Power transmission and transformation equipment operation state dynamic monitoring method and system based on photoelectric detection

Abstract

The invention provides a power transmission and transformation equipment operation state dynamic monitoring method and system based on photoelectric detection, and relates to the technical field of power transmission and transformation equipment. Matching inductance parameters, matching capacitance parameters, original standardized impedance sensing signals and original gain adaptive distribution signals which are collected, preprocessed and fed back by the multispectral sensing unit are fused, associated fault scene diagnosis is executed, and a real-time equipment state is output; and performing insulation failure risk protection prediction according to the real-time equipment state, and outputting a graded early warning instruction. The technical problems that in an existing power transmission and transformation equipment state monitoring technology, single-dimensional perception sensing is adopted for isolated monitoring analysis, cross-equipment fault conduction verification is lacked, and consequently early equipment fault diagnosis is delayed, and the confidence coefficient is low are solved. The technical effects of remarkably improving the fault diagnosis confidence and the early warning timeliness, effectively avoiding the non-planned shutdown and prolonging the service life of the key equipment are achieved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power transmission and transformation equipment, in particular to a power transmission and transformation equipment operation state dynamic monitoring method and system based on photoelectric detection. BACKGROUND

[0002] The existing power transmission and transformation equipment state monitoring technology uses single-dimensional perception sensors for isolated monitoring and analysis, such as relying only on local discharge detection or infrared temperature measurement, and cannot simultaneously capture the electric-thermal-mechanical multi-physical field coupling effect, resulting in early weak characteristics of insulation deterioration being submerged by noise.

[0003] At the same time, the modeling of electrical correlation across devices and the fault conduction verification mechanism are missing, which cannot trace the propagation path of the fault among associated devices such as transformers, bushings, and busbars; the above defects further cause diagnostic delay and confidence crisis, forcing maintenance personnel to manually review when facing alarms, which significantly increases the risk of unplanned shutdown and equipment life loss.

[0004] In summary, the existing power transmission and transformation equipment state monitoring technology has the technical problem of using single-dimensional perception sensors for isolated monitoring and analysis and lacking cross-device fault conduction verification, resulting in delayed diagnosis of early device faults and low confidence. SUMMARY

[0005] The present application provides a power transmission and transformation equipment operation state dynamic monitoring method and system based on photoelectric detection, which is used to solve the technical problem of the existing power transmission and transformation equipment state monitoring technology using single-dimensional perception sensors for isolated monitoring and analysis and lacking cross-device fault conduction verification, resulting in delayed diagnosis of early device faults and low confidence.

[0006] In view of the above problems, the present application provides a power transmission and transformation equipment operation state dynamic monitoring method and system based on photoelectric detection.

[0007] In a first aspect of the present application, a power transmission and transformation equipment operation state dynamic monitoring method based on photoelectric detection is provided, the method comprising: pre-deploying a multi-spectral sensing unit on the exposed surface of power transmission and transformation equipment, wherein the power transmission and transformation equipment is a transformer bushing, and the multi-spectral sensing unit is composed of an ultraviolet photon counter, an infrared focal plane array, and a fiber grating sensor; in an operation state dynamic monitoring scenario, retrieving matching inductance parameters and matching capacitance parameters from a pre-stored impedance parameter mapping library according to the pulse signal spectrum characteristics output by the ultraviolet photon counter; converting the original high-impedance sensing signal output by the fiber grating sensor into an original standardized impedance sensing signal through an active matching circuit connected to the output end of the fiber grating sensor; performing dynamic gain adjustment on the original radiation signal output by the infrared focal plane array through a gain segmentation controller connected to the infrared focal plane array, and outputting an original gain adaptation electrical signal; fusing the matching inductance parameters, the matching capacitance parameters, the original standardized impedance sensing signal, and the original gain adaptation electrical signal, performing associated fault scenario diagnosis, and outputting a real-time equipment state; and performing insulation failure risk protection prediction according to the real-time equipment state, and outputting a hierarchical early warning instruction.

[0008] In a second aspect of the present application, a power transmission and transformation equipment operation state dynamic monitoring system based on photoelectric detection is provided, the system comprising: a sensing unit deployment module for pre-deploying a multi-spectral sensing unit on the exposed surface of power transmission and transformation equipment, wherein the power transmission and transformation equipment is a transformer bushing, and the multi-spectral sensing unit is composed of an ultraviolet photon counter, an infrared focal plane array, and a fiber grating sensor; a spectrum characteristic matching module for retrieving matching inductance parameters and matching capacitance parameters from a pre-stored impedance parameter mapping library according to the pulse signal spectrum characteristics output by the ultraviolet photon counter in an operation state dynamic monitoring scenario; a signal conversion execution module for converting the original high-impedance sensing signal output by the fiber grating sensor into an original standardized impedance sensing signal through an active matching circuit connected to the output end of the fiber grating sensor; a gain adjustment execution module for performing dynamic gain adjustment on the original radiation signal output by the infrared focal plane array through a gain segmentation controller connected to the infrared focal plane array, and outputting an original gain adaptation electrical signal; a fault association diagnosis module for fusing the matching inductance parameters, the matching capacitance parameters, the original standardized impedance sensing signal, and the original gain adaptation electrical signal, performing associated fault scenario diagnosis, and outputting a real-time equipment state; and a hierarchical prediction analysis module for performing insulation failure risk protection prediction according to the real-time equipment state, and outputting a hierarchical early warning instruction.

