Cable equipment positioning system and method based on multi-source data fusion

Abstract

The invention relates to a cable equipment positioning system and method based on multi-source data fusion. The system comprises a reference state mining module, a non-consistent state identification module, a detector matching module, a signal confidence retrieval module and a fault positioning output module. The reference state mining module is used for acquiring the normal operation reference state of the cable in the target area; the non-consistent state identification module is used for identifying an abnormal area deviating from a reference state; the detector matching module is used for selecting an adaptive detector array according to the abnormal region features; the signal confidence retrieval module is used for establishing a confidence interval of the detection signal; and the fault positioning output module is used for accurately positioning a fault cable equipment distribution area and outputting a result to a user side. Through multi-source data fusion, the quasi-fault area is identified in advance, the detector array is deployed in a targeted manner, and the technical effects of improving the accuracy, timeliness and efficiency of cable equipment fault early warning and positioning are achieved.

Description

Technical Field

[0001] The present invention relates to the field of equipment fault detection, and in particular to a cable equipment positioning system and method based on multi-source data fusion. Background Art

[0002] Cable equipment is an important part of the power and communication systems, and its stable operation plays a vital role in ensuring energy supply and information transmission. With the acceleration of urbanization and the continuous advancement of underground pipeline construction, the number of buried cable equipment is increasing, its spatial distribution is more complex, and the working environment is more variable. In this case, fault detection and precise positioning of cable equipment have become key links in operation and maintenance.

[0003] At present, the fault location of cable equipment mainly adopts the post-detection method, that is, after the fault occurs, the fault point is determined by manual inspection, traditional cable fault locator and other means. This method has obvious lag, and detection work can only be carried out after the fault has occurred and caused a certain impact, which often leads to long-term power outages or communication interruptions, causing inconvenience to social production and residents' lives. At the same time, traditional positioning methods usually require comprehensive inspection of the entire cable line, which is labor-intensive, inefficient, and the positioning accuracy is greatly limited in complex environments. Summary of the invention

[0004] The present invention aims to solve the technical problems in the prior art of cable equipment fault positioning, such as serious hysteresis, low positioning efficiency and insufficient accuracy, and provides a cable equipment positioning system and method based on multi-source data fusion to solve the problems.

[0005] The technical solution of the present invention to solve the above technical problems is as follows: In the first aspect, the present invention provides a cable equipment positioning system based on multi-source data fusion, including: a reference state mining module, which is used to receive the operation control parameters, cable circuit topology and environmental parameters of the target area cable, perform frequent mode mining, and obtain the reference state, wherein the reference state has a topological area identifier; an inconsistent state identification module, which is used to extract the inconsistent state topological area identifier of the reference state and the monitoring state from the topological area identifier, and set it as the quasi-fault cable equipment distribution area; a detector matching module, which is used to match the detector array according to the quasi-fault cable equipment type and distribution mode in the quasi-fault cable equipment distribution area, wherein any detector of the detector array has a preset detection parameter label; a signal confidence retrieval module, which is used to retrieve the return signal confidence interval that satisfies the operation control parameters, the preset detection parameter label and the cable circuit topology for each detector in the detector array; a fault location output module, which is used to extract the detection topology area when the return signal acquisition value of the detector array does not belong to the return signal confidence interval, and set it as the fault cable equipment distribution area and send it to the user end.

[0006] Optionally, the benchmark state mining module includes: a first-level constraint construction unit, used to construct a first-level mining constraint condition based on the operation control parameters, the cable circuit topology and the environmental parameters; a first-level sample retrieval unit, used to retrieve the first-level state record value of the first-level sample cable circuit topology that meets the first-level mining constraint condition, wherein the first-level state record value has a first topological area record identifier, and the first-level sample cable circuit topology has a first-level operation control parameter label and a first-level environmental parameter label; a second-level constraint construction unit, used to construct a second-level mining constraint condition based on the first-level operation control parameter label, the cable circuit topology and the first-level environmental parameter label; a second-level sample retrieval unit, used to retrieve the second-level state record value of the second-level sample cable circuit topology that meets the second-level mining constraint condition, wherein the second-level state record value has a second topological area record identifier; an interval statistical identification unit, used to perform concentrated interval statistics on the first-level state record value and the second-level state record value with the same attributes to obtain the benchmark state, and configure the topological area identifier for the benchmark state according to the first topological area record identifier and the second topological area record identifier.

[0007] Optionally, the first-level constraint construction unit includes: a threshold configuration subunit, used to configure a first Euclidean distance threshold, a topological similarity threshold and a second Euclidean distance threshold; a condition satisfaction subunit, used to regard the first-level mining constraint condition as being satisfied when the first Euclidean distance between the sample operation control parameter and the operation control parameter is less than the first Euclidean distance threshold, and the topological similarity between the sample cable circuit topology and the cable circuit topology is greater than the topological similarity threshold, and the second Euclidean distance between the sample environment parameter and the environment parameter is less than the second Euclidean distance threshold; and a condition non-satisfaction subunit, used to regard the first-level mining constraint condition as being not satisfied otherwise.

[0008] Optionally, the signal confidence retrieval module includes: a retrieval constraint construction unit, used to construct a retrieval constraint condition based on the operation control parameters, the preset detection parameter label and the cable circuit topology; a fault-free sample collection unit, used to collect the first return signal attribute record value set until the Nth return signal attribute record value set of the cable fault-free sample group that meets the retrieval constraint condition; a signal trend analysis unit, used to traverse the first return signal attribute record value set until the Nth return signal attribute record value set to perform a central trend analysis, obtain the first return signal confidence interval until the Nth return signal confidence interval, and add it into the return signal confidence interval.

