Circuit breaker failure control method and system

Abstract

The application discloses a circuit breaker fault control method and system, by acquiring the multi-source sensing data of the target circuit breaker in the opening and closing process of each operating stage, extracting the main data features of the multi-source sensing data of each operating stage, determining the coupling strength of each operating stage, and constructing an operating stage fault state transition model to represent the propagation path and probability distribution of the fault among the operating stages, and combining the multi-source sensing data of the current operating stage of the target circuit breaker, updating the state transition probability of the fault of the target circuit breaker in each operating stage, and identifying the precursor fault occurrence stage according to the state transition probability of each operating stage, and then taking the precursor fault occurrence stage as the starting point of fault suppression, optimizing the control parameters of the target circuit breaker in each operating stage according to the propagation path of the fault among the operating stages, realizing the fault precursor sensing and identification, and improving the robustness of fault control.

Description

Technical Field

[0001] This invention relates to the field of power equipment technology, and in particular to a circuit breaker fault control method and system. Background Technology

[0002] Circuit breakers are the core switching equipment of power systems, playing a crucial role in circuit switching and fault isolation. Circuit breaker failures can lead to power outages, equipment damage, and even safety accidents. Therefore, accurate diagnosis and effective control of circuit breaker faults are the core research direction of intelligent operation and maintenance of power equipment.

[0003] Existing circuit breaker fault control methods are mainly divided into two categories: data-driven and linear algorithm-based. Although they have achieved fault identification and control in some simple scenarios, linear algorithm-based methods use linear or low-order nonlinear algorithms, which are difficult to match the nonlinear and time-varying coupling characteristics of circuit breakers. They lack the ability to detect and identify fault precursors, resulting in high rates of missed and false faults. They also have weak ability to identify cross-stage related faults and cannot achieve early warning of faults. At the same time, existing circuit breaker fault control strategies do not fully consider the impact of environmental interference, equipment parameter perturbations, and multi-source interference, resulting in poor robustness, lack of multi-source interference decoupling ability, and easy to cause secondary faults. Summary of the Invention

[0004] In order to solve the above-mentioned technical problems, the present invention provides a circuit breaker fault control method and system.

[0005] The technical solution of this invention is as follows:

[0006] In a first aspect, the present invention provides a circuit breaker fault control method, comprising:

[0007] Acquire multi-source sensor data of the disconnectors of the target circuit breaker during each operating stage of the opening and closing process; wherein, the operating stages include the preparation stage, energy storage stage, opening and closing execution stage and lockout reset stage;

[0008] Extract the master data features of the multi-source sensor data for each of the operation stages, and determine the coupling strength of each operation stage based on each operation stage and its master data features.

[0009] Based on each of the aforementioned operational phases, the master data characteristics of each operational phase, and the coupling strength, an operational phase fault state transition model is constructed; wherein, the operational phase fault state transition model is used to characterize the propagation path and probability distribution of faults between each of the aforementioned operational phases;

[0010] Based on the fault state transition model of the operation phase, and combined with the multi-source sensor data of the current operation phase of the target circuit breaker, the state transition probability of the fault of the target circuit breaker in each operation phase is updated, and the precursor fault occurrence phase is identified according to the state transition probability of each operation phase.

[0011] Taking the occurrence stage of the precursor fault as the starting point of fault suppression, the control parameters of the target circuit breaker in each of the operating stages are optimized according to the propagation path of the fault between each of the operating stages until the state transition probability of each of the operating stages meets the preset safety threshold, and then the optimization of the control parameters of each of the operating stages is stopped.

[0012] Preferably, the multi-source sensing data in the preparation stage includes coil excitation current and control circuit voltage; the energy storage stage includes energy storage pressure and motor speed; the execution stage includes arc intensity, contact displacement and main circuit current; and the lockout reset stage includes lockout position signal and reset spring force.

[0013] The step of extracting the master data features of multi-source sensor data for each of the aforementioned operational stages, and determining the coupling strength of each operational stage based on each operational stage and its master data features, includes:

[0014] Extract multiple time-domain features and multiple frequency-domain features from the multi-source sensor data of each of the aforementioned operational stages;

[0015] Principal component analysis is performed on multiple time-domain features and multiple frequency-domain features of each of the aforementioned operating stages to obtain multiple principal component features of each of the aforementioned operating stages;

[0016] For the aforementioned operational phase, the fault identification degree of each principal component feature is determined based on multiple principal component features of the operational phase.

[0017] Based on the fault identification degree, the master data features of each of the operation stages are determined; wherein the fault identification degree of the master data features is greater than a preset fault identification degree threshold.

[0018] The master data features of each of the aforementioned operational stages are standardized to obtain the standardized master data features of each of the aforementioned operational stages;

[0019] The coupling strength of each operational stage is determined based on the standardized master data characteristics of each operational stage.

[0020] Preferably, determining the coupling strength of each of the operational stages based on the standardized master data characteristics of each operational stage includes:

[0021] For any two of the aforementioned operational phases, the mutual information, phase synchronization degree, and fault propagation probability of the two operational phases are determined based on the master data characteristics of the two operational phases.

