous information and data include: temperature model, clock model, current overload model and metering inaccuracy model.
[0057] Temperature model, which can use gray prediction model, BP neural network, Prophet model and other models to predict the impact of ambient temperature on the meter. Changes in temperature environment will affect the characteristics of the internal components of the meter, thereby affecting the meter readings. Generally speaking, an increase in temperature will increase the meter reading, while a decrease in temperature will reduce the meter reading. The increase in temperature causes the danger of the meter getting hot. Long-term overheating will cause insulation degradation and short circuit, burn the meter, and cause fire; obtain temperature change data in each region, study the impact of seasonal temperature changes on the increase of various abnormal quantities of the meter, analyze the health of the meter, and form a temperature model. The input information of the temperature model includes meter equipment identification, three-phase current, temperature inside the meter, regional temperature and seasonal temperature, etc., and the output information includes user number, data time, temperature inside the meter, whether the temperature is abnormal, etc. The temperature model calculates the impact vector set of temperature environment changes on equipment abnormalities through output information.
[0058] Based on past experience and work order information, the clock model is constructed and trained using models such as the grey prediction model, BP neural network, and autoregressive moving average model. When the meter clock is normal, data can be collected normally. However, due to battery undervoltage, overvoltage, damage to the clock module itself, etc., data loss or collection failure may occur. The above reasons are digitized and used as training data for the clock model. The operating logic of the clock model is as follows: Step 1, the clock of the meter is diagnosed by sending verification data through the power grid system. If there is an abnormality, the clock model obtains the line loss measurement data of the meter. Step 2, based on the line loss measurement data, the clock model calculates the amount of electricity that can be reduced. If the amount of electricity that can be reduced is greater than the set value, and the value of the excess part is L, the clock model outputs the corresponding abnormal value according to the value L. The clock model calculates the influence vector set of clock display abnormality and clock chip damage on equipment abnormality based on the abnormal value.
[0059] Based on past experience and work order information, the current overcurrent model is constructed and trained using models such as the gray prediction model, BP neural network, and autoregressive moving average model. When the meter is collecting data normally, if the current is too large, an abnormal current overcurrent will be generated. Excessive load current will reduce the life of the meter, and will also affect the accuracy and stability of the meter, causing damage to the meter. The method for judging excessive current includes the following steps:
[0060] S31. Obtain the three-phase current data of the power collection equipment on that day, and count the number of times the current of any phase is greater than 1*range.
[0061] S32. If the number of times that the phase current is greater than 1*range / 96>0.7 on that day, it is determined that the current is too large.
[0062] The current overcurrent model is used to calculate the impact vector set of the overcurrent on the device abnormality according to the number of times the current is too large.
[0063] Based on past experience and work order information, the metering inaccuracy model is constructed and trained using models such as the grey prediction model, BP neural network, and autoregressive moving average model. When the meter metering power increases or decreases abnormally, the meter stops when there is voltage but no power load, etc., the load curve characteristics are analyzed. The metering inaccuracy model is used to calculate the influence vector set of voltage fluctuations, internal metering circuit failures, burnout, electromagnetic interference, or meter theft on equipment abnormalities based on the degree of deviation of the abnormality (i.e., the peaks and troughs of the load curve).
[0064] S12. Dimensionless processing is performed on the failure list of the day to obtain a vector set of the parent sequence. Similarly, dimensionless processing is performed on the impact vector set to reduce the difference in absolute values of the data.
[0065] S13, all the influence vector sets in S11 are calculated one by one by the grey correlation method, and the correlation coefficient between the subsequence and the parent sequence in each influence vector set is obtained, that is, the weight of the collection failure corresponding to a certain type of abnormal cause is calculated, so as to determine the diagnosis priority. Among them, the correlation coefficient formula is:
[0066]
[0067] Where i is the number of the subsequence in the influence vector set, and the value range of i is 1, 2, 3, ..., n; j is the number of the dimension, and the value range of j is 1, 2, 3, ..., n; ρ is the resolution coefficient, ρ∈(0,1), generally 0.5; Y t is the vector set of the parent sequence, expressed as Y t =[Y t (1) Y t (2) Y t (3), ..., Y t (n)]; X ri is the influence vector set, expressed as X ri =[X r1 (1) X r 2(2),X r3 (3), ..., X r n (n)]. The grey correlation coefficients of the subsequences in the same influence vector set are averaged to obtain the grey correlation degree. The grey correlation degrees of all influence vector sets are calculated in turn.
[0068] S14. All of the above-mentioned influencing factors may be the direct cause of the equipment failure. In order to accurately determine the priority of the abnormal causes that lead to the equipment failure, multiple gray correlations are arranged from large to small, and the abnormal cause corresponding to the influence vector set with the largest gray correlation is determined as the abnormal cause of the power collection equipment collection failure.
