Intelligent state detection system for dry-type transformer

Abstract

The invention relates to the technical field of transformer detection, in particular to an intelligent state detection system for a dry-type transformer, and the system comprises a data obtaining module which is used for obtaining the environment parameters and operation parameters of the dry-type transformer; the load type determination module is used for determining the operation load type of the dry-type transformer based on the operation parameters; the potential fault positioning module is used for responding to a low load type and positioning a potential fault position based on the operation parameters; the detection path planning module is used for determining a detection path based on the potential fault position and the three-dimensional structure model; and the path adjustment module is used for determining a fault influence area corresponding to the detection position based on the fault detection result, and adjusting a subsequent detection path based on the fault influence area. According to the method, through the technical means of combining iron core vibration distortion analysis under the low-load working condition with dynamic path planning and adjustment, the beneficial effects of accurately positioning potential faults in the low-load period and adaptively optimizing the detection path are achieved.

Description

Technical Field

[0001] This invention relates to the field of transformer testing technology, and in particular to an intelligent condition detection system for dry-type transformers. Background Technology

[0002] Dry-type transformers are widely used in high-rise buildings, commercial centers, industrial parks, and other locations due to their advantages such as being oil-free, fire-resistant, and environmentally friendly. However, during long-term operation, potential faults in dry-type transformers, such as loose cores, aging insulation, and poor contact of connectors, are difficult to detect in a timely manner through conventional means. If these faults are not identified and located early, they may lead to sudden equipment failures and affect the reliability of power supply.

[0003] Traditional condition monitoring of dry-type transformers mainly relies on periodic power outage preventative tests and offline testing, such as insulation resistance testing and partial discharge testing. These methods have the following drawbacks: first, they require equipment shutdown, affecting normal power supply; second, the testing cycle is long, making real-time monitoring and early warning impossible; and third, the test results are greatly affected by load conditions—vibration and noise under high loads can easily mask fault characteristics, while fault signals are weak under low loads, easily leading to missed detections.

[0004] Chinese Patent Application Publication No. CN102759670A discloses a comprehensive evaluation method for the operating status of dry-type transformers. This method integrates the influence of various parameters on the operating status of dry-type transformers, providing a reliable level analysis of their operating status. This allows operators to monitor the operating condition of the dry-type transformer in real time and perform timely condition-based maintenance. The method includes the following steps: 1) Detecting parameter data of the operating conditions and operating environment of the dry-type transformer, and obtaining parameter data of the transformer's inherent characteristics and operating history; 2) Normalizing the parameter data of the operating conditions, operating environment, inherent characteristics, and operating history of the dry-type transformer; 3) Obtaining the membership degree of each parameter and its corresponding evaluation citation through the membership function of each parameter and citation, and obtaining a single-factor evaluation matrix; 4) Combining the weights of each parameter to obtain a comprehensive evaluation matrix; 5) Obtaining a comprehensive evaluation value of the operating status of the dry-type transformer; the evaluation value reflects the operating status of the dry-type transformer.

[0005] Therefore, the existing technology has the following problems: The comprehensive evaluation based solely on multi-parameter data failed to accurately identify the weak fault signals under low load conditions, and the detection path could not be dynamically adjusted according to the detection results, resulting in a low rate of missed faults and low detection efficiency. Summary of the Invention

[0006] To address this, the present invention provides an intelligent condition detection system for dry-type transformers, which overcomes the problems in the prior art that it fails to accurately identify the weak fault signals under low load conditions and cannot dynamically adjust the detection path based on the detection results, resulting in a low rate of missed fault detection and low detection efficiency.

[0007] To achieve the above objectives, the present invention provides an intelligent condition detection system for dry-type transformers, comprising: The data acquisition module is used to obtain the environmental parameters and operating parameters of the dry-type transformer. A load type determination module, connected to the data module, is used to determine the operating load type of the dry-type transformer based on the operating parameters, wherein the load type includes a high load type and a low load type; A potential fault location module is connected to the data acquisition module and the load type determination module respectively. In response to the dry-type transformer operating load type being low load type, it acquires the vibration signal of the iron core in the dry-type transformer from the operating parameters, determines the vibration distortion index of the dry-type transformer based on the vibration signal, and locates the potential fault location of the dry-type transformer based on the vibration distortion index. A detection path planning module, which is connected to the potential fault location module, is used to determine the detection path for the fault location based on the potential fault location of the dry-type transformer and the three-dimensional structural model of the dry-type transformer. A fault detection module, which is connected to the detection path planning module, is used to perform fault detection on the dry-type transformer based on the detection path and obtain fault detection results. A path adjustment module, which is connected to the detection path planning module and the fault detection module respectively, is used to determine the fault impact area of ​​the detection location corresponding to the fault detection result based on the fault detection result, and adjust the subsequent detection path of the fault detection module based on the fault impact area.

[0008] Furthermore, the load type determination module determines the operating load type of the dry-type transformer based on the operating parameters, wherein, The load type determination module determines the real-time load of the dry-type transformer based on the operating parameters; If the real-time load is greater than or equal to the preset real-time load, the load type determination module determines that the operating load type of the dry-type transformer is a high load type; If the real-time load is less than the preset real-time load, the load type determination module determines that the operating load type of the dry-type transformer is a low load type.

[0009] Furthermore, the potential fault location module determines the vibration distortion index of the dry-type transformer based on the vibration signal, wherein, The potential fault location module performs spectral analysis on the vibration signal and extracts the amplitude of the fundamental frequency component and the amplitude of the harmonic component. The vibration distortion index is determined based on the ratio of the harmonic component amplitude to the fundamental frequency component amplitude.

[0010] Furthermore, the potential fault location module locates the potential fault location of the dry-type transformer based on the vibration distortion index, wherein, If the vibration distortion index is greater than or equal to the preset vibration distortion index, the potential fault location module determines that the dry-type transformer has a potential fault, and determines the location of the potential fault based on the acquisition location of the vibration signal.

[0011] Furthermore, the preset vibration distortion index is determined based on the statistical characteristics of the historical vibration signals of the dry-type transformer under normal operating conditions.