[0009] Optionally, the signal trend analysis unit includes: a characteristic value interval analysis subunit, used to traverse the first return signal attribute record value set until the Nth return signal attribute record value set to perform central trend analysis, and obtain the first return signal characteristic value interval until the Nth return signal characteristic value interval; a rated interval acquisition subunit, used to obtain the first return signal rated interval until the Nth return signal rated interval; a confidence interval construction subunit, used to compare the first return signal characteristic value interval and the first return signal rated interval, extract the maximum value of the lower limit of the interval and the minimum value of the upper limit of the interval, and construct the first return signal confidence interval; a multi-signal interval construction subunit, used until the Nth return signal confidence interval is constructed.

[0010] Optionally, the fault location output module includes: a signal deviation calculation unit, used to calculate the feedback signal deviation vector matrix of the first feedback signal collection value and the feedback signal confidence interval of the first detector of the detector array; a fault probability discrimination unit, used to process the feedback signal deviation vector matrix through a fault probability discriminator associated with the detector model of the first detector to obtain a fault probability prediction value; a fault judgment comparison unit, used to regard the first feedback signal collection value as inconsistent with the feedback signal confidence interval when the fault probability prediction value is greater than or equal to a fault probability threshold.

[0011] Optionally, the fault probability discrimination unit includes: a data set configuration subunit, used to configure the return signal deviation vector matrix data set of the detector model, retrieve the proportion of the number of failures of the detection cable equipment that satisfies the return signal deviation vector matrix data set, and set it as fault probability identification data; a discriminator training subunit, used to supervise the training of a number of front sub-discriminators based on the return signal deviation vector matrix data set and the fault probability identification data; a fitter training subunit, used to supervise the training of the output fitter based on a number of output results of the several front sub-discriminators and the fault probability identification data; a discriminator construction subunit, used to merge the output layers of the several front sub-discriminators with the input layer of the output fitter to obtain the fault probability discriminator.

[0012] In a second aspect, the present invention provides a cable equipment positioning method based on multi-source data fusion, including: receiving operation control parameters, cable circuit topology and environmental parameters of the target area cable, performing frequent mode mining, and obtaining a baseline state, wherein the baseline state has a topological area identifier; extracting the inconsistent state topological area identifier between the baseline state and the monitoring state from the topological area identifier, and setting it as a quasi-fault cable equipment distribution area; matching a detector array according to the quasi-fault cable equipment type and distribution mode of the quasi-fault cable equipment distribution area, wherein any one of the detectors in the detector array has a preset detection parameter label; for each detector in the detector array, retrieving a return signal confidence interval that satisfies the operation control parameters, the preset detection parameter label and the cable circuit topology; when the return signal acquisition value of the detector array does not belong to the return signal confidence interval, extracting the detection topological area, setting it as the faulty cable equipment distribution area and sending it to the user end.

[0013] The beneficial effects of the present invention are: The operating control parameters, cable circuit topology and environmental parameters of the target area cable are received through the benchmark state mining module, and frequent mode mining is performed to obtain the benchmark state, wherein the benchmark state has a topological area identifier, thereby establishing the benchmark operating mode of the cable equipment in different areas, providing a reference basis for subsequent fault warning. The inconsistent state identification module extracts the inconsistent state topological area identifier of the benchmark state and the monitoring state from the topological area identifier, and sets it as the distribution area of ​​the quasi-fault cable equipment, thereby finding the area that deviates from the normal operating state, preliminarily determining the location where there may be hidden faults, and narrowing the detection range of the entire line to a specific area to improve the detection efficiency. The detector matching module matches the detector array according to the type and distribution mode of the quasi-fault cable equipment in the distribution area of ​​the quasi-fault cable equipment, wherein any detector in the detector array has a preset detection parameter label, thereby realizing the customized configuration of the detection means. The signal confidence retrieval module retrieves the confidence interval of the return signal that meets the operating control parameters, the preset detection parameter label and the cable circuit topology for each detector in the detector array, thereby determining the normal variation range of the detection signal according to the actual operating environment and characteristics of the cable, establishing a benchmark for signal evaluation, and providing a basis for the identification of abnormal signals. When the return signal acquisition value of the detector array does not belong to the return signal confidence interval, the fault location output module extracts the detection topology area, sets it as the fault cable equipment distribution area and sends it to the user end, so as to accurately identify the fault location and promptly transmit the fault information to the user for timely processing.

[0014] Through the above technical scheme, the present invention realizes the transformation from passive response to active early warning, uses priori state to determine the quasi-fault area, and then conducts precise detection on the quasi-fault area, thereby improving the predictability, accuracy and efficiency of cable equipment fault location, and effectively reducing the losses and impacts caused by the fault. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 A schematic diagram of the structure of a cable equipment positioning system based on multi-source data fusion provided by the present invention; Figure 2 A schematic flow chart of a cable equipment positioning method based on multi-source data fusion provided by the present invention.

[0016] In the accompanying drawings, the components represented by the reference numerals are as follows: A reference state mining module 11, an inconsistent state identification module 12, a detector matching module 13, a signal confidence retrieval module 14, and a fault location output module 15. DETAILED DESCRIPTION

[0017] 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 those skilled in the art without creative work are within the scope of protection of the present invention.

[0018] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0019] In the description of the present invention, the term "for example" is used to mean "used as an example, illustration or explanation". Any embodiment described as "for example" in the present invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is given to enable any technician in the field to implement and use the present invention. In the following description, details are listed for the purpose of explanation. It should be understood that a person of ordinary skill in the art can recognize that the present invention can be implemented without using these specific details. In other examples, well-known structures and processes will not be elaborated in detail to avoid obscuring the description of the present invention with unnecessary details. Therefore, the present invention is not intended to be limited to the embodiments shown, but is consistent with the widest scope consistent with the principles and features disclosed in the present invention.