[0009] The one or more technical solutions provided in the present application have at least the following technical effects or advantages: The method provided in this application pre-deploys a multispectral sensing unit on the exposed surface of a power transmission and transformation equipment, wherein the power transmission and transformation equipment is a transformer bushing, and the multispectral sensing unit consists of an ultraviolet photon counter, an infrared focal plane array, and a fiber Bragg grating sensor. In a dynamic monitoring scenario of operating status, based on the spectral characteristics of the pulse signal output by the ultraviolet photon counter, matching inductance parameters and matching capacitance parameters are retrieved from a pre-stored impedance parameter mapping library. Through an active matching circuit connected to the output end of the fiber Bragg grating sensor, the original high-impedance sensing signal output by the fiber Bragg grating sensor is converted into an original standardized impedance sensing signal. Through a gain segmentation controller connected to the infrared focal plane array, dynamic gain adjustment is performed on the original radiation signal output by the infrared focal plane array, and an original gain adaptation electrical signal is output. The matching inductance parameters, matching capacitance parameters, original standardized impedance sensing signal, and original gain adaptation electrical signal are fused to perform associated fault scenario diagnosis and output real-time equipment status. Based on the real-time equipment status, insulation failure risk protection prediction is performed, and graded early warning commands are output. It achieves early and accurate warning of insulation failure risk of power transmission and transformation equipment based on multi-physical quantity fusion perception and cross-equipment status backtracking, significantly improves the confidence of fault diagnosis and the timeliness of warning, and effectively avoids unplanned outages and extends the service life of key equipment. Attached Figure Description

[0010] Figure 1 A schematic diagram of the dynamic monitoring method for the operating status of power transmission and transformation equipment based on photoelectric detection provided in this application; Figure 2 A schematic diagram of the structure of the dynamic monitoring system for the operating status of power transmission and transformation equipment based on photoelectric detection provided in this application.

[0011] Explanation of reference numerals in the attached diagram: Sensing unit deployment module 11, Spectrum feature matching module 12, Signal conversion execution module 13, Gain adjustment execution module 14, Fault association diagnosis module 15, Graded prediction analysis module 16. Detailed Implementation

[0012] This application provides a method and system for dynamic monitoring of the operating status of power transmission and transformation equipment based on photoelectric detection. It addresses the technical problems of existing power transmission and transformation equipment condition monitoring technologies, which rely on isolated monitoring and analysis using single-dimensional sensing and lack cross-equipment fault propagation verification, leading to delayed early fault diagnosis and low confidence levels. The method achieves early and accurate warning of insulation failure risks in power transmission and transformation equipment based on multi-physical quantity fusion sensing and cross-equipment status backtracking. This significantly improves fault diagnosis confidence and warning timeliness, effectively avoiding unplanned outages and extending the service life of critical equipment.

[0013] The technical solutions of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. It should be understood that the present invention is not limited to the exemplary embodiments described herein. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention. It should also be noted that, for ease of description, only the parts related to the present invention are shown in the accompanying drawings, not all of them.

[0014] Example 1, as Figure 1 As shown, this application provides a method for dynamic monitoring of the operating status of power transmission and transformation equipment based on photoelectric detection, the method comprising: A100: A multispectral sensing unit is pre-deployed on the exposed surface of a power transmission and transformation equipment, wherein the power transmission and transformation equipment is a transformer bushing, and the multispectral sensing unit consists of an ultraviolet photon counter, an infrared focal plane array, and a fiber optic grating sensor.

[0015] Specifically, this embodiment integrates a multispectral sensing unit consisting of an ultraviolet photon counter, an infrared focal plane array, and a fiber optic grating sensor directly on the exposed surface of the transformer bushing. By capturing the corona discharge photon flux, temperature field distribution gradient, and mechanical strain physical quantities, it achieves in-situ fusion sensing of multiple physical quantities of the device body, including electrical, thermal, and mechanical properties.

[0016] A200: In the scenario of dynamic monitoring of operating status, based on the spectral characteristics of the pulse signal output by the ultraviolet photon counter, the matching inductor parameters and matching capacitor parameters are retrieved from the pre-stored impedance parameter mapping library.

[0017] Furthermore, based on the spectral characteristics of the pulse signal output by the ultraviolet photon counter, the matching inductor parameters and matching capacitor parameters are retrieved from the pre-stored impedance parameter mapping library. Step A200 of the method provided by this invention further includes: A210: After performing a fast Fourier transform on the original pulse timing signal output by the ultraviolet photon counter, the spectral distribution of a predefined frequency band is extracted, wherein the predefined frequency band is preferably a 1MHz-1GHz band.

[0018] A220: Identify the main peak frequency from the said spectral distribution.

[0019] A230: Using the main peak frequency as the search key, retrieve the matching target parameter group from the impedance parameter mapping library, wherein the target parameter group consists of the matching inductance parameter and the matching capacitance parameter.

[0020] Further, the matching inductance parameter and the matching capacitance parameter are written into the adjustable inductance module and the adjustable capacitance module of the dynamic adjustable Π type matching network through a digital control interface respectively, and the dynamic adjustable Π type matching network arranged at the output end of the ultraviolet photon counter is switched to the target parameter group.

[0021] Specifically, in the process of dynamic monitoring of the operation state of the power transmission and transformation equipment, the matching inductance parameter and the matching capacitance parameter are directly retrieved and output from the pre-constructed impedance parameter mapping library according to the pulse signal spectrum characteristics output by the ultraviolet photon counter.