[0022] After normalizing the mutual information, phase synchronization degree, and fault propagation probability of the two operation stages respectively, the normalized mutual information, phase synchronization degree, and fault propagation probability are weighted and summed to obtain the coupling strength of the two operation stages.

[0023] Based on the coupling strength between every two operating phases, a coupling strength matrix for each operating phase is determined; wherein, the elements in the coupling strength matrix represent the coupling strength between every two operating phases.

[0024] Preferably, the step of constructing an operational phase fault state transition model based on each operational phase, the master data characteristics of each operational phase, and the coupling strength includes:

[0025] Based on the directed graph topology, each of the aforementioned operational stages is treated as a node, the fault propagation path between the aforementioned operational stages is treated as a directed edge, and the coupling strength between the aforementioned operational stages is mapped to the weight of the directed edge to construct a directed fault propagation graph; wherein, the fault propagation path is determined according to the numerical relationship of the coupling strength between the aforementioned operational stages.

[0026] Based on the master data characteristics of each of the aforementioned operational stages, construct a Gaussian fault membership function for each of the aforementioned operational stages;

[0027] Using the directed edge topology of the directed fault propagation graph as the fault propagation path constraint, and combining the Gaussian fault membership function to output the fault confidence of each node, the fault confidence of each node is corrected according to the coupling strength between each of the operation stages, and the inter-stage state transition probability function is obtained as the fault state transition model of the operation stage.

[0028] Preferably, the step of updating the state transition probability of the target circuit breaker's fault in each of the operating stages based on the fault state transition model of the operating stage, combined with multi-source sensor data of the current operating stage of the target circuit breaker, and identifying the precursor fault occurrence stage according to the state transition probability of each operating stage, includes:

[0029] Acquire multi-source sensor data of the target circuit breaker during its current operating phase, and extract the current master data features of the multi-source sensor data during the current operating phase;

[0030] The current master data features are input into the operational phase fault state transition model to update the fault confidence and state transition probability of each node.

[0031] The operating stage with the highest state transition probability among all operating stages is determined as the precursor fault occurrence stage.

[0032] Preferably, the step of taking the occurrence stage of the precursor fault as the starting point for fault suppression, and optimizing the control parameters of the target circuit breaker in each of the operating stages according to the propagation path of the fault between each operating stage, until the state transition probability of each operating stage meets a preset safety threshold, and then stopping the optimization of the control parameters in each operating stage, includes:

[0033] Starting from the stage of the precursor fault occurrence, the system backtracks step by step along the reverse propagation path of the directed edge topology of the directed fault propagation graph. For each operating stage backtracked step by step, the control parameters of each operating stage are optimized by game theory in combination with Nash equilibrium constraints.

[0034] After optimizing the control parameters for each operating phase, the state transition probability of the target circuit breaker's fault in each operating phase is updated again.

[0035] If the state transition probability of any of the updated operating stages is greater than the preset safety threshold, continue to perform game-theoretic optimization of the control parameters of each operating stage by backtracking step by step, in combination with Nash equilibrium constraints, until the state transition probability of all the updated operating stages is greater than the preset safety threshold, and then stop optimizing the control parameters of each operating stage.

[0036] Preferably, the step of backtracking stepwise along the directed edge topology of the directed fault propagation graph, starting from the stage of the precursor fault occurrence, and optimizing the control parameters of each operational stage through game theory in conjunction with Nash equilibrium constraints, includes:

[0037] Starting from the stage of the precursor fault occurrence, the system backtracks step by step along the reverse propagation path of the directed edge topology of the directed fault propagation graph, traversing each backtracked operational stage.

[0038] For each backtracked operation phase, the optimization objective is to minimize the fault confidence after the control parameters are optimized in the operation phase, and a control parameter optimization model is constructed with Nash equilibrium constraints and control parameter range constraints as constraints.

[0039] For each of the operation stages, the optimization model for the control parameters is optimized and solved, and the optimized control parameters corresponding to the operation stage are determined based on the optimal solution.

[0040] Secondly, the present invention also provides a circuit breaker fault control system, comprising:

[0041] The data acquisition unit is used to acquire multi-source sensor data of the disconnector of the target circuit breaker during each operating stage of the opening and closing process; wherein, the operating stages include the preparation stage, energy storage stage, opening and closing execution stage and lockout reset stage;

[0042] The coupling strength determination unit is used to extract the master data features of the multi-source sensing data of each of the operation stages, and determine the coupling strength of each operation stage based on each operation stage and the master data features of each operation stage.

[0043] A state transition unit is used to construct a state transition model for each operating stage based on the master data characteristics and coupling strength of each operating stage; wherein the state transition model for each operating stage is used to characterize the propagation path and probability distribution of the fault between each operating stage.

[0044] The fault occurrence stage identification unit is used to update the state transition probability of the fault of the target circuit breaker in each of the operating stages based on the fault state transition model of the operating stage and the multi-source sensor data of the current operating stage of the target circuit breaker, and to identify the precursor fault occurrence stage according to the state transition probability of each of the operating stages.