[0012] Furthermore, the detection path planning module determines the detection path for the fault location based on the potential fault location of the dry-type transformer and its three-dimensional structural model. The detection path planning module performs spatial accessibility analysis on the potential fault location based on the three-dimensional structural model to determine at least one candidate observation point. The detection path planning module plans an initial movement path from the current position to the candidate observation point with the goal of minimizing the path length. The detection path planning module further optimizes the initial movement path based on detection performance constraints to determine the final detection path.

[0013] Furthermore, the path adjustment module determines the fault-affected area at the detection location corresponding to the fault detection result based on the fault detection result, wherein, The path adjustment module contains a pre-stored fault propagation mechanism model, which includes fault propagation rules and influence range determination rules corresponding to different fault types. The path adjustment module extracts fault type parameters and fault severity parameters from the fault detection results; The path adjustment module calculates the fault influence area corresponding to the detection location of the fault detection result based on the fault type parameter and the fault severity parameter, combined with the fault propagation mechanism model and the three-dimensional structural model of the dry-type transformer.

[0014] Furthermore, the path adjustment module, based on the fault type parameter and the fault severity parameter, and in conjunction with the fault propagation mechanism model and the three-dimensional structural model of the dry-type transformer, calculates the fault influence area corresponding to the detection location of the fault detection result, wherein... The fault propagation mechanism model includes an overheating fault propagation sub-model, a discharge fault propagation sub-model, and a mechanical fault propagation sub-model. If the fault type parameter indicates that the fault type is an overheating fault, the path adjustment module calls the overheating fault propagation sub-model, determines the heat source intensity based on the fault severity parameter, and determines the heat-affected area by combining the thermal conduction characteristics and spatial distance of each component in the three-dimensional structural model. If the fault type parameter indicates that the fault type is a discharge fault, the path adjustment module calls the discharge fault propagation sub-model, determines the discharge energy based on the fault severity parameter, and determines the discharge influence area by combining the discharge resistance characteristics and spatial distribution of the insulating material in the three-dimensional structural model. If the fault type parameter indicates that the fault type is a mechanical fault, the path adjustment module calls the mechanical fault propagation sub-model, determines the vibration source intensity based on the fault severity parameter, and determines the mechanical influence area by combining the mechanical connection relationship and vibration transmission characteristics of each component in the three-dimensional structural model.

[0015] Furthermore, the path adjustment module adjusts the subsequent detection path of the fault detection module based on the fault-affected area, wherein, The path adjustment module performs spatial overlay analysis on the fault-affected area and the detection path determined by the detection path planning module; If there is a path segment in the detection path that is located within the fault-affected area, the path adjustment module triggers the detection path planning module to replan the path segment and generate a corrected path that bypasses the fault-affected area. If there is no path segment in the detection path located within the fault-affected area, the path adjustment module maintains the detection path unchanged.

[0016] Furthermore, the path adjustment module adjusts the subsequent detection path of the fault detection module based on the fault-affected area, and also based on the following conditions: If the potential fault location is located within the fault influence area, the path adjustment module cancels the independent detection of the potential fault location and marks the potential fault location as being covered by the current fault; If the fault-affected area covers multiple locations to be detected, the path adjustment module merges the detection tasks of the multiple locations to be detected and performs a comprehensive detection at the boundary of the fault-affected area.

[0017] Compared with the prior art, the beneficial effects of the present invention are that the load type determination module can accurately identify the low load operating window of the dry-type transformer, eliminate misjudgments caused by instantaneous load fluctuations, and ensure that the potential fault location module is triggered only under ideal conditions where the core vibration signal mainly comes from the magnetostrictive effect. This effectively avoids the interference of the winding vibration component on the core state analysis under high load, provides reliable triggering conditions for the accurate calculation of the core vibration distortion index and potential fault location, and significantly improves the signal-to-noise ratio and accuracy of core fault detection.

[0018] Furthermore, the potential fault location module can accurately extract the spectral features reflecting the magnetostrictive nonlinear characteristics of the iron core from the core vibration signal under low load conditions. By calculating the ratio of weighted harmonic analysis to the fundamental frequency, it obtains the vibration distortion index that quantitatively characterizes the health status of the iron core, providing a reliable quantitative basis for subsequent determination of potential fault locations.

[0019] Furthermore, the potential fault location module can accurately identify and precisely locate potential faults in the iron core based on the spectral characteristics of the core vibration signal under low load conditions, providing clear target location information for subsequent detection path planning, and effectively solving the technical problem of weak fault signals, difficulty in identification and location under low load conditions in traditional methods.

[0020] Furthermore, the detection path planning module can plan the optimal detection path for the fault detection equipment that meets both the detection efficiency requirements and the shortest path, while ensuring the detection effect. This effectively solves the problem of low detection efficiency caused by blind detection or unreasonable paths, and provides accurate navigation guidance for subsequent refined fault detection.

[0021] Furthermore, the path adjustment module can scientifically and accurately calculate the spatial range of fault impact based on different fault types and severity, combined with the three-dimensional structural characteristics and physical propagation laws of the equipment. This provides a quantitative basis for the dynamic adjustment of the subsequent detection path, effectively preventing the detection equipment from entering dangerous areas or repeatedly detecting areas already affected by faults, thereby improving the safety of the detection process and the utilization efficiency of detection resources.

[0022] Furthermore, the path adjustment module can dynamically optimize subsequent detection paths and detection tasks based on real-time detected faults and their impact range. This ensures the safety of the detection process (avoiding entry into fault-affected areas) and avoids the waste of detection resources (canceling duplicate detections and merging related tasks). It realizes intelligent and adaptive adjustment of detection tasks, significantly improving the operating efficiency and intelligence level of the entire detection system. Attached Figure Description

[0023] Figure 1This is a schematic diagram of the intelligent condition detection system for dry-type transformers in this embodiment; Figure 2 This is a flowchart illustrating the workflow of the load type determination module in the intelligent condition detection system for dry-type transformers in this embodiment. Figure 3 This is a flowchart illustrating the process of the potential fault location module in the intelligent condition detection system for dry-type transformers in this embodiment. Detailed Implementation

[0024] To make the objectives and advantages of the present invention clearer, the present invention will be further described below with reference to embodiments; it should be understood that the specific embodiments described herein are merely for explaining the present invention and are not intended to limit the present invention.

[0025] Preferred embodiments of the present invention will now be described with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are merely illustrative of the technical principles of the present invention and are not intended to limit the scope of protection of the present invention.