[0022] First, the original pulse time sequence signal output by the ultraviolet photon counter is subjected to fast Fourier transform, and the energy distribution characteristics in the preset 1MHz to 1GHz frequency band are extracted, and then the main peak frequency point with significant energy aggregation is identified from the above spectrum distribution as the core feature.

[0023] Finally, the identified main peak frequency is used as the only retrieval key to directly match and output the target parameter group from the pre-stored impedance parameter mapping library, and the target parameter group is composed of the matching inductance parameter and the matching capacitance parameter.

[0024] At the same time, it should be understood that the impedance parameter mapping library accumulatively stores the matching inductance and capacitance parameter combinations corresponding to different main peak frequencies.

[0025] The matching inductance parameter and the matching capacitance parameter are respectively written into the adjustable inductance module and the adjustable capacitance module of the dynamic adjustable Π type matching network in real time through a digital control interface, and the Π type matching network arranged at the output end of the ultraviolet photon counter is switched to the target parameter group state, realizing the impedance adaptive matching of the partial discharge pulse signal in the 1MHz-1GHz frequency band with the back-end processing circuit, maximizing the signal transmission energy efficiency and suppressing high-frequency reflection interference.

[0026] A300: The active matching circuit connected through the output end of the fiber grating sensor converts the original high-impedance sensing signal output by the fiber grating sensor into an original standardized impedance sensing signal.

[0027] Specifically, in the embodiment, the active matching circuit is directly connected through the output end of the fiber grating sensor arranged on the surface of the transformer bushing, and the original high-impedance sensing signal of the sensor is converted into an electrical signal with a standard impedance level in real time, eliminating signal reflection loss caused by high-frequency transmission line effect and ensuring that the mechanical strain physical quantity still maintains the integrity of the amplitude-frequency characteristics after long-distance transmission.

[0028] A400: The gain segment controller connected through the infrared focal plane array performs dynamic gain adjustment on the original radiation signal output by the infrared focal plane array, and outputs an original gain adapted electrical signal.

[0029] Further, the gain segment controller connected to the infrared focal plane array performs dynamic gain adjustment on the original radiation signal output by the infrared focal plane array, and outputs an original gain adaptation electrical signal. The method step A400 provided by the application further comprises: A410: performing time domain analysis on the original radiation signal, and outputting a transient signal-to-noise ratio and a peak-to-peak dynamic range.

[0030] A420: defining a gain adjustment evaluation function.

[0031] A430: iteratively adjusting gain parameters by a gradient descent method, so that the change rate of the transient signal-to-noise ratio in P consecutive iteration periods is less than a preset threshold, and the updated peak-to-peak dynamic range adapts to the analog-to-digital converter range, and outputting gain optimization parameters.

[0032] A440: loading the gain optimization parameters into the gain segment controller, performing amplitude scaling on the original radiation signal, and outputting the original gain adaptation electrical signal.

[0033] Further, the analog-to-digital converter is configured between the gain segment controller output end and a measurement device, and the infrared focal plane array, the gain segment controller and the measurement device constitute an infrared radiation signal acquisition and conditioning chain.

[0034] Specifically, the embodiment performs closed-loop dynamic gain adjustment operation on the original radiation electrical signal output by the infrared focal plane array through the gain segment controller physically connected to the output end of the infrared focal plane array, optimizes the signal amplitude in real time according to the back-end analog-to-digital converter range constraint and signal quality requirement, outputs a gain adaptation electrical signal strictly adapted to the input voltage range of the analog-to-digital converter and maintaining the original temperature field distribution characteristics, and ensures the fidelity and availability of infrared thermal image data in the subsequent processing chain.

[0035] First, high-precision time domain waveform analysis is performed on the original radiation signal output by the infrared focal plane array, the transient signal-to-noise ratio and the peak-to-peak dynamic range of the signal in a preset time window are extracted, and quantized evaluation parameters representing signal quality and amplitude characteristics are generated. The transient signal-to-noise ratio is defined as the ratio of the signal effective value to the noise root mean square, and the peak-to-peak dynamic range is defined as the voltage difference between the maximum peak value and the minimum valley value of the signal.

[0036] The pre-defined gain adjustment evaluation function is a multi-objective optimization model, which takes the transient signal-to-noise ratio stability and the peak-to-peak dynamic range adaptation degree as dual optimization objectives, establishes a mathematical mapping relationship between the gain parameters and the signal quality indicators, and provides a quantitative evaluation criterion for gradient descent iteration.

[0037] Preferably, the gain adjustment evaluation function is as follows: ; wherein, is the first transient signal-to-noise ratio change rate in the iteration period, is the current signal peak-to-peak dynamic range measured value, is the analog-to-digital converter range adaptation target value, such as 0-3.3V, and α and β are normalization weight coefficients, with default α=0.7, β=0.3. The weight coefficients here are adjustable.

[0038] The gradient descent method is used to perform multi-period iterative optimization on the gain parameter. In the continuous iteration period, the gain value is dynamically adjusted and the evaluation function output is calculated in real time. When the transient signal-to-noise ratio change rate is lower than the preset threshold (such as 0.5%) and the updated peak-to-peak dynamic range completely falls within the analog-to-digital converter range (such as an error of less than 2%), the iteration is terminated, and the gain optimization parameter meeting the double constraint conditions is output.