[0020] Embodiment 1, as Figure 1As shown, an embodiment of the present invention provides a cable equipment positioning system based on multi-source data fusion, including a reference state mining module 11, an inconsistent state identification module 12, a detector matching module 13, a signal confidence retrieval module 14 and a fault location output module 15. Among them, the reference state mining module 11 is used to receive the operation control parameters, cable circuit topology and environmental parameters of the target area cable, perform frequent mode mining, and obtain the reference state, wherein the reference state has a topological area identifier; the inconsistent state identification module 12 is used to extract the inconsistent state topological area identifier of the reference state and the monitoring state from the topological area identifier, and set it as the quasi-fault cable equipment distribution area; the detector matching module 13 is used to match the detector array according to the quasi-fault cable equipment type and distribution mode in the quasi-fault cable equipment distribution area, wherein any detector in the detector array has a preset detection parameter label; the signal confidence retrieval module 14 is used to retrieve the return signal confidence interval that satisfies the operation control parameters, the preset detection parameter label and the cable circuit topology for each detector in the detector array; the fault location output module 15 is used to extract the detection topology area when the return signal acquisition value of the detector array does not belong to the return signal confidence interval, set it as the fault cable equipment distribution area and send it to the user end.

[0021] Specifically, the reference state mining module 11 is responsible for modeling and mining the normal operating state of the target area cable. The reference state mining module 11 receives the operation control parameters, cable circuit topology information and environmental parameters of the target area cable. Among them, the operation control parameters include but are not limited to the electrical operation parameters such as the voltage level, load current, and power factor of the cable; the cable circuit topology characterizes the physical connection relationship, length, type distribution and other structural characteristics of the cable in the target area; the environmental parameters include temperature, humidity, geological conditions and other external factors that may affect the operation state of the cable. The reference state mining module 11 performs frequent modal mining on these multi-source heterogeneous data to extract the operation characteristic mode of the target area cable under normal conditions from a large amount of historical operation data to form a reference state. It is worth noting that the obtained reference state has a topological area identification, that is, each reference state corresponds to a specific area or section in the cable network, which provides a spatial reference basis for subsequent fault location. Through the reference state mining module 11, the system can establish a reference model of the cable equipment under normal operating conditions, laying the foundation for subsequent abnormal state identification.

[0022] The inconsistent state identification module is responsible for comparing the current monitoring state of the cable equipment with the reference state obtained by the reference state mining module 11, so as to identify the area where the abnormality exists. Specifically, the inconsistent state identification module 12 extracts the topological area where there is a significant difference between the reference state and the real-time monitoring state by analyzing the topological area identifier. When the parameters of the monitoring state are inconsistent with the parameters of the reference state, the topological area identifiers corresponding to these inconsistent states are extracted and set as the distribution area of ​​quasi-fault cable equipment. Among them, the quasi-fault indicates that there is a potential fault risk in the area, but it has not yet developed into a complete fault state. Through the inconsistent state identification module 12, the system can identify the areas where problems may occur in the cable network before the fault is fully formed, thereby providing a target area for subsequent precise detection and fault prevention, reducing the detection range, and improving the efficiency and accuracy of fault location.

[0023] The detector matching module 13 is responsible for selecting the most suitable detection equipment according to the characteristics of the quasi-fault cable equipment. This module receives the information of the distribution area of ​​the quasi-fault cable equipment determined by the non-consistent state identification module 12, and analyzes the type and distribution mode of the quasi-fault cable equipment in the area. The detector matching module 13 selects and deploys the most suitable detector array according to the characteristics of different types of quasi-fault cable equipment. For example, electromagnetic detectors are matched for metal cable equipment, while acoustic positioning detectors are matched for direct buried cable equipment or cable equipment in pipes, and geological radars are matched for non-metallic cable equipment. This targeted detector selection mechanism significantly improves the accuracy of fault location. Among them, each detector in the detector array has a preset detection parameter label, which contains key information such as the technical parameters, applicable conditions, and sensitivity range of the detector, which provides the necessary reference basis for subsequent signal processing and analysis. Through the detector matching module 13, the most suitable detection method can be selected according to the characteristics of different cable types, thereby improving the pertinence and effectiveness of fault location.

[0024] The signal confidence retrieval module 14 is responsible for processing each detector in the detector array individually and establishing a standard reference interval for the return signal of each detector. Specifically, the signal confidence retrieval module 14 first establishes the constraints of the signal retrieval according to the operation control parameters of the current cable system, the preset detection parameter labels of the detectors, and the cable circuit topology information. Then, based on these constraints, the module retrieves historical data that matches the current situation and extracts the normal return signal characteristics of the detector under a specific combination of conditions. By performing statistical analysis and trend research on these historical return signal data, the signal confidence retrieval module 14 establishes a return signal confidence interval for each detector. These confidence intervals represent the reasonable range of variation of the detector return signal under normal operating conditions, providing a reference standard for subsequent abnormal signal judgment. Through the signal confidence retrieval module 14, the system can accurately identify abnormal fluctuations in the detection signal, thereby improving the accuracy and reliability of fault judgment and laying a data foundation for the final fault location.

[0025] The fault location output module 15 is responsible for determining and outputting the accurate location information of the faulty cable equipment. This module analyzes the actual signal data obtained by the detector array and compares it with the return signal confidence interval established by the signal confidence retrieval module 14 to achieve accurate fault location. Specifically, when the return signal value collected by the detector in the detector array falls outside the preset return signal confidence interval, the fault location output module 15 will determine that there is an abnormality in the topological area corresponding to the detector. Subsequently, the module accurately extracts the topological area information corresponding to these abnormal detection signals and determines it as the distribution area of ​​the faulty cable equipment. After the fault location output module 15 completes the fault area, it sends the positioning result to the user end through the system interface so that maintenance personnel can obtain fault information in time and take corresponding maintenance measures. This real-time fault information transmission mechanism significantly improves the timeliness of cable system fault handling.