[0069] The abnormal diagnosis method in this embodiment can be visualized, and multi-dimensional statistics and presentations are performed on multi-dimensional data, cause analysis results, equipment failure prediction models, and processing feedback information and other process and result data. Its visualization processing includes the use of big data and artificial intelligence technology to simulate the on-site operation and maintenance fault knowledge map and operation and maintenance knowledge base formed by R&D experts based on equipment fault analysis experience, automatically locate on-site problems, feedback solutions, reduce the skill reserve requirements of on-site operation and maintenance personnel, provide fault solution references and learning for operation and maintenance personnel, and provide guidance on the operation process of on-site fault handling.
[0070] In another embodiment, a power harvesting device abnormality diagnosis system is provided, which uses the power harvesting device abnormality diagnosis method described above. The abnormality diagnosis system includes:
[0071] The data collection module is used to collect relevant multi-dimensional data from each power collection system and various sources (including power consumption information collection system, marketing business system, dispatching platform and GIS system) to support grid business monitoring and operation and maintenance.
[0072] The storage and statistics module reads and writes information through ORACLE, stores data in the distributed data warehouse and performs parallel operations on the big data platform to meet the needs of intelligent diagnosis and analysis. The storage and statistics module is used to store multi-dimensional data, clean and solidify data, and is also used to count the failure list of the day.
[0073] The intelligent diagnosis module uses stream processing programs, artificial intelligence algorithms, etc. to integrate and analyze different fault diagnosis rules, fault handling rules, and historical fault handling in real time, and executes data collection failure diagnosis steps and equipment fault diagnosis steps.
[0074] The application layer module pushes strategies based on the diagnosis results of the abnormal diagnosis method to achieve functions such as indicator control, equipment operation monitoring, visualization processing, and operation and maintenance management.
[0075] In summary, the failure of information collection of power collection equipment may be caused by factors such as parameters, clocks, temporary offline or physical failure of equipment. Therefore, the abnormal diagnosis method in this embodiment can send parameters and perform clock synchronization on the electric meter and concentrator every morning to determine whether it is a communication or software failure. Send it multiple times before 8 o'clock, and then focus on abnormalities such as no data from the electric meter under the concentrator and failure to collect data for three consecutive days, and combine the equipment fault diagnosis steps to determine whether it is a device or hardware failure. After determining the cause of the fault, the intelligent diagnosis work order is given to the municipal company, and the on-site operation and maintenance personnel repair and verify the fault information, and then feedback to the superior to eliminate the fault. If it is a new fault or a fault that is not diagnosed by intelligent diagnosis, it will be retained and summarized.
[0076] The present invention proposes an abnormality diagnosis method and system suitable for power collection equipment. Based on big data and artificial intelligence technology, it cleans, analyzes and diagnoses various fault anomalies of power collection equipment, shortens the fault discovery time, strengthens the inspection of equipment such as electric meters, concentrators, and collectors, and realizes qualitative analysis of faults. Accurately identify the cause of the fault, and provide targeted solutions at the same time, improve the responsiveness of the collection equipment fault handling, improve the efficiency of operation and maintenance implementation, and ensure the safety and standardization of the fault handling process. At the same time, it can provide strong data support and decision-making support for the life cycle management and precise rotation of power collection equipment. It is of great significance to the research on collection and metering operation and maintenance and fault analysis and processing.
[0077] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.
[0078] The above-mentioned embodiments only express several implementation methods of the present invention, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, and these all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.
Claims
1. An abnormality diagnosis method applicable to power collection equipment, which uses the collected multi-dimensional data and the daily failure list to perform abnormality diagnosis, characterized in that: The abnormality diagnosis method includes an equipment fault diagnosis method, and the equipment fault diagnosis method includes the following steps: S11, calculating multiple impact vector sets according to the multi-dimensional data and the daily failure list, each of the impact vector sets corresponding to an abnormal cause; S12, performing dimensionless processing on the failure list of the day to obtain a vector set of a parent sequence; S13, all the influence vector sets are calculated one by one by the grey correlation method to obtain the correlation coefficient between the subsequence in each influence vector set and the parent sequence; wherein the correlation coefficient formula is: Where i is the number of the subsequence in the influence vector set, and the value range of i is 1, 2, 3, ..., n; j is the number of the dimension, and the value range of j is 1, 2, 3, ..., n; ρ is the resolution coefficient, ρ∈(0,1); Y t is the vector set of the parent sequence, expressed as Y t =[Y t (1) Y t (2) Y t (3), ..., Y t (n)]; X ri is the influence vector set, expressed as X ri =[X r1 (1) X r2 (2) X r3 (3), ..., X rn (n)]; The grey correlation coefficients of the subsequences in the same influence vector set are averaged to obtain the grey correlation coefficients; the grey correlation coefficients of all influence vector sets are calculated in sequence; S14. Determine the abnormal cause corresponding to the influence vector set with the largest grey correlation as the abnormal cause of the power collection device collection failure.