[0039] The gain optimization parameter obtained by the gradient descent optimization is loaded into the digital register of the gain segmented controller. The original radiation signal is subjected to linear amplitude scaling transformation through the hardware circuit, and the gain adaptation electrical signal meeting the input level requirements of the analog-to-digital converter and retaining the original temperature field spatial distribution characteristics is output.

[0040] At the same time, it should be understood that the analog-to-digital converter is physically connected between the output end of the gain segmented controller and the input end of the infrared measuring device, and the three constitute an infrared radiation signal acquisition and conditioning chain, wherein the gain segmented controller undertakes the function of dynamic adaptation of signal amplitude, ensuring that the signal input to the analog-to-digital converter is neither saturated nor distorted nor higher than the quantization noise floor.

[0041] The embodiment achieves the technical effect of ensuring real-time adaptive gain control and high-fidelity transmission of infrared radiation signals in a wide dynamic range, and provides high-confidence infrared thermal image data that strictly meets the analog-to-digital converter range constraints, eliminates saturation distortion and quantization noise interference, and completely retains the original temperature field spatial distribution characteristics for subsequent multi-physical quantity fusion fault diagnosis.

[0042] A500: fuse the matching inductance parameter, the matching capacitance parameter, the original standardized impedance sensing signal, and the original gain adaptation electrical signal, perform associated fault scene diagnosis, and output real-time device status.

[0043] Further, the matching inductance parameter, the matching capacitance parameter, the original standardized impedance sensing signal, and the original gain adaptation electrical signal are fused, the associated fault scene diagnosis is performed, and the real-time device status is output. The method steps A500 provided by the present application further comprise: A510: obtaining multiple sets of perception intervals of multiple sample fault states, and performing perception data association rule mining based on the multiple sets of perception intervals of the multiple sample fault states, and constructing a fault scene decision tree.

[0044] A520: mapping the matched inductance parameter, matched capacitance parameter, original standardized impedance sensing signal and original gain adaptation electric signal into a feature vector, loading to the fault scene decision tree to perform fault fusion decision, and outputting the real-time device state.

[0045] A530: if the power transmission and transformation device is in a non-fault state, the real-time device state output is an empty set.

[0046] Further, if the fusion decision confidence of the real-time device state is lower than a confidence threshold, state transition backtracking verification is triggered, and fault attribute time consistency check is performed.

[0047] Further, the method steps provided by the present application further comprise: A5001: if the real-time device state is a fault state, taking the power transmission and transformation device as a fault tracing starting point, performing state backtracking on electrically associated devices according to a preset topology level, and obtaining the operating states of K fault-associated devices.

[0048] A5002: taking the real-time device state and the states of the K fault-associated devices as starting points, performing state transition backtracking to obtain backtracked device states and K associated backtracked states.

[0049] A5003: if the backtracked device states and the real-time device state are consistent in fault attributes, performing insulation failure risk protection prediction according to the real-time device state, and outputting a hierarchical early warning instruction.

[0050] Further, taking the real-time device state and the states of the K fault-associated devices as starting points, performing state transition backtracking to obtain backtracked device states and K associated backtracked states, the method step A5002 provided by the present application further comprises: A5002-1: predefining a first device discrete state set of the power transmission and transformation device.

[0051] A5002-2: predefining K device discrete state sets of the K fault-associated devices.

[0052] A5002-3: taking the first device discrete state set and the K device discrete state sets as a joint state space, and locally calling a first fault time sequence data set and K fault time sequence data sets.

[0053] A5002-4: constructing a cross-device state transition probability matrix based on the first device discrete state set and the K device discrete state sets.

[0054] A5002-5: Perform transition conditional probability calculation on the first fault time series data set and K fault time series data sets, and use the calculation results to fill in the data of the cross-device state transition probability matrix.

[0055] A5002-6: Perform Viterbi backward path search on the cross-device state transition probability matrix along the time axis with the real-time device state and K associated device states as initial states, and output the backtracking device state and K backtracking states.

[0056] Specifically, in the process of dynamic monitoring of the operation state of the power transmission and transformation device, four types of heterogeneous physical characteristics, i.e., matched inductance parameters, matched capacitance parameters, original standardized impedance sensing signals, and original gain adaptation electrical signals, are fused to construct a multi-source data joint analysis framework. Based on the device ontology electric-thermal-mechanical multi-physical quantity coupling characteristics, associated fault scene diagnosis is performed to output real-time device state judgment results covering typical fault types such as insulation degradation, partial discharge, poor contact, and mechanical deformation, thereby forming a complete closed loop from physical signal perception to fault decision.

[0057] The judgment model relied on for the real-time device state judgment is a fault scene decision tree. The construction method is to obtain, through an interactive interface, a plurality of sets of multi-physical quantity sensing data intervals composed of pulse signal spectral features output by an ultraviolet photon counter, temperature field distribution gradients collected by an infrared focal plane array, mechanical strain micro-variables sensed by a fiber grating sensor, and high-frequency impedance response signals processed by a dynamically adjustable Π-type matching network under a plurality of historical sample fault states.

[0058] Based on a hybrid driving mechanism of a statistical learning algorithm and a device physical operation mechanism, deep association rules between different fault types and multi-physical quantity characteristics are deeply mined. For example, an insulation degradation fault is not only associated with a sudden drop of more than 30% in the real part of the impedance in a specific frequency band, but also accompanied by a local temperature rise of more than 15°C and a thermal distribution variance of more than 8°C / cm², and a mechanical deformation micro-strain presents a stepwise increase in the same time interval. Typical sensing interval coupling characteristics are constructed to build a layered fault scene decision tree diagnosis model with physical interpretability as the core, complete fault scene coverage, and transparent diagnosis logic.