[0026] The cable fault is located in advance through the coordinated work of the reference state mining module 11, the inconsistent state identification module 12, the detector matching module 13, the signal confidence retrieval module 14 and the fault location output module 15. Different from the traditional post-detection method, the quasi-fault area is determined by using the prior state, and then the appropriate detector is selected for precise detection, such as electromagnetic detectors for metal cables, acoustic positioning detectors for directly buried or in-pipe cables, and geological radar for non-metallic cables. After that, the signal confidence interval and fault discrimination mechanism are used to achieve accurate analysis and abnormal judgment of the return signal. When the return signal acquisition value of the detector array does not belong to the preset return signal confidence interval, the corresponding detection topology area can be accurately extracted and determined as the distribution area of ​​the faulty cable equipment, thereby improving the predictability, accuracy and efficiency of fault location, reducing the false alarm rate and missed alarm rate, and providing support for preventive maintenance and rapid fault handling of the power system.

[0027] Further, the reference state mining module 11 includes a primary constraint construction unit, a primary sample retrieval unit, a secondary constraint construction unit, a secondary sample retrieval unit and an interval statistical identification unit. Among them, the primary constraint construction unit is used to construct a primary mining constraint condition according to the operation control parameters, the cable circuit topology and the environmental parameters; the primary sample retrieval unit is used to retrieve the primary state record value of the primary sample cable circuit topology that meets the primary mining constraint condition, wherein the primary state record value has a first topological area record identifier, and the primary sample cable circuit topology has a primary operation control parameter label and a primary environmental parameter label; the secondary constraint construction unit is used to construct a secondary mining constraint condition according to the primary operation control parameter label, the cable circuit topology and the primary environmental parameter label; the secondary sample retrieval unit is used to retrieve the secondary state record value of the secondary sample cable circuit topology that meets the secondary mining constraint condition, wherein the secondary state record value has a second topological area record identifier; the interval statistical identification unit is used to perform the same attribute concentrated interval statistics on the primary state record value and the secondary state record value, obtain the reference state, and configure the topological area identifier for the reference state according to the first topological area record identifier and the second topological area record identifier.

[0028] In a preferred embodiment, the baseline state mining module 11 includes a primary constraint construction unit, a primary sample retrieval unit, a secondary constraint construction unit, a secondary sample retrieval unit and an interval statistical identification unit, thereby adopting a multi-level constraint construction and sample retrieval strategy to improve the accuracy and comprehensiveness of baseline state mining.

[0029] The first-level constraint construction unit is responsible for constructing preliminary mining constraints, namely, first-level mining constraints, based on the current operation control parameters, cable circuit topology, and environmental parameters. The first-level mining constraints define the basic condition framework for the system to find similar cases in historical data, providing a screening basis for subsequent sample retrieval. Then, based on the first-level constraints, the first-level sample retrieval unit retrieves cable circuit samples that meet the first-level mining constraints from the historical database and extracts the state record values ​​of these samples, namely, the first-level state record values. Each first-level state record value is associated with a specific topological area and is marked by the first topological area record identifier. In addition, the retrieved first-level sample cable circuit topology is also accompanied by a first-level operation control parameter label and a first-level environmental parameter label, which provide a reference for the construction of the second-level constraints.

[0030] The secondary constraint construction unit uses the primary operation control parameter labels, the current cable circuit topology, and the primary environmental parameter labels extracted from the primary samples to construct the secondary mining constraints. This hierarchical constraint construction method enables the system to retrieve samples under the secondary mining constraints. Although the samples retrieved through the secondary mining constraints have a low direct similarity with the current state, they are similar to the primary samples, so they still have important reference value and can effectively expand the sample pool that can be used for benchmark state modeling. Subsequently, the secondary sample retrieval unit retrieves the cable circuit samples based on the secondary mining constraints and extracts the state record values ​​of these samples as the secondary state record values. Each secondary state record value also has a topological area identifier, namely the second topological area record identifier, which is used to mark the specific topological position corresponding to the state value.

[0031] Afterwards, the interval statistical identification unit performs concentrated interval statistical analysis of the same attributes on the state record values ​​(primary state record values ​​and secondary state record values) obtained from the two levels, extracts the parameter interval that can represent the normal operating state, and forms the final reference state. At the same time, the unit also configures the corresponding reference state configuration topological area identifier for each reference state according to the first topological area record identifier and the second topological area record identifier, and establishes the corresponding relationship between the reference state and the topological position of the cable network.

[0032] Through multi-level constraints and hierarchical retrieval methods, the benchmark state mining module 11 can effectively extract parameter models that can accurately characterize the normal operating status of cable equipment from a large amount of historical data, laying a data foundation for subsequent inconsistent state identification.

[0033] Furthermore, the first-level constraint construction unit includes a threshold configuration subunit, a condition satisfaction subunit and a condition non-satisfaction subunit. Among them, the threshold configuration subunit is used to configure the first Euclidean distance threshold, the topological similarity threshold and the second Euclidean distance threshold; the condition satisfaction subunit is used to regard the first-level mining constraint condition as being satisfied when the first Euclidean distance between the sample operation control parameter and the operation control parameter is less than the first Euclidean distance threshold, and the topological similarity between the sample cable circuit topology and the cable circuit topology is greater than the topological similarity threshold, and the second Euclidean distance between the sample environmental parameter and the environmental parameter is less than the second Euclidean distance threshold; the condition non-satisfaction subunit is used to regard the first-level mining constraint condition as being not satisfied otherwise.

[0034] In a preferred embodiment, the primary constraint construction unit includes a threshold configuration subunit, a condition satisfaction subunit and a condition non-satisfaction subunit, which are used to accurately evaluate the similarity between the historical sample and the current state.

[0035] The threshold configuration subunit is responsible for setting the key threshold parameters for judging whether the sample meets the constraint conditions, which are the first Euclidean distance threshold, the topological similarity threshold and the second Euclidean distance threshold. Among them, the first Euclidean distance threshold is used to measure the similarity of the operation control parameters, the topological similarity threshold is used to measure the similarity of the cable circuit topology structure, and the second Euclidean distance threshold is used to measure the similarity of the environmental parameters.