[0045] The control parameter optimization unit is used to optimize the control parameters of the target circuit breaker in each of the operating stages, taking the early fault occurrence stage as the starting point for fault suppression, according to the propagation path of the fault between each of the operating stages, until the state transition probability of each of the operating stages meets the preset safety threshold, and then stop optimizing the control parameters of each of the operating stages.

[0046] Thirdly, the present invention also provides an electronic device, the electronic device including a memory and a processor, the memory storing a computer program, which, when executed by the processor, causes the processor to perform the circuit breaker fault control method as described in the first aspect.

[0047] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, the computer-readable storage medium storing computer-executable instructions for performing the circuit breaker fault control method as described in the first aspect.

[0048] The beneficial effects of this invention are:

[0049] This invention acquires multi-source sensor data of the disconnectors of a target circuit breaker during each operating stage of the opening and closing process, extracts the master data features of the multi-source sensor data of each operating stage, determines the coupling strength of each operating stage, and thus constructs an operating stage fault state transition model to characterize the propagation path and probability distribution of the fault between each operating stage. Furthermore, it combines the multi-source sensor data of the target circuit breaker's current operating stage to update the fault state transition probability of the target circuit breaker in each operating stage, and identifies the precursor fault occurrence stage based on the state transition probability of each operating stage. Then, using the precursor fault occurrence stage as the starting point for fault suppression, it optimizes the control parameters of the target circuit breaker in each operating stage according to the fault propagation path between each operating stage, thereby achieving dynamic truncation of the fault propagation chain and adaptive closed-loop parameter control. This realizes fault precursor perception and identification, avoiding missed or false fault detection. Furthermore, it utilizes the operating stage fault state transition model to deduce the fault state transition evolution trend of each operating stage, improving the ability to identify cross-stage associated faults and enhancing the robustness of fault control. Attached Figure Description

[0050] Figure 1 A flowchart of a circuit breaker fault control method;

[0051] Figure 2 This is a schematic diagram of a circuit breaker fault control system. Detailed Implementation

[0052] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0053] like Figure 1 As shown, the present invention provides a circuit breaker fault control method, comprising the following steps:

[0054] Step 1: Obtain multi-source sensor data of the disconnector of the target circuit breaker during each operating stage of the opening and closing process; the operating stages include the preparation stage, energy storage stage, opening and closing execution stage and lockout reset stage.

[0055] Based on the opening and closing timing characteristics of the disconnector of the target circuit breaker, the operation process can be precisely divided into four stages. The preparation stage is the system self-check and parameter initialization stage before the disconnector operates. The energy storage stage corresponds to the energy accumulation process of the spring or hydraulic mechanism. The opening and closing execution stage covers the transient dynamic behavior of the actual separation / contact of the main contacts. The lockout and reset stage completes the mechanical position locking and the state signal returning to zero. There is a strong nonlinear coupling relationship between the four stages, and their state variables are mutually constrained and their evolution paths are highly chaotic.

[0056] By acquiring multi-source sensor data for each operational stage through multiple sensors, and performing spatiotemporal synchronization calibration and wavelet threshold denoising preprocessing on the multi-source sensor data for each operational stage, it is ensured that heterogeneous signals such as coil current, vibration acceleration, and arc intensity are accurately aligned under time deviation and spatial error constraints. This supports the millisecond-level identification of micro-feature amplitude, time sequence difference, and mutation rate by the adaptive partitioning model for each operational stage, making the stage partitioning accuracy consistently above 99.5%.

[0057] In one example, the multi-source sensing data in the preparation stage includes coil excitation current and control circuit voltage; the energy storage stage includes energy storage pressure and motor speed; the execution stage includes arc intensity, contact displacement and main circuit current; and the lockout reset stage includes lockout position signal and reset spring force.

[0058] Among them, the coil excitation current is the current generated when the coil in the circuit breaker is energized under controlled conditions in the preparatory stage to produce a magnetic field. Its waveform distortion rate directly reflects the core saturation degree and the trend of insulation degradation between coil turns. The control circuit voltage is the output voltage of the control circuit that drives the coil. Its drop amplitude and response delay jointly characterize the increase in control circuit contact resistance and the aging state of relay contacts. The energy storage pressure reflects the energy storage integrity of the spring or hydraulic mechanism, while the motor speed fluctuation reveals the coupling effect of mechanical clearance and lubrication degradation in the transmission system. The transient peak value and rising slope of the arc intensity envelop the contact erosion degree and the attenuation law of the dielectric recovery strength. The micron-level offset of the contact displacement trajectory and the zero-crossing drift of the main circuit current jointly characterize the risk of arc reignition. The jitter spectrum of the lockout position signal and the attenuation rate of the reset spring force accurately map the fatigue accumulation effect of the mechanical interlocking mechanism.

[0059] Based on this, the master data features of multi-source sensor data in each operation stage are extracted. The coupling strength of each operation stage is determined according to the operation stage and its master data features. This includes: extracting multiple time-domain and frequency-domain features of the multi-source sensor data in each operation stage; performing principal component analysis on the multiple time-domain and frequency-domain features of each operation stage to obtain multiple principal component features for each operation stage; determining the fault identification degree of each principal component feature for each operation stage based on its multiple principal component features; determining the master data features of each operation stage based on the fault identification degree; wherein the fault identification degree of the master data features is greater than a preset fault identification degree threshold; standardizing the master data features of each operation stage to obtain standardized master data features for each operation stage; and determining the coupling strength of each operation stage based on the standardized master data features.