[0059] The model root node of the fault scene decision tree is preliminarily classified according to the physical nature of the fault, the intermediate nodes are embedded with multi-source signal time domain and frequency domain association constraint rules, and the leaf nodes output comprehensive diagnosis results fused with historical statistical probability and real-time feature matching degree, thereby ensuring that the fault diagnosis process not only has adaptive optimization capability under data driving, but also strictly follows the inherent law of the device physical failure mechanism, and finally forms a reliable, traceable, and reproducible industrial-level intelligent diagnosis system.

[0060] The matching inductance parameter, which characterizes the high-frequency discharge sensitivity, the matching capacitance parameter, which reflects the low-frequency overheat sensitivity, the original standardized impedance sensing signal that transmits the physical quantity of mechanical strain, and the original gain adaptation signal that retains the original characteristics of the infrared temperature field distribution are four types of heterogeneous signals mapped to a unified dimension feature space. A fused feature vector is formed through feature weighting and standardization.

[0061] The obtained vectors are then loaded into a pre-constructed fault scenario decision tree, and node splitting and rule matching are performed layer by layer. The real-time device status based on the consistency of multiple physical quantities is output, realizing cross-level analysis from signal-level fusion to decision-level diagnosis. The fusion decision confidence of the real-time device status represents the degree of inherent consistency of the multi-source physical feature vectors in the rule matching process of the decision tree, the support strength of the statistical probability of historical fault samples, and the degree of conformity between the real-time status and the physical failure mechanism of the device. Its value comprehensively reflects the credibility of the diagnostic results in two dimensions: data-driven adaptability and physical mechanism reliability.

[0062] If the rule matching results of all branch nodes of the fault scenario decision tree do not reach the preset fault judgment threshold, the power transmission and transformation equipment is clearly determined to be in a non-fault state. At this time, the real-time equipment status output is an empty set symbol rather than a probability value or fuzzy level. The uncertainty expression is eliminated by the discretized output form, providing an absolutely reliable risk-free benchmark for subsequent risk warning.

[0063] If the confidence level of the fusion decision of the real-time device status is lower than the preset threshold (e.g., 85%), it indicates that there is a transient conflict or sensor abnormality among the multi-source features. At this time, the verification process based on the state transition backtracking mechanism is automatically triggered. The historical state sequence is decoded in reverse by the Viterbi algorithm and the consistency of the current fault attribute with the time dimension of the historical state evolution law is verified. For example, it is verified whether the partial discharge fault meets the physical law of continuous evolution from the insulation deterioration state, thereby distinguishing between true faults and transient interference.

[0064] Specifically, if the real-time device status is faulty and its fusion decision confidence is lower than the confidence threshold, the power transmission and transformation equipment will be used as the starting point for fault tracing. According to the predefined electrical connection topology hierarchy, the operating status of other electrically associated devices will be collected level by level to obtain the current operating status data of K fault-related devices, providing a data foundation for subsequent cross-device state transition analysis.

[0065] After obtaining the status of K fault-related devices, the real-time device status and the status of K fault-related devices are used as the starting point for analysis. Based on the state transition relationship between devices, reverse tracing is performed in the time dimension. By analyzing the state evolution pattern, the historical backtracking device status and K related backtracking status are derived, thereby constructing a complete fault development time sequence profile.

[0066] The technical implementation of the device state backtracking is as follows: All discrete state sets that the power transmission and transformation equipment may be in are precisely defined in advance. The first device discrete state set should comprehensively cover various operating state modes of the equipment, including but not limited to normal operating state, insulation deterioration state, partial discharge state, overheating state and other typical fault states, and each state has a clear state identifier and characteristic description.

[0067] K fault-associated equipment are respectively pre-defined with their corresponding device discrete state sets. Each state set should be defined according to the technical characteristics and operating characteristics of the associated equipment to ensure that various operating states of the equipment can be accurately represented. The union set of these state sets constitutes a complete state space for cross-device state analysis.

[0068] The first device discrete state set and the K device discrete state sets are integrated to build a unified joint state space, which can represent the global operating state of the entire electrical association system. The first fault time series data set and the K fault time series data sets are called from the local storage. These data sets contain rich historical operating state records, which provide data support for state transition probability statistics.

[0069] Based on the defined joint state space, a cross-device state transition probability matrix is constructed. The matrix is a high-dimensional probability model, whose dimension is determined by the number of discrete states of all devices. Each element in the matrix represents the transition probability of the power transmission and transformation equipment and its associated equipment from one joint state to another joint state, reflecting the statistical law of state evolution between devices.

[0070] The first fault time series data set and the K fault time series data sets are subjected to in-depth transition condition probability statistical analysis. The transition probability values are calculated by counting the frequency of various state transition events in the historical data, and the statistical results are used to fill the cross-device state transition probability matrix, so that the matrix can accurately reflect the real state transition law.

[0071] With real-time device state and K associated device states as initial state, on the basis of the constructed cross-device state transition probability matrix, start the path search process based on Viterbi algorithm, which traces back the system state evolution trajectory along the time axis frame by frame; the algorithm first initializes the path probability and path record of each state node, sets the initial state probability as the current known state, then iteratively calculates from the latest time to the historical time, in each time step, the algorithm calculates all possible path probabilities from the previous state to the current state, and retains the maximum probability path reaching each state and records the corresponding state sequence index according to the dynamic programming principle; when the algorithm completes the iteration calculation of all time steps, select the state node with the maximum probability value from the final time as the path endpoint, and the complete optimal state sequence can be obtained by backtracking along the recorded state index; finally output the optimal backtracking device state and K associated backtracking state sequences derived based on the maximum a posteriori probability criterion.