[0036] The condition-satisfied subunit is used to implement the specific judgment logic of the first-level constraint conditions, that is, the sample is judged to meet the first-level mining constraint conditions if and only if the following three conditions are met at the same time: first, the Euclidean distance between the sample operation control parameter and the current operation control parameter is less than the preset first Euclidean distance threshold, indicating that the two are close enough in electrical parameters; second, the topological similarity between the sample cable circuit topology and the current cable circuit topology is greater than the preset topological similarity threshold, indicating that the two have sufficient similarity in physical structure; third, the Euclidean distance between the sample environmental parameter and the current environmental parameter is less than the preset second Euclidean distance threshold, indicating that the two are close enough in external environmental conditions. The condition-unsatisfied subunit handles the situation where the sample does not meet the first-level mining constraint conditions, that is, when any of the above three conditions is not met, the condition-unsatisfied subunit determines that the sample does not meet the first-level mining constraint conditions, and is thus excluded from the scope of the first-level sample retrieval.

[0037] Through constraint construction based on multi-dimensional similarity calculation, the first-level constraint construction unit can accurately identify historical samples that are highly similar to the current state, ensuring the accuracy and reliability of first-level sample retrieval and laying the foundation for accurate mining of baseline states.

[0038] Furthermore, the signal confidence retrieval module includes a retrieval constraint construction unit, a fault-free sample collection unit and a signal trend analysis unit. Among them, the retrieval constraint construction unit is used to construct retrieval constraints according to the operation control parameters, the preset detection parameter labels and the cable circuit topology; the fault-free sample collection unit is used to collect the first return signal attribute record value set of the cable fault-free sample group that meets the retrieval constraint conditions until the Nth return signal attribute record value set; the signal trend analysis unit is used to traverse the first return signal attribute record value set until the Nth return signal attribute record value set to perform a central trend analysis, obtain the first return signal confidence interval until the Nth return signal confidence interval, and add it into the return signal confidence interval.

[0039] In a feasible implementation, the signal confidence retrieval module 14 includes a retrieval constraint construction unit, a fault-free sample collection unit and a signal trend analysis unit, which together complete the establishment of the return signal confidence interval, thereby ensuring the accuracy of the detection signal discrimination.

[0040] The retrieval constraint construction unit first constructs the retrieval constraint conditions according to the current operation control parameters, the preset detection parameter labels of the detector and the cable circuit topology information. The retrieval constraint conditions define the screening criteria for finding similar fault-free cases in the historical data, ensuring that the retrieved samples have characteristics that match the current situation, thereby ensuring the applicability and effectiveness of the signal confidence interval established subsequently. Based on the above retrieval constraints, the fault-free sample acquisition unit selects the cable fault-free sample group that meets the conditions from the historical database. For these samples, the fault-free sample acquisition unit extracts their multi-dimensional return signal attribute record values ​​to form a complete data set from the first return signal attribute record value set to the Nth return signal attribute record value set. Among them, N represents the number of return signal attributes that the detector can collect, and different types of detectors may have different numbers of signal attributes. Subsequently, the signal trend analysis unit systematically traverses and analyzes the collected multi-dimensional signal attribute record value set. The signal trend analysis unit extracts the confidence interval that can characterize the change law of the signal attribute under normal conditions by performing a central trend analysis on the historical record values ​​of each dimensional signal attribute. Thus, a complete interval set from the first feedback signal confidence interval to the Nth feedback signal confidence interval is formed, and these intervals are added to the feedback signal confidence interval.

[0041] By extracting signal features based on historical fault-free samples, the signal confidence retrieval module 14 establishes a return signal judgment standard for each detector, so that the system can accurately identify abnormal patterns in the detection signal and provide data support for the final fault location.

[0042] Furthermore, the signal trend analysis unit includes a characteristic value interval analysis subunit, a rated interval acquisition subunit, a confidence interval construction subunit and a multi-signal interval construction subunit. Among them, the characteristic value interval analysis subunit is used to traverse the first feedback signal attribute record value set until the Nth feedback signal attribute record value set to perform a central trend analysis, and obtain the first feedback signal characteristic value interval until the Nth feedback signal characteristic value interval; the rated interval acquisition subunit is used to obtain the first feedback signal rated interval until the Nth feedback signal rated interval; the confidence interval construction subunit is used to compare the first feedback signal characteristic value interval and the first feedback signal rated interval, extract the interval lower limit maximum value and the interval upper limit minimum value, and construct the first feedback signal confidence interval; the multi-signal interval construction subunit is used until the Nth feedback signal confidence interval is constructed.

[0043] Specifically, the signal trend analysis unit includes a characteristic value interval analysis subunit, a rated interval acquisition subunit, a confidence interval construction subunit and a multi-signal interval construction subunit, which together complete the conversion process from sample data to confidence intervals to achieve accurate signal trend analysis.

[0044] The eigenvalue interval analysis subunit first systematically traverses the set of multi-dimensional feedback signal attribute record values ​​collected. The eigenvalue interval analysis subunit performs a central trend analysis on the historical record values ​​for each dimensional signal attribute, and extracts the interval range that can characterize the signal change characteristics under normal conditions. Through data-driven analysis, a complete feature description from the first feedback signal eigenvalue interval to the Nth feedback signal eigenvalue interval is obtained. The rated interval acquisition subunit is responsible for extracting the rated working interval of each dimensional feedback signal, that is, the normal change range of the signal specified by the equipment specifications, technical standards or design parameters. The rated interval acquisition subunit obtains a complete set of rated intervals from the first feedback signal rated interval to the Nth feedback signal rated interval, providing a theoretical reference for the subsequent confidence interval construction.