[0026] Please see Figures 1-3 As shown, Figure 1 This is a schematic diagram of the intelligent condition detection system for dry-type transformers in this embodiment; Figure 2 This is a flowchart illustrating the workflow of the load type determination module in the intelligent condition detection system for dry-type transformers in this embodiment. Figure 3 This is a flowchart illustrating the process of the potential fault location module in the intelligent condition detection system for dry-type transformers in this embodiment.

[0027] This embodiment provides an intelligent condition detection system for dry-type transformers, including: The data acquisition module is used to obtain the environmental parameters and operating parameters of the dry-type transformer. A load type determination module, connected to the data module, is used to determine the operating load type of the dry-type transformer based on the operating parameters, wherein the load type includes a high load type and a low load type; A potential fault location module is connected to the data acquisition module and the load type determination module respectively. In response to the dry-type transformer operating load type being low load type, it acquires the vibration signal of the iron core in the dry-type transformer from the operating parameters, determines the vibration distortion index of the dry-type transformer based on the vibration signal, and locates the potential fault location of the dry-type transformer based on the vibration distortion index. A detection path planning module, which is connected to the potential fault location module, is used to determine the detection path for the fault location based on the potential fault location of the dry-type transformer and the three-dimensional structural model of the dry-type transformer. A fault detection module, which is connected to the detection path planning module, is used to perform fault detection on the dry-type transformer based on the detection path and obtain fault detection results. A path adjustment module, which is connected to the detection path planning module and the fault detection module respectively, is used to determine the fault impact area of ​​the detection location corresponding to the fault detection result based on the fault detection result, and adjust the subsequent detection path of the fault detection module based on the fault impact area.

[0028] In this embodiment of the invention, the environmental parameters of the dry-type transformer include, but are not limited to, ambient temperature, ambient humidity, ventilation conditions, and ambient noise. The operating parameters include, but are not limited to, load current, primary side voltage, secondary side voltage, active power, reactive power, power factor, and winding temperature. The fault detection module includes multiple detection units, which are deployed on a mobile detection platform or handheld detection device according to the detection task requirements. Specifically, an infrared thermal imaging detection unit includes an uncooled infrared focal plane detector and an electrically adjustable focusing lens. The operating wavelength is 8-14μm, the temperature resolution is 0.05℃, and the spatial resolution is not less than 640×480 pixels. When the detection device moves along the detection path to the target observation point, the infrared thermal imaging detection unit automatically adjusts the lens focal length according to the shooting angle and distance set by the detection path planning module, performing high-resolution thermal imaging scanning of the potential fault location and its surrounding area. The infrared thermal imaging detection unit extracts temperature distribution data of the scanned area, including the maximum temperature T_max, the average temperature T_avg, and the temperature gradient. The system will output the coordinates of hotspot locations as part of the fault detection results.

[0029] The partial discharge detection unit includes an ultra-high frequency (UHF) sensor and an ultrasonic sensor. The UHF sensor operates in the 300MHz-1.5GHz band and is used to detect electromagnetic wave signals generated by partial discharge; the ultrasonic sensor operates in the 20kHz-200kHz band and is used to detect acoustic wave signals generated by partial discharge. When the detection equipment reaches the target observation point, the partial discharge detection unit activates the acquisition mode and continuously acquires partial discharge signals for at least 50 power frequency cycles. The partial discharge detection unit extracts features from the acquired signals, including discharge amplitude, discharge phase distribution spectrum, discharge count, and discharge type identification results, and outputs this data as part of the fault detection result.

[0030] The ultrasonic imaging detection unit includes a phased array ultrasonic probe and a signal processing module, operating at a frequency of 1MHz-5MHz. When the detection equipment moves to the target observation point, the ultrasonic imaging detection unit emits an ultrasonic beam towards the potential fault location and receives the reflected echo. Based on the received echo signal, the ultrasonic imaging detection unit performs imaging processing to generate ultrasonic B-scan or C-scan images of the tested area, used to detect hidden faults such as internal insulation defects, delamination, and air gaps. The ultrasonic imaging detection unit outputs the ultrasonic images and identified defect features as part of the fault detection results.

[0031] The visual inspection unit includes a high-definition industrial camera and a supplementary lighting device, with a resolution of no less than 1920×1080 pixels, and features autofocus and optical zoom. When the inspection equipment moves to the target observation point, the visual inspection unit takes multi-angle photos or videos of the potential fault location to acquire images of the equipment's appearance. The visual inspection unit has a built-in image recognition algorithm that can automatically identify abnormal features such as insulation surface cracks, discharge marks, loose connectors, and foreign object adhesion, and outputs the identification results as part of the fault detection results.

[0032] The fault detection module also includes a multi-sensor fusion processing submodule, connected to each of the aforementioned detection units. This submodule receives raw data and preliminary identification results collected by each detection unit, performs time synchronization and spatial registration, and correlates data acquired by different sensors at the same detection location. The submodule uses DS evidence theory or a Bayesian network to fuse multi-source information, generating a comprehensive fault determination result for the current detection location. This comprehensive fault determination result includes: whether a fault exists, the fault type, the fault severity, the fault location coordinates, and the confidence level. The fault detection module outputs this comprehensive fault determination result as the fault detection result to the path adjustment module. The data structure of the fault detection result includes: detection location coordinates (x, y, z), detection timestamp, fault type label, fault severity level (1-5), confidence level (0-100%), and the associated raw detection data index.

[0033] Specifically, the load type determination module determines the operating load type of the dry-type transformer based on the operating parameters, wherein, The load type determination module determines the real-time load of the dry-type transformer based on the operating parameters; If the real-time load is greater than or equal to the preset real-time load, the load type determination module determines that the operating load type of the dry-type transformer is a high load type; If the real-time load is less than the preset real-time load, the load type determination module determines that the operating load type of the dry-type transformer is the low-load type.

[0034] In an embodiment of the present invention, the load type determination module extracts the effective value of the primary side current of the dry-type transformer from the operating parameters obtained by the data acquisition module, and calculates the real-time load rate in combination with the rated current. Specifically, the real-time load rate L is calculated by the following formula: L = I / I_N × 100%, where I is the current effective value of the primary side current, and I_N is the rated current of the dry-type transformer.