[0072] The sequence reflects the most likely state evolution process in the historical time period, providing decision basis for fault diagnosis in the time dimension.

[0073] Extract key fault feature indicators from the optimal backtracking device state and K associated backtracking state sequences, including calculating the duration of specific fault attributes, statistical key state transition node conditional probability, and evaluating the confidence of cross-device fault propagation path, outputting the backtracking device state and K associated backtracking states.

[0074] Compare the backtracking device state and real-time device state for fault attribute consistency, if a high degree of consistency is found in the fault attribute, then make insulation failure risk protection prediction according to the real-time device state, output corresponding level warning instruction according to the risk level, and provide support for operation and maintenance decision.

[0075] This embodiment achieves the construction of a fault diagnosis and warning system based on multi-physical quantity fusion and cross-device state backtracking, significantly improves the accuracy, reliability and decision credibility of power transmission and transformation device state monitoring, realizes early and accurate identification and risk grading warning of insulation failure and other faults, thereby effectively avoiding unplanned device shutdown and major faults, and improving the safety of power grid operation and the intelligent level of operation and maintenance.

[0076] A600: Make insulation failure risk protection prediction according to the real-time device state, and output graded warning instructions.

[0077] Based on the real-time equipment status, combined with the fault type, confidence level, and cross-equipment status backtracking verification results output by its multi-physical quantity fusion diagnosis, the risk of insulation failure of power transmission and transformation equipment such as transformer bushings in the current and future period is comprehensively predicted. Based on the prediction results, a graded early warning instruction including risk level, risk location, risk evolution trend, and disposal suggestions is generated to support operation and maintenance personnel in taking differentiated prevention and control measures, thereby realizing closed-loop management from status monitoring to risk early warning.

[0078] This embodiment achieves early and accurate warning of insulation failure risk in power transmission and transformation equipment based on multi-physical quantity fusion sensing and cross-equipment status backtracking, significantly improving the confidence of fault diagnosis and the timeliness of warning, effectively avoiding unplanned outages and extending the service life of critical equipment.

[0079] Example 2 is based on the same inventive concept as the photoelectric detection-based dynamic monitoring method for the operating status of power transmission and transformation equipment in the previous examples, such as... Figure 2 As shown, this application provides a dynamic monitoring system for the operating status of power transmission and transformation equipment based on photoelectric detection, wherein the system includes: The sensing unit deployment module 11 is used to pre-deploy a multispectral sensing unit on the exposed surface of the power transmission and transformation equipment, wherein the power transmission and transformation equipment is a transformer bushing, and the multispectral sensing unit is composed of an ultraviolet photon counter, an infrared focal plane array, and a fiber optic grating sensor.

[0080] The spectrum feature matching module 12 is used to retrieve and output matching inductor parameters and matching capacitor parameters from a pre-stored impedance parameter mapping library based on the spectrum features of the pulse signal output by the ultraviolet photon counter in a dynamic monitoring scenario of operation status.

[0081] The signal conversion execution module 13 is used to convert the original high-impedance sensing signal output by the fiber Bragg grating sensor into the original standardized impedance sensing signal through the active matching circuit connected to the output end of the fiber Bragg grating sensor.

[0082] The gain adjustment execution module 14 is used to perform dynamic gain adjustment on the original radiation signal output by the infrared focal plane array through the gain segment controller connected to the infrared focal plane array, and output the original gain adaptation electrical signal.

[0083] The fault association diagnosis module 15 is used to integrate the matching inductor parameters, matching capacitor parameters, original standardized impedance sensing signals and original gain adaptation electrical signals, perform associated fault scenario diagnosis, and output real-time equipment status.

[0084] The graded early warning analysis module 16 is used to predict the risk of insulation failure based on the real-time equipment status and output graded early warning commands.

[0085] Further, the fault correlation diagnosis module 15 is further used for: If the real-time device state is a fault state, taking the power transmission and transformation device as a fault tracing starting point, performing state backtracking on the electrical correlation device according to a preset topology level, obtaining the operation state of the K fault correlation devices; taking the real-time device state and the K fault correlation device states as starting points, performing state transition backtracking to obtain the backtracked device state and the K correlation backtracked states; if the backtracked device state and the real-time device state have consistency in fault attributes, performing insulation failure risk protection prediction according to the real-time device state, and outputting a hierarchical early warning instruction.

[0086] Further, the fault correlation diagnosis module 15 is further used for: Predefining a first device discrete state set of the power transmission and transformation device; predefining K device discrete state sets of the K fault correlation devices; taking the first device discrete state set and the K device discrete state sets as a joint state space, locally calling a first fault time series data set and K fault time series data sets; constructing a cross-device state transition probability matrix based on the first device discrete state set and the K device discrete state sets; performing transition conditional probability calculation on the first fault time series data set and the K fault time series data sets, and using the calculation results to perform data filling on the cross-device state transition probability matrix; taking the real-time device state and the K correlation device states as initial states, performing Viterbi reverse path search on the cross-device state transition probability matrix along a time axis, and outputting the backtracked device state and the K correlation backtracked states.