[0045] The confidence interval construction subunit compares the characteristic value interval with the rated interval, extracts the maximum value of the lower limit of the interval and the minimum value of the upper limit of the interval, and thus constructs the first feedback signal confidence interval that meets both the statistical characteristics of historical data and the theoretical rated requirements. This double insurance interval construction can improve the reliability and accuracy of the confidence interval. The multi-signal interval construction subunit is responsible for applying the above confidence interval construction process to the feedback signals of all dimensions in turn, extending from the first dimension to the Nth dimension, and finally completing the construction of all feedback signal confidence intervals.

[0046] By integrating statistical features with theoretical norms for signal trend analysis, the signal trend analysis unit can establish rigorous judgment criteria for each dimension of the return signal, thereby improving the accuracy and reliability of the system in identifying abnormal signals in complex environments.

[0047] Furthermore, the fault location output module includes a signal deviation calculation unit, a fault probability determination unit and a fault determination comparison unit. The signal deviation calculation unit is used to calculate the return signal deviation vector matrix of the first return signal acquisition value of the first detector of the detector array and the return signal confidence interval; the fault probability determination unit is used to process the return signal deviation vector matrix through the fault probability discriminator associated with the detector model of the first detector to obtain the fault probability prediction value; the fault determination comparison unit is used to regard the first return signal acquisition value as inconsistent with the return signal confidence interval when the fault probability prediction value is greater than or equal to the fault probability threshold.

[0048] In a preferred embodiment, the fault location output module 15 includes a signal deviation calculation unit, a fault probability determination unit and a fault judgment comparison unit, which together complete the entire process from signal anomaly detection to fault judgment, and achieve accurate fault judgment and location output.

[0049] First, the signal deviation calculation unit processes each detector in the detector array. Taking the first detector as an example, the unit calculates the degree of deviation between the first return signal acquisition value of the first detector and the pre-established return signal confidence interval. By quantifying this deviation relationship, a return signal deviation vector matrix is ​​generated. The matrix comprehensively characterizes the difference characteristics between the actual signal and the normal range, and provides a data basis for the subsequent fault probability assessment. The fault probability discrimination unit processes the return signal deviation vector matrix in combination with the detector model characteristics. Specifically, the fault probability discrimination unit calls the fault probability discriminator specifically associated with the detector model of the first detector, performs a deep analysis of the signal deviation characteristics, and outputs a fault probability prediction value. This special discrimination mechanism based on detector characteristics improves the accuracy and adaptability of fault identification. Subsequently, the fault judgment comparison unit performs the final fault judgment by comparing the fault probability prediction value with the preset fault probability threshold. When the fault probability prediction value is greater than or equal to the fault probability threshold, it is determined that there is a substantial inconsistency between the first return signal acquisition value and the return signal confidence interval, that is, it is considered that the topological area corresponding to the detector may have a fault.

[0050] Through fault judgment based on probability evaluation, the fault location output module 15 can effectively filter random noise in signal fluctuations, improve the robustness and reliability of fault judgment, and provide an accurate basis for the final fault area positioning.

[0051] Furthermore, the fault probability discrimination unit includes a data set configuration subunit, a discriminator training subunit, a fitter training subunit and a discriminator construction subunit. Among them, the data set configuration subunit is used to configure the return signal deviation vector matrix data set of the detector model, retrieve the proportion of the number of faults of the detection cable equipment that meets the return signal deviation vector matrix data set, and set it as the fault probability identification data; the discriminator training subunit is used to supervise the training of several front sub-discriminators according to the return signal deviation vector matrix data set and the fault probability identification data; the fitter training subunit is used to supervise the training of the output fitter according to several output results of several front sub-discriminators and the fault probability identification data; the discriminator construction subunit is used to merge the output layers of several front sub-discriminators and the input layer of the output fitter to obtain the fault probability discriminator.

[0052] In a preferred embodiment, the fault probability discrimination unit specifically includes a data set configuration subunit, a discriminator training subunit, a fitter training subunit and a discriminator construction subunit, so as to construct a fault probability discriminator by adopting an integrated method of machine learning, thereby improving the accuracy and reliability of fault probability discrimination.

[0053] The dataset configuration subunit is responsible for establishing the data foundation for model training. The dataset configuration subunit configures the historical return signal deviation vector matrix as a training dataset for a specific detector model. At the same time, by retrieving the historical fault records of the cable equipment associated with these matrices, the proportion of fault times is calculated as fault probability identification data. These identification data provide a mapping relationship between the signal deviation characteristics and the actual fault probability, providing a reference for subsequent supervised learning. Then, the discriminator training subunit uses a supervised learning method to train multiple front-end sub-discriminators based on the above-mentioned return signal deviation vector matrix dataset. These front-end sub-discriminators use different model structures or parameter configurations, which can capture abnormal features in the return signal deviation vector matrix from different angles or levels, and provide a multi-dimensional decision-making basis for the final fault probability judgment.

[0054] The fitter training subunit uses the multi-channel output results of the front sub-discriminator as a new feature vector, combined with the original fault probability identification data, to train a special output fitter. This fitter can effectively integrate the judgment results of multiple sub-discriminators, eliminate the bias of a single model, and further improve the accuracy of fault probability prediction. Afterwards, the discriminator construction subunit structurally merges the output layers of several front sub-discriminators with the input layer of the output fitter to form an end-to-end fault probability discriminator. This integrated learning architecture effectively combines the advantages of multiple models and improves the system's ability to recognize and resist various fault modes.

[0055] Through the fault probability judgment mechanism based on the machine learning integration method, a dedicated judgment model can be established for different types of detectors to achieve accurate identification of complex cable fault modes, providing a high-reliability judgment basis for the fault location output module.