[0060] Specifically, by performing Fourier transform and wavelet packet decomposition on the multi-source sensor data of each operating stage, frequency domain features including amplitude spectral entropy, frequency band energy ratio, harmonic distortion rate, and instantaneous frequency variance are extracted; in addition, time domain features such as mean, variance, skewness, kurtosis, zero-crossing rate, and impulse factor of the multi-source sensor data of each operating stage are statistically analyzed.

[0061] Principal component analysis was performed on multiple time-domain and frequency-domain features during the operation phase, and principal components with a cumulative contribution rate of ≥95% were extracted as principal component features.

[0062] Fault identification score is the ability of a feature to distinguish between normal and fault states. Principal component features with fault identification scores greater than a threshold (e.g., 0.7) can be retained as master data features during the operational phase. Furthermore, Z-score standardization is applied to these features to eliminate the influence of dimensions. The fault identification score is:

[0063] In the formula, For fault identification, , These are the mean values ​​of the feature under normal and fault conditions, respectively. , These represent the standard deviations of the features under normal and fault conditions, respectively.

[0064] Step 2: Extract the master data features of multi-source sensor data in each operation stage, and determine the coupling strength of each operation stage based on each operation stage and its master data features.

[0065] Among them, the master data feature is the criterion signal component that best represents the essence of fault evolution in each operating stage. By identifying the master data feature of multi-source sensor data in each operating stage, the fault evolution dynamics of that stage can be accurately characterized.

[0066] Among them, coupling strength is a quantitative measure of the nonlinear interaction between different operating stages. The greater the correlation between different operating stages, the higher the coupling strength and the more significant the fault propagation path.

[0067] In one example, the coupling strength of each operational phase is determined based on the standardized master data characteristics of each operational phase. This includes: for any two operational phases, determining the mutual information, phase synchronization, and fault propagation probability of the two operational phases based on their master data characteristics; normalizing the mutual information, phase synchronization, and fault propagation probability of the two operational phases respectively, and then weighting and summing the normalized mutual information, phase synchronization, and fault propagation probability to obtain the coupling strength of the two operational phases; and determining the coupling strength matrix of each operational phase based on the coupling strength of every two operational phases. The elements in the coupling strength matrix represent the coupling strength of every two operational phases.

[0068] Specifically, for each operational phase, taking any two operational phases (such as phase V and phase X) as the calculation objects, mutual information is the value of the mutual information between the master data features of phase V and phase X. It characterizes the degree of information sharing between the master data features of the two phases; the larger the value, the stronger the correlation between the feature information of the two phases. The mutual information is calculated as follows:

[0069] ;

[0070] In the formula, This represents the mutual information value of the master data features between stage V and stage X. This is the main data feature set for stage V. This is the main data feature set for stage X. The joint probability density function of the main data features v and x is obtained from the statistical distribution of the two-stage main data features. The marginal probability density function of the main data feature v. It is the marginal probability density function of the main data feature x.

[0071] Phase synchronization is a measure of the dynamic changes in the main data characteristics of stages V and X. It calculates the phase difference and synchronization of the characteristic signals, characterizing the degree of temporal synchronization between the physical actions of the two stages. The phase synchronization is calculated as follows:

[0072] ;

[0073] In the formula, For phase synchronization, The number of samples for the two-stage master data feature signals. Let be the phase difference between the two stages of the main data feature signals at the i-th sampling point. This is the mean of the phase differences across all sampling points.

[0074] The fault propagation probability is determined based on the physical coupling mechanism of each operating stage of the circuit breaker, combined with the fault types corresponding to the master data characteristics of the two stages. This determines the probability of a fault in stage V propagating to stage X, characterizing the degree of fault cascading between the two stages. The fault propagation probability is calculated as follows:

[0075] ;

[0076] In the formula, Let V be the probability of a fault propagating to stage X. The fault confidence level for stage V. Cosine similarity of the principal data features between stages V and X. Taken as the time-series decay coefficient, adjacent stages (e.g., preparation → energy storage). A value of 1 indicates a skipping of one stage (e.g., preparation → execution). The value is 0.7, spanning two stages (e.g., preparation → locking reset). The value is 0.4. , The failure risk coefficients for stages V and X are 0.4 for the execution stage, 0.3 for the energy storage stage, 0.2 for the preparation stage, and 0.1 for the lockout reset stage, representing the impact weight of stage failures.

[0077] After normalizing the mutual information, phase synchronization, and fault propagation probability of stages V and X, and then weighting and summing the normalized mutual information, phase synchronization, and fault propagation probability, the coupling strength of stages V and X is obtained. Based on the coupling strength of the pairwise stage combinations, the coupling strength matrix of each operating stage is determined. This matrix has a dimension of 4×4, with the main diagonal elements being 1.0, reflecting the complete consistency of self-coupling in each stage; the off-diagonal elements take values ​​in the range [0,1]. The closer the value is to 1, the stronger the nonlinear coupling and the closer the dynamic coordination between the two stages.