[0035] A load type determination threshold, denoted as L_th, is preset in the load type determination module. The load type determination threshold L_th is determined according to the economic operation range and typical load characteristics of the dry-type transformer, and usually takes a fixed value between 30% and 50%. In this embodiment, preferably, the load type determination threshold L_th takes the value of 30%, which corresponds to the general empirical value of the balance point between the no-load loss and the load loss of the dry-type transformer.

[0036] The load type determination module compares the calculated real-time load rate L with the load type determination threshold L_th: If the real-time load rate L is greater than or equal to the load type determination threshold L_th, that is, L ≥ L_th, then the load type determination module determines that the operating load type of the current dry-type transformer is the high-load type. The high-load type corresponds to the transformer being in the normal operating load range, where the winding electrodynamic force is significant and the vibration signal contains a large component of winding vibration, which is not suitable for core state analysis. If the real-time load rate L is less than the load type determination threshold L_th, that is, L < L_th, then the load type determination module determines that the operating load type of the current dry-type transformer is the low-load type. The low-load type corresponds to the transformer being in a light-load or no-load state, where the winding current is small and the electrodynamic force can be ignored, and the core vibration signal mainly comes from the magnetostrictive effect, which is an ideal window period for analyzing the core state.

[0037] To further improve the accuracy of load type determination, the load type determination module can also introduce a duration determination condition. Specifically, the load type determination module records the duration T_low during which the real-time load rate L is continuously lower than the load type determination threshold L_th, and presets a minimum stable time threshold T_min. The minimum stable time threshold T_min is determined according to the thermal time constant of the transformer and the load fluctuation characteristics. In this embodiment, it is preferably 10 minutes, which is sufficient to exclude misjudgments caused by instantaneous load fluctuations. Only when L < L_th and T_low ≥ T_min, the load type determination module finally determines it as the low-load type and triggers the potential fault location module to start core state analysis.

[0038] Through the above technical solution, the load type determination module can accurately identify the low-load operating window of the dry-type transformer, eliminate misjudgments caused by instantaneous load fluctuations, and ensure that the potential fault location module is triggered only under ideal conditions where the core vibration signal mainly originates from the magnetostrictive effect. This effectively avoids interference from the winding vibration component under high load on the core state analysis, provides reliable triggering conditions for the accurate calculation of the core vibration distortion index and potential fault location, and significantly improves the signal-to-noise ratio and accuracy of core fault detection.

[0039] Specifically, the potential fault location module determines the vibration distortion index of the dry-type transformer based on the vibration signal, wherein, The potential fault location module performs spectral analysis on the vibration signal and extracts the amplitude of the fundamental frequency component and the amplitude of the harmonic component. The vibration distortion index is determined based on the ratio of the harmonic component amplitude to the fundamental frequency component amplitude.

[0040] In this embodiment of the invention, after receiving a low-load type signal triggered by the load type determination module, the potential fault location module acquires vibration signals collected by a vibration sensor array arranged on the surface of the dry-type transformer core from the data acquisition module. The vibration sensor array includes multiple triaxial accelerometers, respectively installed at key locations on the upper clamp, lower clamp, and core column surfaces of the core, for collecting vibration data in the x, y, and z directions. The sampling frequency is set to no less than 10kHz to meet the requirements for collecting high-frequency harmonic components.

[0041] The potential fault location module preprocesses the acquired vibration signal, including removing the DC component, windowing filtering, and eliminating trend terms, to improve the accuracy of the spectrum analysis. Subsequently, the potential fault location module performs a Fast Fourier Transform on the preprocessed vibration signal, converting the time-domain signal to the frequency domain to obtain the spectral distribution of the vibration signal.

[0042] The potential fault location module extracts the fundamental frequency component amplitude and harmonic component amplitude from the spectrum. The fundamental frequency component corresponds to the operating frequency of the dry-type transformer. Considering that dry-type transformers typically operate in a 50Hz power frequency system, the fundamental frequency of the magnetostriction effect in the iron core is 100Hz (i.e., twice the power supply frequency). Therefore, the fundamental frequency component amplitude A1 is the amplitude corresponding to the 100Hz frequency point. The harmonic components include at least two higher-order harmonic components, specifically integer multiples of 100Hz, including the 2nd harmonic amplitude A2 corresponding to 200Hz, the 3rd harmonic amplitude A3 corresponding to 300Hz, and the 4th harmonic amplitude A4 corresponding to 400Hz.

[0043] The potential fault location module calculates the vibration distortion index based on the extracted fundamental frequency component amplitude and harmonic component amplitude. In one specific embodiment, the vibration distortion index D_v is calculated according to the following formula: ; Where A1 is the amplitude of the fundamental frequency component, A_i is the amplitude of the i-th harmonic component, w_i is the weighting coefficient of the i-th harmonic, and n is the selected highest harmonic order. The weighting coefficient w_i is pre-calibrated based on the magnetostrictive nonlinear characteristics of the iron core and the sensitivity of different types of faults to each harmonic. In this embodiment, preferably, the 2nd to 5th harmonics (i.e., 200Hz, 300Hz, 400Hz, and 500Hz) are selected, and the weighting coefficients are set to w2=1.0, w3=0.8, w4=0.6, and w5=0.4, respectively, to reflect the dominant role of low-frequency harmonics on the state of the iron core.

[0044] To ensure the stability and reliability of the vibration distortion index calculation, the potential fault location module also introduces a time averaging mechanism. Specifically, the potential fault location module continuously collects vibration signals for multiple time windows, each with a length of 1 second, calculates the vibration distortion index D_v(t) for each time window, and then calculates the average value over a preset time period as the final vibration distortion index. The preset time period is determined based on the stability of the vibration signal; in this embodiment, it is preferably 10 seconds, i.e., the average value of 10 time windows is taken as the output.

[0045] Through the above technical solution, the potential fault location module can accurately extract the spectral features reflecting the magnetostrictive nonlinear characteristics of the iron core from the core vibration signal under low load conditions. By calculating the ratio of weighted harmonic analysis to the fundamental frequency, the vibration distortion index that quantitatively characterizes the health status of the iron core is obtained, providing a reliable quantitative basis for the subsequent determination of potential fault locations.

[0046] Specifically, the potential fault location module locates the potential fault location of the dry-type transformer based on the vibration distortion index, wherein... If the vibration distortion index is greater than or equal to the preset vibration distortion index, the potential fault location module determines that the dry-type transformer has a potential fault, and determines the location of the potential fault based on the acquisition location of the vibration signal.