[0056] Embodiment 2, as Figure 2 As shown, based on the same inventive concept of a cable equipment positioning system based on multi-source data fusion provided in Embodiment 1, an embodiment of the present invention further provides a cable equipment positioning method based on multi-source data fusion, including: Receiving operation control parameters, cable circuit topology and environmental parameters of the target area cable, performing frequent mode mining, and obtaining a reference state, wherein the reference state has a topological area identifier; Extract the topological region identifier of the inconsistent state between the reference state and the monitoring state from the topological region identifier, and set it as the distribution area of ​​the quasi-fault cable equipment; According to the type and distribution mode of the quasi-faulty cable equipment in the distribution area of ​​the quasi-faulty cable equipment, a detector array is matched, wherein any detector in the detector array has a preset detection parameter label; For each detector of the detector array, a confidence interval of a return signal satisfying an operation control parameter, a preset detection parameter label and a cable circuit topology is retrieved; When the return signal acquisition value of the detector array does not belong to the return signal confidence interval, the detection topology area is extracted, set as the faulty cable equipment distribution area and sent to the user end.

[0057] Furthermore, the operation control parameters, cable circuit topology and environmental parameters of the target area cables are received, and frequent mode mining is performed to obtain the baseline state, including: According to the operation control parameters, cable circuit topology and environmental parameters, the first-level mining constraints are constructed; Retrieving a first-level state record value of a first-level sample cable circuit topology that meets a first-level mining constraint condition, wherein the first-level state record value has a first topology area record identifier, and the first-level sample cable circuit topology has a first-level operation control parameter tag and a first-level environment parameter tag; According to the first-level operation control parameter labels, cable circuit topology and first-level environmental parameter labels, the second-level mining constraints are constructed; Retrieving a secondary state record value of a secondary sample cable circuit topology that satisfies the secondary mining constraint condition, wherein the secondary state record value has a second topology region record identifier; The primary state record value and the secondary state record value are statistically analyzed in a concentrated interval with the same attribute to obtain a reference state, and a topological area identifier is configured for the reference state according to the first topological area record identifier and the second topological area record identifier.

[0058] Furthermore, based on the operation control parameters, cable circuit topology and environmental parameters, the first-level mining constraints are constructed, including: Configure the first Euclidean distance threshold, topological similarity threshold, and second Euclidean distance threshold; When the first Euclidean distance between the sample operation control parameters and the operation control parameters is less than the first Euclidean distance threshold, and the topological similarity between the sample cable circuit topology and the cable circuit topology is greater than the topological similarity threshold, and the second Euclidean distance between the sample environmental parameters and the environmental parameters is less than the second Euclidean distance threshold, it is considered that the first-level mining constraint condition is met; Otherwise, it is considered that the first-level mining constraints are not met.

[0059] Furthermore, for each detector of the detector array, a confidence interval of a return signal satisfying the operation control parameter, the preset detection parameter label and the cable circuit topology is retrieved, including: Construct search constraints based on operation control parameters, preset detection parameter labels and cable circuit topology; Collecting the first return signal attribute record value set until the Nth return signal attribute record value set of the cable fault-free sample group that meets the retrieval constraint condition; Traversing the first return signal attribute record value set until the Nth return signal attribute record value set to perform central tendency analysis, obtaining the first return signal confidence interval until the Nth return signal confidence interval, and adding it into the return signal confidence interval, further comprising: Traversing the first return signal attribute record value set until the Nth return signal attribute record value set to perform central trend analysis, and obtaining the first return signal characteristic value interval until the Nth return signal characteristic value interval; Obtaining a first return signal rated interval until an Nth return signal rated interval; Compare the characteristic value interval of the first return signal with the rated interval of the first return signal, extract the maximum value of the lower limit of the interval and the minimum value of the upper limit of the interval, and construct a confidence interval of the first return signal; Until the Nth feedback signal confidence interval is constructed.

[0060] Furthermore, when the return signal acquisition value of the detector array does not belong to the return signal confidence interval, it includes: Calculate a return signal deviation vector matrix of a first return signal acquisition value of a first detector of the detector array and a return signal confidence interval; Processing the return signal deviation vector matrix through a fault probability discriminator associated with the detector model of the first detector to obtain a fault probability prediction value; When the fault probability prediction value is greater than or equal to the fault probability threshold, it is considered that the first return signal acquisition value is inconsistent with the return signal confidence interval.

[0061] Further, the fault probability prediction value is obtained by processing the return signal deviation vector matrix through a fault probability discriminator associated with the detector model of the first detector, including: A return signal deviation vector matrix data set of the configuration detector model is retrieved, and the failure times ratio of the detection cable equipment that satisfies the return signal deviation vector matrix data set is retrieved and set as the failure probability identification data; According to the feedback signal deviation vector matrix data set and fault probability identification data, supervise the training of several front-end sub-discriminators; According to the output results of the plurality of pre-discriminators and the fault probability identification data, supervise the training of the output fitter; The output layers of several pre-discriminators are combined with the input layer of the output fitter to obtain a fault probability discriminator.

[0062] It should be noted that in the above embodiments, the description of each embodiment has its own emphasis, and for parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0063] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0064] The present invention is described with reference to flowcharts and / or block diagrams of systems, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded computer, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0065] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1A function specified in one or more boxes.

[0066] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0067] Although preferred embodiments of the present invention have been described, additional changes and modifications may occur to these embodiments once those skilled in the art understand the basic inventive concepts.

[0068] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention belong to the scope of the present invention and its equivalent technologies, the present invention is also intended to include these changes and variations.