[0078] Step 3: Based on each operating stage and the master data characteristics and coupling strength of each operating stage, construct the operating stage fault state transition model; wherein, the operating stage fault state transition model is used to characterize the propagation path and probability distribution of faults between each operating stage.

[0079] Among them, the fault state transition model in the operation phase models the fault propagation evolution process between each operation phase in the form of a directed weighted graph, which can characterize the fault propagation path and transition probability between the four phases of preparation, energy storage, execution and lockout reset.

[0080] Step 4: Based on the fault state transition model of the operation phase, and combined with the multi-source sensor data of the current operation phase of the target circuit breaker, update the state transition probability of the fault of the target circuit breaker in each operation phase, and identify the stage of the precursor fault occurrence according to the state transition probability of each operation phase.

[0081] Specifically, by acquiring the current operating stage of the target circuit breaker and the multi-source sensor data of that stage in real time, the state transition probability of each operating stage is dynamically updated using the operating stage fault state transition model. The operating stage with the highest state transition probability is identified as the precursor fault occurrence stage, which is the operating stage with the most significant precursor fault signals, providing a precise spatiotemporal anchor point for subsequent graded early warning and self-healing control.

[0082] Step 5: Taking the occurrence stage of the precursor fault as the starting point of fault suppression, optimize the control parameters of the target circuit breaker in each operating stage according to the propagation path of the fault between each operating stage, until the state transition probability of each operating stage meets the preset safety threshold, and stop optimizing the control parameters of each operating stage.

[0083] Since fault propagation has time-dependent characteristics and nonlinear transitions between stages, this application takes the occurrence stage of the precursor fault as the starting point for fault suppression. According to the constraints of the fault propagation path between each operating stage, the control parameters of each operating stage are optimized to ensure that the positive diffusion of the fault on the propagation path is suppressed, while blocking its nonlinear transition to subsequent stages, until the state transition probability of each operating stage converges to below the preset safety threshold, thereby realizing the active truncation and closed-loop suppression of the fault propagation chain.

[0084] This application embodiment acquires multi-source sensor data of the disconnector of the target circuit breaker during each operating stage of the opening and closing process, extracts the master data features of the multi-source sensor data of each operating stage, determines the coupling strength of each operating stage, and thus constructs an operating stage fault state transition model to characterize the propagation path and probability distribution of the fault between each operating stage. It also combines the multi-source sensor data of the target circuit breaker in the current operating stage to update the fault state transition probability of the target circuit breaker in each operating stage, and identifies the stage of precursor fault occurrence based on the state transition probability of each operating stage. Then, taking the stage of precursor fault occurrence as the starting point of fault suppression, it optimizes the control parameters of the target circuit breaker in each operating stage according to the propagation path of the fault between each operating stage, thereby realizing the dynamic truncation of the fault propagation chain and the adaptive closed-loop control of parameters, realizing the perception and identification of fault precursors, avoiding fault omission and misjudgment, and using the operating stage fault state transition model to deduce the fault state transition evolution trend of each operating stage, improving the ability to identify cross-stage associated faults and improving the robustness of fault control.

[0085] In one example, a fault state transition model for each operational stage is constructed based on the master data characteristics and coupling strength of each operational stage. This includes: constructing a directed fault propagation graph based on a directed graph topology, treating each operational stage as a node, the fault propagation path between each operational stage as a directed edge, and mapping the coupling strength between each operational stage to the weight of the directed edge; wherein the fault propagation path is determined according to the numerical relationship of the coupling strength between each operational stage; constructing a Gaussian fault membership function for each operational stage based on the master data characteristics of each operational stage; using the directed edge topology of the directed fault propagation graph as the fault propagation path constraint, and combining the Gaussian fault membership function to output the fault confidence of each node, and correcting the fault confidence of each node according to the coupling strength between each operational stage, to obtain the inter-stage state transition probability function as the operational stage fault state transition model.

[0086] By comparing the numerical relationships of the coupling strengths between different operational stages, the subsequent operational stage corresponding to each stage in the fault propagation path is determined. For example, if the coupling strengths of stage X with stages V, Y, and Z are 0.82, 0.65, and 0.31 respectively, then a fault in stage X is most likely to propagate to stage V, followed by stage Y, and has the lowest probability of propagating to stage Z. Therefore, the weights of the directed edges from stage X to stage V are set to 0.82, for stage Y to 0.65, and for stage Z to 0.31, forming a fault propagation priority sequence. Based on this priority sequence, a dynamic fault propagation path tree, i.e., a directed fault propagation graph, is constructed.

[0087] Among them, the Gaussian fault membership function for each operational stage quantifies the fault probability distribution. The S-type Gaussian mixture membership function is selected to balance the sensitivity and discriminative power of precursor micro-features. The Gaussian fault membership function is as follows:

[0088] ;

[0089] In the formula, Let i be the main data feature vector of stage i. For the principal data feature vector in stage i The output value of the fault membership, that is, the fault confidence; This is the normal feature threshold. The standard deviation of the main data feature vectors, where the normal feature threshold is... The value is taken from the main data feature vector. The average value under normal operating conditions.