[0047] If the vibration distortion index is less than the preset vibration distortion index, the potential fault location module determines that the dry-type transformer does not have a potential fault.

[0048] Specifically, the preset vibration distortion index is determined based on the statistical characteristics of the historical vibration signals of the dry-type transformer under normal operating conditions.

[0049] In an embodiment of the present invention, a preset vibration distortion index, denoted as D_th, is pre-stored in the potential fault location module. The preset vibration distortion index D_th is determined according to the statistical characteristics of historical vibration signals collected when the dry-type transformer is in a healthy state at the time of factory shipment or in the initial operation stage. Specifically, during the healthy operation period after the first operation or major overhaul of the dry-type transformer, vibration data during low-load periods for 30 consecutive days is collected, the average value of the vibration distortion index is calculated for each day, and then the statistical upper limit (such as the 95% quantile) of these 30 average values is taken as the reference value D_base. On this basis, considering measurement errors and safety margins, the preset vibration distortion index D_th = k×D_base is set, where k is a safety factor with a value range of 1.2 to 1.5, and k = 1.3 is preferably selected in this embodiment.

[0050] The potential fault location module compares the calculated current vibration distortion index D_v with the preset vibration distortion index D_th: If the vibration distortion index D_v is greater than or equal to the preset vibration distortion index D_th, that is, D_v≥D_th, the potential fault location module determines that there is a potential fault in the core of the dry-type transformer. At this time, the potential fault location module further determines the potential fault location based on the acquisition position of the vibration signal. Specifically, the data acquisition module includes a plurality of vibration sensors distributed at different positions of the core of the dry-type transformer, and each sensor has a fixed spatial coordinate. The potential fault location module determines the area where the sensor with the vibration distortion index D_v exceeding the preset threshold D_th is located as the potential fault location. When the vibration distortion indices of multiple adjacent sensors all exceed the threshold, the potential fault location module performs spatial interpolation based on the spatial coordinates of each sensor and the corresponding distortion index to generate a spatial distribution cloud map of the vibration distortion index, and determines the central coordinate of the continuous area where the distortion index is greater than or equal to the preset threshold as the potential fault location. If the vibration distortion index D_v is less than the preset vibration distortion index D_th, that is, D_v<D_th, the potential fault location module determines that there is no potential fault in the core of the dry-type transformer, records this result in the system log, and waits for the next low-load window period to perform monitoring again.

[0051] Through the above technical solution, the potential fault location module can accurately identify and precisely locate potential faults in the core based on the spectral characteristics of the core vibration signal under low-load conditions, provide clear target position information for subsequent detection path planning, and effectively solve the technical problems that the fault signal is weak, difficult to identify and locate under low load in traditional methods.

[0052] Specifically, the detection path planning module determines a detection path for the fault location based on the potential fault location of the dry-type transformer and the three-dimensional structure model of the dry-type transformer, where The detection path planning module performs spatial accessibility analysis on the potential fault location based on the three-dimensional structural model to determine at least one candidate observation point. The detection path planning module plans an initial movement path from the current position to the candidate observation point with the goal of minimizing the path length. The detection path planning module further optimizes the initial movement path based on detection performance constraints to determine the final detection path.

[0053] In this embodiment of the invention, the detection path planning module first obtains the spatial coordinates P_fault(x,y,z) of the potential fault location from the potential fault location module, and then reads a pre-built three-dimensional structural model of the dry-type transformer from the system database. The three-dimensional structural model includes the geometric dimensions, spatial coordinates, material properties, and interconnections of all components such as the core, windings, clamps, insulators, connecting bars, and shell; the model's accuracy reaches the millimeter level.

[0054] The detection path planning module performs spatial accessibility analysis on the potential fault location based on the three-dimensional structural model to determine at least one candidate observation point. Specifically, the detection path planning module constructs a spherical search space centered on the potential fault location P_fault with a preset search radius R (preferably R = 1.5 meters in this embodiment). Within this search space, all unobstructed accessible spatial areas are identified based on the three-dimensional structural model, and locations obstructed by other components or with insufficient safe distance from energized components are eliminated. The remaining accessible spatial areas are discretized into grid points, with each grid point serving as a candidate observation point. The candidate observation point must meet the following basic conditions: unobstructed line of sight to the potential fault location, a safe distance from energized components (determined according to the voltage level of the dry-type transformer; in this embodiment, the safe distance for a 10kV dry-type transformer is not less than 0.7 meters), and physical accessibility for the detection equipment.

[0055] The detection path planning module aims to minimize the path length by planning initial movement paths from the current position P_current of the detection device to each candidate observation point. The module employs the A algorithm for path search, comprehensively considering obstacle distribution and safety distance constraints in three-dimensional space to generate the shortest feasible path from the current position to each candidate observation point, and records the corresponding path length L_path. The heuristic function of the A algorithm uses Euclidean distance to ensure search efficiency and path optimality.

[0056] The detection path planning module further optimizes the initial movement path based on detection performance constraints to determine the final detection path. The detection performance constraints are determined according to the detection methods used by the fault detection module, and in this embodiment, they include two main constraints: The first constraint is visibility. When the fault detection module mainly uses infrared thermal imaging or visual inspection, the visibility of the observation point to the potential fault location is crucial. The detection path planning module calculates the visibility coefficient V_ij of the potential fault location observed from each candidate observation point based on the three-dimensional structural model. The visibility coefficient comprehensively considers factors such as whether the line of sight is obstructed, the angle between the observation angle and the normal of the fault surface, and the effective field of view coverage. The visibility coefficient V_ij ranges from 0 to 1, with a larger value indicating a better observation effect. The detection path planning module selects candidate observation points with a visibility coefficient V_ij greater than a preset visibility threshold V_th (V_th=0.8 in this embodiment) as valid observation points.

[0057] Secondly, there is the signal attenuation constraint. When the fault detection module mainly uses partial discharge detection or ultrasonic detection, the signal propagation distance has a significant impact on the detection sensitivity. The detection path planning module calculates the signal propagation path attenuation value A_signal from the potential fault location to each candidate observation point based on the three-dimensional structural model and material properties. The signal attenuation value comprehensively considers air propagation attenuation and obstacle penetration attenuation. The detection path planning module selects candidate observation points whose signal propagation path attenuation value A_signal is less than a preset attenuation threshold A_th (A_th = 20dB in this embodiment) as valid observation points.