Claims

1. A cable equipment positioning system based on multi-source data fusion, characterized in that: include: A reference state mining module is used to receive operation control parameters, cable circuit topology and environmental parameters of the target area cable, perform frequent mode mining, and obtain a reference state, wherein the reference state has a topological area identifier; An inconsistent state identification module is used to extract the inconsistent state topology area identifier of the reference state and the monitoring state from the topology area identifier, and set it as a quasi-fault cable equipment distribution area; A detector matching module, used to match the detector array according to the type and distribution mode of the quasi-faulty cable equipment in the distribution area of ​​the quasi-faulty cable equipment, wherein any detector of the detector array has a preset detection parameter label; A signal confidence retrieval module, used for retrieving, for each detector of the detector array, a return signal confidence interval that satisfies the operation control parameter, the preset detection parameter label and the cable circuit topology; The fault location output module is used to extract the detection topology area when the return signal collection value of the detector array does not belong to the return signal confidence interval, set it as the faulty cable equipment distribution area and send it to the user end.

2. The system according to claim 1, characterized in that The benchmark state mining module includes: A primary constraint construction unit, used to construct a primary mining constraint condition according to the operation control parameter, the cable circuit topology and the environmental parameter; A primary sample retrieval unit, used to retrieve a primary state record value of a primary sample cable circuit topology that satisfies the primary mining constraint condition, wherein the primary state record value has a first topology area record identifier, and the primary sample cable circuit topology has a primary operation control parameter tag and a primary environment parameter tag; A secondary constraint construction unit, used to construct a secondary mining constraint condition according to the primary operation control parameter label, the cable circuit topology and the primary environmental parameter label; A secondary sample retrieval unit, used to retrieve a secondary state record value of a secondary sample cable circuit topology that meets the secondary mining constraint condition, wherein the secondary state record value has a second topology area record identifier; An interval statistics identification unit is used to perform concentrated interval statistics of the primary state record value and the secondary state record value with the same attribute to obtain the reference state, and configure the topology area identifier for the reference state according to the first topology area record identifier and the second topology area record identifier.

3. The system according to claim 2, characterized in that The primary constraint construction unit comprises: A threshold configuration subunit, used to configure a first Euclidean distance threshold, a topological similarity threshold, and a second Euclidean distance threshold; A condition satisfaction subunit, used for considering that the first-level mining constraint condition is satisfied when the first Euclidean distance between the sample operation control parameter and the operation control parameter is less than the first Euclidean distance threshold, and the topological similarity between the sample cable circuit topology and the cable circuit topology is greater than the topological similarity threshold, and the second Euclidean distance between the sample environment parameter and the environment parameter is less than the second Euclidean distance threshold; The condition is not satisfied subunit, and is used for otherwise, it is considered that the first-level mining constraint condition is not satisfied.

4. The system according to claim 1, characterized in that The signal confidence retrieval module comprises: A search constraint construction unit, used to construct a search constraint condition according to the operation control parameter, the preset detection parameter label and the cable circuit topology; A fault-free sample collection unit, used to collect the first return signal attribute record value set to the Nth return signal attribute record value set of the cable fault-free sample group that meets the retrieval constraint condition; A signal trend analysis unit is used to traverse the first feedback signal attribute record value set until the Nth feedback signal attribute record value set to perform central trend analysis, obtain the first feedback signal confidence interval until the Nth feedback signal confidence interval, and add it into the feedback signal confidence interval.

5. The system according to claim 4, characterized in that The signal trend analysis unit comprises: A characteristic value interval analysis subunit, used for traversing the first return signal attribute record value set until the Nth return signal attribute record value set to perform a central trend analysis, and obtaining the first return signal characteristic value interval until the Nth return signal characteristic value interval; A rated interval acquisition subunit, used to obtain the rated interval of the first return signal until the rated interval of the Nth return signal; A confidence interval construction subunit, used to compare the characteristic value interval of the first return signal with the rated interval of the first return signal, extract the maximum value of the lower limit of the interval and the minimum value of the upper limit of the interval, and construct the confidence interval of the first return signal; The multi-signal interval construction subunit is used to construct the Nth feedback signal confidence interval.

6. The system according to claim 1, wherein: The fault location output module comprises: A signal deviation calculation unit, used to calculate a return signal deviation vector matrix between a first return signal acquisition value of a first detector of the detector array and a return signal confidence interval; a fault probability discriminating unit, configured to process the return signal deviation vector matrix through a fault probability discriminator associated with a detector model of the first detector to obtain a fault probability prediction value; The fault judgment comparison unit is used to regard that the first feedback signal acquisition value is inconsistent with the feedback signal confidence interval when the fault probability prediction value is greater than or equal to the fault probability threshold.

7. The system according to claim 6, characterized in that The fault probability determination unit comprises: A data set configuration subunit, used to configure a return signal deviation vector matrix data set of the detector model, retrieve the proportion of the number of failures of the detection cable equipment that satisfies the return signal deviation vector matrix data set, and set it as the failure probability identification data; A discriminator training subunit, used for supervised training of a plurality of front-end sub-discriminators according to the feedback signal deviation vector matrix data set and the fault probability identification data; A fitter training subunit, used for supervising the training of an output fitter according to the output results of the plurality of pre-discriminators and the fault probability identification data; The discriminator construction subunit is used to merge the output layers of the plurality of front sub-discriminators with the input layer of the output fitter to obtain the fault probability discriminator.

8. A cable equipment positioning method based on multi-source data fusion, characterized in that: include: Receiving operation control parameters, cable circuit topology and environmental parameters of the target area cable, performing frequent mode mining, and obtaining a reference state, wherein the reference state has a topological area identifier; Extracting the inconsistent state topology area identifier of the reference state and the monitoring state from the topology area identifier, and setting it as the quasi-fault cable equipment distribution area; Matching a detector array according to the type and distribution mode of the quasi-faulty cable equipment in the distribution area of ​​the quasi-faulty cable equipment, wherein any detector of the detector array has a preset detection parameter label; For each detector of the detector array, retrieving a confidence interval of a return signal that satisfies the operation control parameter, the preset detection parameter label and the cable circuit topology; When the return signal acquisition value of the detector array does not belong to the return signal confidence interval, the detection topology area is extracted, set as the faulty cable equipment distribution area and sent to the user end.

Images