[0087] Further, the gain adjustment execution module 14 is further used for: Performing time domain analysis on the original radiation signal to output a transient signal-to-noise ratio and a peak-to-peak value dynamic range; predefining a gain adjustment evaluation function; iteratively adjusting a gain parameter by a gradient descent method so that a change rate of the transient signal-to-noise ratio in P consecutive iteration periods is less than a preset threshold value, and an updated peak-to-peak value dynamic range adapts to an analog-to-digital converter range, and outputting a gain optimization parameter; loading the gain optimization parameter into the gain segmented controller to perform amplitude scaling on the original radiation signal, and outputting the original gain adaptation electrical signal.

[0088] Further, the spectrum feature matching module 12 is further used for: After performing fast Fourier transform on the original pulse time sequence signal output by the ultraviolet photon counter, extracting a spectrum distribution of a pre-defined frequency band, wherein the pre-defined frequency band is preferably a 1MHz-1GHz frequency band; identifying a main peak frequency from the spectrum distribution; taking the main peak frequency as a retrieval key to retrieve a matching target parameter group from the impedance parameter mapping library, wherein the target parameter group is composed of the matching inductance parameter and the matching capacitance parameter.

[0089] Further, the fault correlation diagnosis module 15 is further used for: interactively obtaining multiple groups of sensing intervals of multiple sample fault states, and performing sensing data correlation rule mining based on the multiple groups of sensing intervals of the multiple sample fault states to construct a fault scene decision tree; mapping the matching inductance parameter, the matching capacitance parameter, the original standardized impedance sensing signal and the original gain adaptation electric signal into a feature vector, loading to the fault scene decision tree to perform fault fusion decision, and outputting the real-time device state; if the power transmission and transformation device is in a non-fault state, the real-time device state output is an empty set.

[0090] Further, the fault correlation diagnosis module 15 is further used for: if the fusion decision confidence of the real-time device state is lower than a confidence threshold, triggering state transition backtracking verification to perform fault attribute time consistency checking.

[0091] Further, the fault correlation diagnosis module 15 is further used for: writing the matching inductance parameter and the matching capacitance parameter into the adjustable inductance module and the adjustable capacitance module of the dynamic adjustable Π type matching network through a digital control interface respectively, and driving the dynamic adjustable Π type matching network configured at the output end of the ultraviolet photon counter to switch to the target parameter group.

[0092] Further, the analog-to-digital converter is configured between the gain segmentation controller output end and the measuring device, and the infrared focal plane array, the gain segmentation controller and the measuring device constitute an infrared radiation signal acquisition and conditioning chain.

[0093] Any one of the above methods or steps can be stored as computer instructions or programs in various types of computer memories, and the computer instructions or programs are recognized by various types of computer processors, thereby realizing any one of the above methods or steps.

[0094] Based on the above specific embodiments of the present application, any improvement and modification of the present application made by those skilled in the art without departing from the principles of the present application shall fall within the scope of the patent protection of the present application.

Claims

1. A method for dynamic monitoring of the operating status of power transmission and transformation equipment based on photoelectric detection, characterized in that, The method includes: A multispectral sensing unit is pre-deployed on the exposed surface of a power transmission and transformation equipment, wherein the power transmission and transformation equipment is a transformer bushing, and the multispectral sensing unit consists of an ultraviolet photon counter, an infrared focal plane array, and a fiber optic grating sensor. In the scenario of dynamic monitoring of operating status, the matching inductor parameters and matching capacitor parameters are retrieved and output from the pre-stored impedance parameter mapping library based on the spectral characteristics of the pulse signal output by the ultraviolet photon counter. The active matching circuit connected to the output end of the fiber Bragg grating sensor converts the original high-impedance sensing signal output by the fiber Bragg grating sensor into the original normalized impedance sensing signal. The gain segment controller connected to the infrared focal plane array performs dynamic gain adjustment on the original radiation signal output by the infrared focal plane array and outputs the original gain adaptation electrical signal. By integrating the matching inductor parameters, matching capacitor parameters, original standardized impedance sensing signals, and original gain adaptation electrical signals, the system performs associated fault scenario diagnosis and outputs real-time device status. Based on the real-time equipment status, insulation failure risk prediction is performed, and graded early warning commands are output.

2. The method for dynamic monitoring of the operating status of power transmission and transformation equipment based on photoelectric detection as described in claim 1, characterized in that, The method further includes: If the real-time device status is faulty, then the power transmission and transformation equipment is taken as the starting point for fault tracing, and the status of the electrical associated equipment is traced back according to the preset topology level to obtain the operating status of K fault-related devices. Starting from the real-time device status and the K fault-related device statuses, state transition backtracking is performed to obtain the backtracked device status and the K associated backtracked states; If the retrospective device status and the real-time device status are consistent in terms of fault attributes, then insulation failure risk protection prediction is performed based on the real-time device status, and graded early warning instructions are output.

3. The method for dynamic monitoring of the operating status of power transmission and transformation equipment based on photoelectric detection as described in claim 2, characterized in that, Starting with the real-time device status and the K fault-related device statuses, a state transition backtracking is performed to obtain the backtracked device status and the K associated backtracked states. The method includes: A first discrete state set of the power transmission and transformation equipment is predefined; The K discrete state sets of the K fault-associated devices are predefined; The first device discrete state set and the K device discrete state sets are used as a joint state space, and the first fault time series dataset and the K fault time series dataset are called locally. Construct a cross-device state transition probability matrix based on the first device discrete state set and the K device discrete state sets; The transition conditional probability is calculated for the first fault time series dataset and the K fault time series datasets, and the calculation results are used to fill the data of the cross-device state transition probability matrix. Using the real-time device state and K associated device states as the initial state, a Viterbi reverse path search is performed along the time axis on the cross-device state transition probability matrix to output the backtracked device state and K associated backtracked states.