[0058] After determining the set of valid observation points, the detection path planning module selects the observation point with the shortest path length from the valid observation points as the final detection position P_target, based on the principle of minimizing path length. If the set of valid observation points is empty, the detection performance constraints are relaxed (e.g., the visibility threshold is lowered or the attenuation threshold is increased), and the selection is re-screened until at least one valid observation point is obtained.

[0059] The detection path planning module generates a final detection path from the current position P_current to P_target with P_target as the endpoint, and outputs the path as a sequence of coordinate points to the fault detection module for execution.

[0060] Through the above technical solution, the detection path planning module can plan the optimal detection path for the fault detection equipment that meets the detection efficiency requirements and has the shortest path, while ensuring the detection effect. This effectively solves the problem of low detection efficiency caused by blind detection or unreasonable paths, and provides accurate navigation guidance for subsequent refined fault detection.

[0061] Specifically, the path adjustment module determines the fault-affected area at the detection location corresponding to the fault detection result based on the fault detection result, wherein, The path adjustment module contains a pre-stored fault propagation mechanism model, which includes fault propagation rules and influence range determination rules corresponding to different fault types. The path adjustment module extracts fault type parameters and fault severity parameters from the fault detection results; The path adjustment module calculates the fault influence area corresponding to the detection location of the fault detection result based on the fault type parameter and the fault severity parameter, combined with the fault propagation mechanism model and the three-dimensional structural model of the dry-type transformer.

[0062] Specifically, the path adjustment module, based on the fault type parameter and the fault severity parameter, and in conjunction with the fault propagation mechanism model and the three-dimensional structural model of the dry-type transformer, calculates the fault influence area corresponding to the detection location of the fault detection result, wherein... The fault propagation mechanism model includes an overheating fault propagation sub-model, a discharge fault propagation sub-model, and a mechanical fault propagation sub-model. If the fault type parameter indicates that the fault type is an overheating fault, the path adjustment module calls the overheating fault propagation sub-model, determines the heat source intensity based on the fault severity parameter, and determines the heat-affected area by combining the thermal conduction characteristics and spatial distance of each component in the three-dimensional structural model. If the fault type parameter indicates that the fault type is a discharge fault, the path adjustment module calls the discharge fault propagation sub-model, determines the discharge energy based on the fault severity parameter, and determines the discharge influence area by combining the discharge resistance characteristics and spatial distribution of the insulating material in the three-dimensional structural model. If the fault type parameter indicates that the fault type is a mechanical fault, the path adjustment module calls the mechanical fault propagation sub-model, determines the vibration source intensity based on the fault severity parameter, and determines the mechanical influence area by combining the mechanical connection relationship and vibration transmission characteristics of each component in the three-dimensional structural model.

[0063] In this embodiment of the invention, the path adjustment module extracts fault type parameters and fault severity parameters from the fault detection results. The fault type parameter indicates the specific type of fault detected at the current detection location, including three main categories: overheating faults, discharge faults, and mechanical faults. The fault severity parameter is a quantitative indicator with a value range of 1 to 5 levels, corresponding to five levels: minor, moderate, severe, and critical.

[0064] The path adjustment module contains a pre-stored fault propagation mechanism model. This model is pre-established based on the physical propagation laws of various faults in dry-type transformers and is verified and calibrated through experimental data and simulation analysis. The fault propagation mechanism model includes overheating fault propagation sub-models, discharge fault propagation sub-models, and mechanical fault propagation sub-models, each corresponding to the propagation characteristics of different fault types.

[0065] If the fault type parameter indicates an overheating fault, the path adjustment module calls the overheating fault propagation sub-model. This sub-model is based on the principles of thermal conductivity physics and includes a mapping relationship between heat source intensity and fault severity, a material thermal conductivity parameter library, and rules for calculating heat diffusion in three-dimensional space. The path adjustment module first determines the heat source intensity Q_heat based on the fault severity parameter, with severity levels 1 to 5 corresponding to heat source intensities of 50W, 100W, 200W, 400W, and 800W, respectively. Then, the path adjustment module reads the thermal conductivity coefficient k_thermal, specific heat capacity c_p, and density ρ of various components (such as the core, windings, insulation components, and clamps) around the fault location from the three-dimensional structural model. Based on the thermal conductivity differential equation and the steady-state thermal diffusion approximation, the spatial region centered on the fault location and whose temperature rise exceeds a certain threshold of the ambient temperature (set to 5℃ in this embodiment) is calculated as the thermally affected region. This region typically appears as an ellipsoid in three-dimensional space, with its major axis extending along the direction of better thermal conductivity and its minor axis perpendicular to the direction of poorer thermal conductivity.

[0066] If the fault type parameter indicates a discharge fault, the path adjustment module invokes the discharge fault propagation sub-model. This sub-model is established based on the correlation between discharge energy and insulation damage range, including a mapping relationship between discharge energy and fault severity, a library of insulation material discharge resistance parameters, and spatial distribution rules for discharge effects. The path adjustment module first determines the discharge energy E_discharge based on the fault severity parameter, with severity levels 1 to 5 corresponding to discharge energies of 100pC, 500pC, 2000pC, 5000pC, and 10000pC, respectively. Then, the path adjustment module reads the discharge resistance characteristics parameters of the insulation material surrounding the fault location from the three-dimensional structural model, including surface flashover distance, insulation thickness, and dielectric strength. Based on the discharge energy decay law and the discharge resistance characteristics of the insulation material, the spatial region where the discharge may propagate along the insulation surface or be conducted through the interior of the insulation material is calculated as the discharge influence region. This region typically exhibits a radial distribution centered on the discharge point in three-dimensional space, extending far along the insulation surface but with limited extension perpendicular to the insulation surface.