[0090] By using the directed edge topology of the directed fault propagation graph as the constraint for the fault propagation path of the model, only the state of the preceding stage is allowed to transition to the state of the following stage, with no reverse transitions or cross-order jumps. The propagation path of the model is completely consistent with the directed edges of the directed fault propagation graph, thus achieving an accurate representation of the fault propagation path.

[0091] By correcting the fault confidence of each node through the coupling strength between each operational stage, the inter-stage state transition probability function is obtained as the operational stage fault state transition model:

[0092] ;

[0093] In the formula, Let be the state transition probability from stage i to stage i+1. Let represent the coupling strength between stage i and stage i+1.

[0094] By combining coupling strength and Gaussian mixture membership function, this operational phase fault state transition model ensures that the fault evolution process strictly follows the physical time sequence and coupling mechanism, which not only suppresses false propagation in low-coupling paths but also enhances the precursor response sensitivity of high-coupling paths; by dynamically updating the coupling strength value in conjunction with real-time micro-feature data, the transition probability has online adaptive capability.

[0095] In one example, based on the operational phase fault state transition model and combined with multi-source sensor data of the target circuit breaker's current operational phase, the state transition probability of the target circuit breaker's fault in each operational phase is updated, and the precursor fault occurrence phase is identified according to the state transition probability of each operational phase. This includes: acquiring multi-source sensor data of the target circuit breaker's current operational phase and extracting the current master data features of the multi-source sensor data of the current operational phase; inputting the current master data features into the operational phase fault state transition model to update the fault confidence and state transition probability of each node; and determining the operational phase with the highest state transition probability as the precursor fault occurrence phase.

[0096] Specifically, by acquiring multi-source sensor data of the current operating stage of the target circuit breaker and extracting the current master data features of the multi-source sensor data of the current operating stage according to the aforementioned steps, the current master data features are input into the model to calculate the fault confidence and the transition probability after coupling correction for each stage in real time. At the same time, by comparing the relative magnitude of the transition probabilities of each stage, the stage of the precursor fault occurrence is accurately located. It should be noted that if multiple stages with the same state transition probability occur, the first stage can be selected based on the temporal sequence relationship between the stages. In terms of temporal sequence, the preparation stage takes precedence over the energy storage stage, the energy storage stage takes precedence over the execution stage, and the execution stage takes precedence over the blocking reset stage.

[0097] In some embodiments, starting from the occurrence stage of a precursor fault, the control parameters of the target circuit breaker in each operating stage are optimized according to the propagation path of the fault in each operating stage until the state transition probability of each operating stage meets a preset safety threshold, at which point the optimization of the control parameters of each operating stage is stopped. This includes: starting from the occurrence stage of a precursor fault, backtracking step by step along the backward propagation path of the directed edge topology of the directed fault propagation graph, and performing game-theoretic optimization of the control parameters of each operating stage in combination with Nash equilibrium constraints; after optimizing the control parameters of each operating stage, updating the state transition probability of the target circuit breaker in each operating stage; if the updated state transition probability of any operating stage is greater than the preset safety threshold, continuing to perform game-theoretic optimization of the control parameters of each operating stage in combination with Nash equilibrium constraints until the updated state transition probability of all operating stages is greater than the preset safety threshold, at which point the optimization of the control parameters of each operating stage is stopped.

[0098] Starting from the stage of the precursor fault, the system backtracks step by step along the directed edge topology of the directed fault propagation graph to ensure that the control intervention accurately anchors the source of the fault. Each round of game optimization is embedded with Nash equilibrium constraints. Nash equilibrium constraints are stable strategy combinations reached in non-cooperative games at each stage. Any unilateral deviation at any stage will lead to a decrease in its own fault suppression payoff, thereby ensuring global balance.

[0099] Specifically, starting from the stage of the precursor fault occurrence, the system backtracks step by step along the directed edge topology of the directed fault propagation graph. For each operational stage reached through this backtracking, the control parameters of each operational stage are optimized using a game theory approach, taking into account Nash equilibrium constraints. This includes: starting from the stage of the precursor fault occurrence, backtracking step by step along the directed edge topology of the directed fault propagation graph, traversing each backtracked operational stage; for each backtracked operational stage, minimizing the fault confidence after optimizing the control parameters in that operational stage is the optimization objective, and a control parameter optimization model is constructed using Nash equilibrium constraints and control parameter range constraints as constraints; the optimization model for the control parameters under each operational stage is optimized and solved, and the optimized control parameters corresponding to that operational stage are determined based on the optimal solution.

[0100] The control parameters for each operating stage are used as decision variables. The control parameters for the preparatory stage include the coil pre-charge voltage, the initial compression of the energy storage spring, and the mechanical clearance tolerance. The control parameters for the energy storage stage include the motor drive current limit, the energy storage capacitor charging time constant, and the hydraulic system back pressure threshold. The control parameters for the execution stage include the opening and closing coil pulse width, the contact pressure dynamic compensation coefficient, and the arc-extinguishing chamber airflow disturbance suppression gain. The control parameters for the lockout and reset stage include the mechanical reset spring stiffness coefficient, the position sensor sampling delay compensation, and the lockout electromagnet response time margin.