2. The abnormality diagnosis method applicable to power collection equipment according to claim 1, characterized in that: The multiple influence vector sets are respectively: The influence vector set corresponding to the abnormal cause is the temperature environment change; The impact vector set corresponding to the abnormal causes is clock display abnormality and clock chip damage; The influence vector set corresponding to the abnormal cause of excessive current; The influence vector set corresponding to the abnormal causes is voltage fluctuation and internal metering circuit failure.
3. The abnormality diagnosis method applicable to power collection equipment according to claim 1, characterized in that: The multi-dimensional data is first cleaned and solidified before use; the multi-dimensional data includes environmental data, verification and detection data, automatic data collected by power collection equipment, operation and maintenance closed-loop work orders, equipment fault record data and historical processing data.
4. The abnormality diagnosis method applicable to power collection equipment according to claim 1, characterized in that: One of the impact vector sets is calculated by a temperature model, the input information of the temperature model includes the temperature inside the power collection equipment, the temperature of the area where it is located, and the seasonal temperature, and the output information includes the user number, data time, the temperature inside the meter, and whether the temperature is abnormal. The temperature model calculates the impact vector set of temperature environment changes on equipment abnormalities through the output information.
5. The abnormality diagnosis method applicable to power collection equipment according to claim 1, characterized in that: One of the impact vector sets is calculated by a clock model, and the operation steps of the clock model are as follows: the first step is to obtain the clock anomaly of the online monitoring of the electric meter by default, and if there is no result, call the line loss measurement data of the electric meter; The second step is to calculate the amount of electricity that can be reduced based on the line loss measurement data; if the amount of electricity that can be reduced is greater than the set value, output the corresponding abnormal result according to the numerical value greater than the set value, and the clock model calculates the impact vector set of clock display abnormality and clock chip damage on equipment abnormality based on the abnormal result.
6. The abnormality diagnosis method applicable to power collection equipment according to claim 1, characterized in that: One of the influence vector sets is calculated by a current overcurrent model, and the current overcurrent model is used to determine whether the current is too large. The method for determining whether the current is too large includes the following steps: S31. Obtain the three-phase current data of the power collection device on that day, and count the number of times that the current of any phase is greater than 1*range; S32. If the number of times that the phase current is greater than 1*range / 96>0.7 on that day, it is determined that the current is too large.
7. The abnormality diagnosis method applicable to power collection equipment according to claim 1, characterized in that: One of the influence vector sets is calculated by a metering inaccuracy model, and the input information of the metering inaccuracy model is abnormal increase or decrease information of the power collection device, and information that the power collection device stops when there is voltage but no power load.
8. The abnormality diagnosis method applicable to power collection equipment according to claim 1, characterized in that: The abnormal diagnosis method is visualized, and the visualization processing includes establishing an operation and maintenance knowledge base and establishing a fault operation and maintenance knowledge graph, which is used to perform multi-dimensional statistics and presentation of multi-dimensional data, cause analysis results and processing feedback information.
9. The abnormality diagnosis method applicable to power collection equipment according to claim 1, characterized in that: The abnormality diagnosis method further includes a data acquisition failure diagnosis method, and the data acquisition failure diagnosis method includes the following steps: S21, obtaining an electric meter to be diagnosed according to the failure list of the day; S22: Diagnose whether the concentrator is online through the power grid system. If the concentrator is online, enter S23; if the concentrator is not online, check whether there is a power outage in the local area. If there is no power outage in the local area but the concentrator is not online, or the concentrator is still not online after the power outage, it is necessary to dispatch operation and maintenance personnel to conduct on-site inspection; S23, remotely synchronize the clocks of the electric meter and the concentrator through the power grid system; if the remote synchronization of the clock is successful, the first data supplement is performed; if the first data supplement is successful, it is determined that the clock is abnormal; if the first data supplement fails or the remote synchronization of the clock fails, enter S24; S24, resetting the meter reading parameters of the electric meter and the concentrator through the power grid system; if the resetting is successful, performing a second data supplement; if the second data supplement is successful, determining that the meter reading parameters are abnormal; if the second data supplement fails or the resetting fails, entering S25; S25. Perform a push process on the in-transit process through the power grid system; if the push process is normal, perform a third data supplement; if the third data supplement is successful, determine that the in-transit process is abnormal; if the third data supplement fails or the push process fails, enter the subsequent process.
10. An abnormality diagnosis system for power collection equipment, characterized in that: The abnormality diagnosis method for power collection equipment according to any one of claims 1 to 9 is used, and the abnormality diagnosis system comprises: Data collection module, which is used to collect relevant multi-dimensional data from various power collection systems and related sources to support grid-based business monitoring and operation and maintenance; A storage and statistics module is used to store the multi-dimensional data, clean and solidify the data, and compile statistics on the failure list of the day; An intelligent diagnosis module, which is used to execute the steps of the data acquisition failure diagnosis method and the equipment fault diagnosis method; The application layer module is used to visualize the abnormal diagnosis method.