[0067] If the fault type parameter indicates a mechanical fault, the path adjustment module calls the mechanical fault propagation sub-model. This sub-model is based on vibration propagation mechanisms and includes a mapping relationship between vibration source intensity and fault severity, a structural vibration transfer function library, and rules for vibration propagation in mechanically connected structures. The path adjustment module first determines the vibration source intensity A_vib based on the fault severity parameter. Severity levels 1 to 5 correspond to vibration source intensities of 0.1g, 0.3g, 0.6g, 1.0g, and 2.0g, respectively. Then, the path adjustment module reads the mechanical connection relationships of components around the fault location from the three-dimensional structural model, including bolted connections, welding, clamping, and other connection methods, as well as the stiffness and damping coefficients of each connection. Based on vibration propagation theory, the spatial region with the fault location as the vibration source and the vibration amplitude attenuating to below a certain threshold (set to 0.05g in this embodiment) is calculated as the mechanical influence region. This region typically propagates along the mechanical connection path in three-dimensional space, propagating further at rigid connections and attenuating faster at flexible connections or locations with higher damping.

[0068] The path adjustment module uniformly represents the calculated thermally affected area, discharge affected area, or mechanically affected area as the fault affected area. This area is stored in the form of a three-dimensional spatial coordinate set, specifically including the coordinate sequence of the area boundary points and a description of the spatial range within the area. The fault affected area will serve as the basis for subsequent detection path adjustments.

[0069] Through the above technical solution, the path adjustment module can scientifically and accurately calculate the spatial range of fault impact based on different fault types and severity, combined with the three-dimensional structural characteristics and physical propagation laws of the equipment. This provides a quantitative basis for the dynamic adjustment of the subsequent detection path, effectively preventing the detection equipment from entering dangerous areas or repeatedly detecting areas already affected by faults, thereby improving the safety of the detection process and the utilization efficiency of detection resources.

[0070] Specifically, the path adjustment module adjusts the subsequent detection path of the fault detection module based on the fault-affected area, wherein, The path adjustment module performs spatial overlay analysis on the fault-affected area and the detection path determined by the detection path planning module; If there is a path segment in the detection path that is located within the fault-affected area, the path adjustment module triggers the detection path planning module to replan the path segment and generate a corrected path that bypasses the fault-affected area. If there is no path segment in the detection path located within the fault-affected area, the path adjustment module maintains the detection path unchanged.

[0071] Specifically, the path adjustment module adjusts the subsequent detection path of the fault detection module based on the fault-affected area, and also based on the following conditions: If the potential fault location is located within the fault influence area, the path adjustment module cancels the independent detection of the potential fault location and marks the potential fault location as being covered by the current fault; If the fault-affected area covers multiple locations to be detected, the path adjustment module merges the detection tasks of the multiple locations to be detected and performs a comprehensive detection at the boundary of the fault-affected area.

[0072] In this embodiment of the invention, after calculating the fault impact area corresponding to the current detection position, the path adjustment module immediately initiates dynamic adjustment of the subsequent detection path. This adjustment process is based on spatial overlay analysis and a multi-condition decision-making mechanism to ensure the safety and efficiency of the detection task.

[0073] First, the path adjustment module performs spatial overlay analysis on the calculated fault impact area and the detection path pre-determined by the detection path planning module. The detection path is represented by a series of ordered spatial coordinate points, each corresponding to a position that the detection equipment needs to traverse. The fault impact area is represented in the form of a three-dimensional spatial region, including boundary coordinates and internal spatial range. The path adjustment module uses a spatial intersection algorithm to determine whether each path segment on the detection path intersects with the fault impact area.

[0074] If any path segment in the detection path is located within the fault-affected area (i.e., one or more coordinate points on the path fall within the fault-affected area), the path adjustment module determines that these path segments pose a safety risk or cause detection interference. In this case, the path adjustment module immediately triggers the detection path planning module to perform local replanning on these path segments. The detection path planning module uses the current real-time location of the detection equipment as the starting point and the first path point in the original path not covered by the fault-affected area as the target. Under the constraint of avoiding the fault-affected area, it re-searches the path and generates a corrected path that bypasses the fault-affected area. The corrected path, while meeting safety distance requirements, maintains a similar orientation to the original path as much as possible to minimize the impact of path changes on the overall detection plan. The detection path planning module then sends the generated corrected path to the fault detection module, replacing the affected portion of the original path for continued execution.

[0075] If there is no path segment within the fault-affected area in the detection path, meaning the entire subsequent detection path is outside the fault-affected area, the path adjustment module determines that the current fault does not affect the safety of subsequent detection, maintains the original detection path, and the fault detection module continues to execute the detection task according to the original plan.

[0076] Secondly, the path adjustment module also optimizes and adjusts the content of subsequent detection tasks based on the following conditions: If the potential fault location is within the fault's influence area, meaning the potential fault location initially located by the potential fault location module is already covered by a currently detected fault, the path adjustment module determines that independent detection of that potential fault location is unnecessary. In this case, the path adjustment module cancels the originally planned independent detection task for that potential fault location and marks it as "covered by current fault" in the system task list. Simultaneously, the path adjustment module synchronizes this information to the fault detection module, preventing the detection equipment from making an empty trip to that location. This mechanism effectively avoids wasting detection resources and prevents repeated detection of the same fault area.

[0077] If the fault-affected area covers multiple locations to be detected, meaning the current fault impact is significant and affects multiple target locations in the subsequent detection plan, the path adjustment module initiates a detection task merging mechanism. The path adjustment module identifies all locations to be detected within the fault-affected area and merges these scattered detection points into a single comprehensive detection task. The path adjustment module then triggers the detection path planning module to replan a path directly to the boundary of the fault-affected area and selects an optimal observation point on the boundary. At this optimal observation point, a comprehensive detection covering the entire fault-affected area is performed. This comprehensive detection uses wide-angle scanning or panoramic imaging to acquire the state information of the entire affected area in one go, replacing the original multi-point scattered detection.

[0078] Through the above technical solution, the path adjustment module can dynamically optimize subsequent detection paths and detection tasks based on the faults detected in real time and their impact range. This ensures the safety of the detection process (avoiding entry into the fault-affected area) and avoids the waste of detection resources (canceling duplicate detections and merging related tasks). It realizes intelligent and adaptive adjustment of detection tasks, significantly improving the operating efficiency and intelligence level of the entire detection system.

[0079] The technical solution of the present invention has been described above with reference to the preferred embodiments shown in the accompanying drawings. However, it will be readily understood by those skilled in the art that the scope of protection of the present invention is obviously not limited to these specific embodiments. Without departing from the principles of the present invention, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of the present invention.