[0101] The circuit breaker described in this application is applicable to SF6 circuit breakers, vacuum circuit breakers, etc. The coil pre-charge voltage is a core control parameter in the circuit breaker's preparation stage, directly affecting the excitation build-up rate and the consistency of the initial mechanical response. The initial compression of the energy storage spring is an important physical benchmark for the energy storage accuracy and release stability during the energy storage stage; its deviation will directly cause distortion of the contact movement trajectory during the execution stage. The mechanical clearance tolerance characterizes the sensitivity of the mechanism's assembly tolerance to the nonlinear amplification of the dynamic response. These three factors together constitute the rigid boundary for robust control in the preparation stage. The motor drive current limit determines the energy injection rate and thermal accumulation risk boundary during the energy storage process; the energy storage capacitor charging time constant affects the dynamic matching accuracy of energy conversion; and the hydraulic system back pressure threshold regulates the pressure response hysteresis and overshoot suppression capability of the energy storage actuator. These three factors work together to form a rigid constraint set for the nonlinear coupled control of the energy storage stage. The pulse width of the opening and closing coil directly determines the initial kinetic energy and motion stability of the contacts; the dynamic compensation coefficient of the contact pressure addresses the attenuation of contact force caused by mechanical wear and temperature drift; and the airflow disturbance suppression gain in the arc extinguishing chamber ensures the spatiotemporal consistency of the dielectric recovery strength during the arc extinguishing process.

[0102] By minimizing the fault confidence after optimizing the control parameters as the optimization objective, the objective function can be determined as follows:

[0103] ;

[0104] in, ,in, For stage i, the control parameter variables, For stage i, the control variable The state mapping function between the master data features of stage i can obtain the master data feature response of the multi-source sensor data by changing the control parameters. Substituting the master data feature response into the aforementioned Gaussian fault membership function, the fault confidence of the corresponding stage i can be calculated.

[0105] The Nash equilibrium constraint is:

[0106] ;

[0107] In the formula, These are the optimal control parameters for stage V. These are the optimal control parameters for stage X. The optimal control parameters are in stage X. Given the given conditions, find the control parameters that minimize the objective function value (fault confidence) in stage V. This control parameter The optimal control parameters for stage V. .

[0108] The control parameter range constraint means that the control parameters at each stage of operation must be within the normal physical adjustment range to avoid exceeding the equipment hardware capabilities, that is:

[0109] ;

[0110] In the formula, , These are the control parameters. The minimum and maximum values ​​are determined by the physical limits of the equipment.

[0111] The control parameter optimization model is optimized by using a mathematical solver, and the optimized control parameters for the corresponding operation stage are determined based on the optimal solution.

[0112] like Figure 2 As shown in the figure, this application provides a circuit breaker fault control system, including:

[0113] The data acquisition unit 101 is used to acquire multi-source sensor data of the disconnector of the target circuit breaker during each operating stage of the opening and closing process; wherein, the operating stages include the preparation stage, energy storage stage, opening and closing execution stage and lockout reset stage;

[0114] The coupling strength determination unit 102 is used to extract the main data features of multi-source sensing data in each operating stage, and determine the coupling strength of each operating stage based on each operating stage and the main data features of each operating stage.

[0115] The state transition unit 103 is used to construct a fault state transition model for each operating stage based on the master data characteristics and coupling strength of each operating stage; wherein, the fault state transition model for each operating stage is used to characterize the propagation path and probability distribution of the fault between each operating stage;

[0116] The fault occurrence stage identification unit 104 is used to update the state transition probability of the fault of the target circuit breaker in each operating stage based on the fault state transition model of the operating stage and combined with the multi-source sensor data of the current operating stage of the target circuit breaker, and to identify the precursor fault occurrence stage according to the state transition probability of each operating stage.

[0117] The control parameter optimization unit 105 is used to optimize the control parameters of the target circuit breaker in each operating stage, taking the early warning fault occurrence stage as the starting point for fault suppression and following the propagation path of the fault between each operating stage, until the state transition probability of each operating stage meets the preset safety threshold, and then stop optimizing the control parameters of each operating stage.

[0118] This application provides an electronic device, which includes a memory and a processor. The memory stores a computer program, and when the computer program is executed by the processor, the processor performs the circuit breaker fault control method as described in the above embodiments.

[0119] The present invention also provides a computer-readable storage medium having a computer program stored thereon, the computer-readable storage medium storing computer-executable instructions for performing the circuit breaker fault control method as described in the above embodiments.

[0120] The above description only illustrates preferred embodiments of the present invention and should not be construed as limiting the scope of the claims. The present invention is not limited to the above embodiments, and variations in its specific structure are permitted. In short, all variations made within the scope of the independent claims of the present invention are within the scope of protection of the present invention.