4. The method for dynamic monitoring of the operating status of power transmission and transformation equipment based on photoelectric detection as described in claim 1, wherein a gain segmentation controller connected to the infrared focal plane array performs dynamic gain adjustment on the original radiation signal output by the infrared focal plane array, and outputs an original gain adaptation electrical signal, the method comprising: Perform time-domain analysis on the original radiation signal to output the transient signal-to-noise ratio and peak-to-peak dynamic range; Predefined gain adjustment evaluation function; The gain parameters are iteratively adjusted using the gradient descent method so that the rate of change of the transient signal-to-noise ratio within P consecutive iteration cycles is less than a preset threshold, and the updated peak-to-peak dynamic range is adapted to the range of the analog-to-digital converter, thus outputting the gain optimization parameters. The gain optimization parameters are loaded into the gain segmentation controller to scale the amplitude of the original radiated signal and output the original gain-adapted electrical signal.

5. The method for dynamic monitoring of the operating status of power transmission and transformation equipment based on photoelectric detection as described in claim 1, characterized in that, Based on the spectral characteristics of the pulse signal output by the ultraviolet photon counter, the matching inductor parameters and matching capacitor parameters are retrieved from a pre-stored impedance parameter mapping library. The method includes: After performing a fast Fourier transform on the original pulse timing signal output by the ultraviolet photon counter, the spectral distribution of a predefined frequency band is extracted, wherein the predefined frequency band is preferably a 1MHz-1GHz band; Identify the main peak frequency from the said spectral distribution; Using the main peak frequency as the search key, a matching target parameter group is retrieved from the impedance parameter mapping library, wherein the target parameter group consists of the matching inductance parameter and the matching capacitance parameter.

6. The method for dynamic monitoring of the operating status of power transmission and transformation equipment based on photoelectric detection as described in claim 1, characterized in that, By integrating the matching inductor parameters, matching capacitor parameters, original standardized impedance sensing signals, and original gain adaptation electrical signals, the method performs associated fault scenario diagnosis and outputs real-time device status. The method includes: The system interactively obtains multiple sets of perception intervals for multiple sample fault states, and performs perception data association rule mining based on the multiple sets of perception intervals for multiple sample fault states to construct a fault scenario decision tree. The matching inductor parameters, matching capacitor parameters, original standardized impedance sensing signals, and original gain adaptation electrical signals are mapped into feature vectors, loaded into the fault scenario decision tree to perform fault fusion decision-making, and the real-time device status is output. If the power transmission and transformation equipment is in a non-fault state, the real-time equipment status output is an empty set.

7. The method for dynamic monitoring of the operating status of power transmission and transformation equipment based on photoelectric detection as described in claim 6, characterized in that, If the confidence level of the fusion decision of the real-time device status is lower than the confidence threshold, then the state transition backtracking verification is triggered to perform fault attribute time consistency verification.

8. The method for dynamic monitoring of the operating status of power transmission and transformation equipment based on photoelectric detection as described in claim 5, characterized in that, The matching inductor parameters and matching capacitor parameters are written into the adjustable inductor module and adjustable capacitor module of the dynamically adjustable Π-type matching network respectively through the digital control interface, driving the dynamically adjustable Π-type matching network configured at the output end of the ultraviolet photon counter to switch to the target parameter group.

9. The method for dynamic monitoring of the operating status of power transmission and transformation equipment based on photoelectric detection as described in claim 4, characterized in that, The analog-to-digital converter is configured between the output of the gain segment controller and the measurement device, and the infrared focal plane array, the gain segment controller and the measurement device constitute an infrared radiation signal acquisition and conditioning chain.

10. A dynamic monitoring system for the operating status of power transmission and transformation equipment based on photoelectric detection, characterized in that, The steps for implementing the method according to any one of claims 1 to 9 include: A sensing unit deployment module is used to pre-deploy a multispectral sensing unit on the exposed surface of a power transmission and transformation equipment, wherein the power transmission and transformation equipment is a transformer bushing, and the multispectral sensing unit consists of an ultraviolet photon counter, an infrared focal plane array, and a fiber optic grating sensor. The spectrum feature matching module is used to retrieve and output matching inductor parameters and matching capacitor parameters from a pre-stored impedance parameter mapping library based on the spectrum characteristics of the pulse signal output by the ultraviolet photon counter in a dynamic monitoring scenario of operation status. The signal conversion execution module is used to convert the original high-impedance sensing signal output by the fiber Bragg grating sensor into the original normalized impedance sensing signal through the active matching circuit connected to the output end of the fiber Bragg grating sensor. The gain adjustment execution module is used to perform dynamic gain adjustment on the original radiation signal output by the infrared focal plane array through the gain segment controller connected to the infrared focal plane array, and output the original gain adaptation electrical signal. The fault correlation diagnosis module is used to integrate the matching inductor parameters, matching capacitor parameters, original standardized impedance sensing signals and original gain adaptation electrical signals, perform correlation fault scenario diagnosis, and output real-time equipment status. The graded early warning analysis module is used to predict the risk of insulation failure based on the real-time equipment status and output graded early warning commands.

Images