Claims

1. An intelligent condition detection system for dry-type transformers, characterized in that, include: The data acquisition module is used to obtain the environmental parameters and operating parameters of the dry-type transformer. A load type determination module, connected to the data module, is used to determine the operating load type of the dry-type transformer based on the operating parameters, wherein the load type includes a high load type and a low load type; A potential fault location module is connected to the data acquisition module and the load type determination module respectively. In response to the dry-type transformer operating load type being low load type, it acquires the vibration signal of the iron core in the dry-type transformer from the operating parameters, determines the vibration distortion index of the dry-type transformer based on the vibration signal, and locates the potential fault location of the dry-type transformer based on the vibration distortion index. A detection path planning module, which is connected to the potential fault location module, is used to determine the detection path for the fault location based on the potential fault location of the dry-type transformer and the three-dimensional structural model of the dry-type transformer. A fault detection module, which is connected to the detection path planning module, is used to perform fault detection on the dry-type transformer based on the detection path and obtain fault detection results. A path adjustment module, which is connected to the detection path planning module and the fault detection module respectively, is used to determine the fault impact area of ​​the detection location corresponding to the fault detection result based on the fault detection result, and adjust the subsequent detection path of the fault detection module based on the fault impact area.

2. The intelligent condition detection system for dry-type transformers according to claim 1, characterized in that, The load type determination module determines the operating load type of the dry-type transformer based on the operating parameters, wherein, The load type determination module determines the real-time load of the dry-type transformer based on the operating parameters; If the real-time load is greater than or equal to the preset real-time load, the load type determination module determines that the operating load type of the dry-type transformer is a high load type; If the real-time load is less than the preset real-time load, the load type determination module determines that the operating load type of the dry-type transformer is a low load type.

3. The intelligent condition detection system for dry-type transformers according to claim 1, characterized in that, The potential fault location module determines the vibration distortion index of the dry-type transformer based on the vibration signal, wherein, The potential fault location module performs spectral analysis on the vibration signal and extracts the amplitude of the fundamental frequency component and the amplitude of the harmonic component. The vibration distortion index is determined based on the ratio of the harmonic component amplitude to the fundamental frequency component amplitude.

4. The intelligent condition detection system for dry-type transformers according to claim 1, characterized in that, The potential fault location module locates the potential fault location of the dry-type transformer based on the vibration distortion index, wherein... If the vibration distortion index is greater than or equal to the preset vibration distortion index, the potential fault location module determines that the dry-type transformer has a potential fault, and determines the location of the potential fault based on the acquisition location of the vibration signal.

5. The intelligent condition detection system for dry-type transformers according to claim 4, characterized in that, The preset vibration distortion index is determined based on the statistical characteristics of the historical vibration signals of the dry-type transformer under normal operating conditions.

6. The intelligent condition detection system for dry-type transformers according to claim 1, characterized in that, The detection path planning module determines the detection path for the fault location based on the potential fault location of the dry-type transformer and its three-dimensional structural model. The detection path planning module performs spatial accessibility analysis on the potential fault location based on the three-dimensional structural model to determine at least one candidate observation point. The detection path planning module plans an initial movement path from the current position to the candidate observation point with the goal of minimizing the path length. The detection path planning module further optimizes the initial movement path based on detection performance constraints to determine the final detection path.

7. The intelligent condition detection system for dry-type transformers according to claim 1, characterized in that, The path adjustment module determines the fault-affected area corresponding to the detection location based on the fault detection result, wherein, The path adjustment module contains a pre-stored fault propagation mechanism model, which includes fault propagation rules and influence range determination rules corresponding to different fault types. The path adjustment module extracts fault type parameters and fault severity parameters from the fault detection results; The path adjustment module calculates the fault influence area corresponding to the detection location of the fault detection result based on the fault type parameter and the fault severity parameter, combined with the fault propagation mechanism model and the three-dimensional structural model of the dry-type transformer.

8. The intelligent condition detection system for dry-type transformers according to claim 7, characterized in that, The path adjustment module, based on the fault type parameter and the fault severity parameter, combined with the fault propagation mechanism model and the three-dimensional structural model of the dry-type transformer, calculates the fault influence area corresponding to the detection location of the fault detection result, wherein... The fault propagation mechanism model includes an overheating fault propagation sub-model, a discharge fault propagation sub-model, and a mechanical fault propagation sub-model. If the fault type parameter indicates that the fault type is an overheating fault, the path adjustment module calls the overheating fault propagation sub-model, determines the heat source intensity based on the fault severity parameter, and determines the heat-affected area by combining the thermal conduction characteristics and spatial distance of each component in the three-dimensional structural model. If the fault type parameter indicates that the fault type is a discharge fault, the path adjustment module calls the discharge fault propagation sub-model, determines the discharge energy based on the fault severity parameter, and determines the discharge influence area by combining the discharge resistance characteristics and spatial distribution of the insulating material in the three-dimensional structural model. If the fault type parameter indicates that the fault type is a mechanical fault, the path adjustment module calls the mechanical fault propagation sub-model, determines the vibration source intensity based on the fault severity parameter, and determines the mechanical influence area by combining the mechanical connection relationship and vibration transmission characteristics of each component in the three-dimensional structural model.

9. The intelligent condition detection system for dry-type transformers according to claim 1, characterized in that, The path adjustment module adjusts the subsequent detection path of the fault detection module based on the fault-affected area, wherein, The path adjustment module performs spatial overlay analysis on the fault-affected area and the detection path determined by the detection path planning module; If there is a path segment in the detection path that is located within the fault-affected area, the path adjustment module triggers the detection path planning module to replan the path segment and generate a corrected path that bypasses the fault-affected area. If there is no path segment in the detection path located within the fault-affected area, the path adjustment module maintains the detection path unchanged.

10. The intelligent condition detection system for dry-type transformers according to claim 9, characterized in that, The path adjustment module adjusts the subsequent detection path of the fault detection module based on the fault-affected area, and also based on the following conditions: If the potential fault location is located within the fault influence area, the path adjustment module cancels the independent detection of the potential fault location and marks the potential fault location as being covered by the current fault; If the fault-affected area covers multiple locations to be detected, the path adjustment module merges the detection tasks of the multiple locations to be detected and performs comprehensive detection at the boundary location of the fault-affected area.

Images