Claims

1. A circuit breaker fault control method, characterized in that, include: Acquire multi-source sensor data of the disconnectors of the target circuit breaker during each operating stage of the opening and closing process; wherein, the operating stages include the preparation stage, energy storage stage, opening and closing execution stage and lockout reset stage; Extract the master data features of the multi-source sensor data for each of the operation stages, and determine the coupling strength of each operation stage based on each operation stage and its master data features. Based on each of the aforementioned operational phases, the master data characteristics of each operational phase, and the coupling strength, an operational phase fault state transition model is constructed; wherein, the operational phase fault state transition model is used to characterize the propagation path and probability distribution of faults between each of the aforementioned operational phases; Based on the fault state transition model of the operation phase, and combined with the multi-source sensor data of the current operation phase of the target circuit breaker, the state transition probability of the fault of the target circuit breaker in each operation phase is updated, and the precursor fault occurrence phase is identified according to the state transition probability of each operation phase. Taking the occurrence stage of the precursor fault as the starting point of fault suppression, the control parameters of the target circuit breaker in each of the operating stages are optimized according to the propagation path of the fault between each of the operating stages until the state transition probability of each of the operating stages meets the preset safety threshold, and then the optimization of the control parameters of each of the operating stages is stopped.

2. The circuit breaker fault control method according to claim 1, characterized in that, The multi-source sensing data in the preparation stage includes coil excitation current and control circuit voltage; the energy storage stage includes energy storage pressure and motor speed; the execution stage includes arc intensity, contact displacement and main circuit current; the lockout reset stage includes lockout position signal and reset spring force. The step of extracting the master data features of multi-source sensor data for each of the aforementioned operational stages, and determining the coupling strength of each operational stage based on each operational stage and its master data features, includes: Extract multiple time-domain features and multiple frequency-domain features from the multi-source sensor data of each of the aforementioned operational stages; Principal component analysis is performed on multiple time-domain features and multiple frequency-domain features of each of the aforementioned operating stages to obtain multiple principal component features of each of the aforementioned operating stages; For the aforementioned operational phase, the fault identification degree of each principal component feature is determined based on multiple principal component features of the operational phase. Based on the fault identification degree, the master data features of each of the operation stages are determined; wherein the fault identification degree of the master data features is greater than a preset fault identification degree threshold. The master data features of each of the aforementioned operational stages are standardized to obtain the standardized master data features of each of the aforementioned operational stages; The coupling strength of each operational stage is determined based on the standardized master data characteristics of each operational stage.

3. The circuit breaker fault control method according to claim 2, characterized in that, The step of determining the coupling strength of each operational stage based on the standardized master data characteristics of each operational stage includes: For any two of the aforementioned operational phases, the mutual information, phase synchronization degree, and fault propagation probability of the two operational phases are determined based on the master data characteristics of the two operational phases. After normalizing the mutual information, phase synchronization degree, and fault propagation probability of the two operation stages respectively, the normalized mutual information, phase synchronization degree, and fault propagation probability are weighted and summed to obtain the coupling strength of the two operation stages. Based on the coupling strength between every two operating phases, a coupling strength matrix for each operating phase is determined; wherein, the elements in the coupling strength matrix represent the coupling strength between every two operating phases.

4. The circuit breaker fault control method according to claim 1, characterized in that, The step of constructing a fault state transition model for each operational phase based on the master data characteristics and coupling strength of each operational phase includes: Based on the directed graph topology, each of the aforementioned operational stages is treated as a node, the fault propagation path between the aforementioned operational stages is treated as a directed edge, and the coupling strength between the aforementioned operational stages is mapped to the weight of the directed edge to construct a directed fault propagation graph; wherein, the fault propagation path is determined according to the numerical relationship of the coupling strength between the aforementioned operational stages. Based on the master data characteristics of each of the aforementioned operational stages, construct a Gaussian fault membership function for each of the aforementioned operational stages; Using the directed edge topology of the directed fault propagation graph as the fault propagation path constraint, and combining the Gaussian fault membership function to output the fault confidence of each node, the fault confidence of each node is corrected according to the coupling strength between each of the operation stages, and the inter-stage state transition probability function is obtained as the fault state transition model of the operation stage.

5. The circuit breaker fault control method according to claim 4, characterized in that, The step of updating the state transition probability of the target circuit breaker's fault in each of the operating stages based on the fault state transition model of the operating phase, combined with multi-source sensor data of the current operating phase of the target circuit breaker, and identifying the precursor fault occurrence stage based on the state transition probability of each operating stage, includes: Acquire multi-source sensor data of the target circuit breaker during its current operating phase, and extract the current master data features of the multi-source sensor data during the current operating phase; The current master data features are input into the operational phase fault state transition model to update the fault confidence and state transition probability of each node. The operating stage with the highest state transition probability among all operating stages is determined as the precursor fault occurrence stage.

6. The circuit breaker fault control method according to claim 4, characterized in that, The step of taking the occurrence stage of the precursor fault as the starting point for fault suppression, and optimizing the control parameters of the target circuit breaker in each of the operating stages according to the propagation path of the fault between each operating stage, until the state transition probability of each operating stage meets a preset safety threshold, and then stopping the optimization of the control parameters in each operating stage, includes: Starting from the stage of the precursor fault occurrence, the system backtracks step by step along the reverse propagation path of the directed edge topology of the directed fault propagation graph. For each operating stage backtracked step by step, the control parameters of each operating stage are optimized by game theory in combination with Nash equilibrium constraints. After optimizing the control parameters for each operating phase, the state transition probability of the target circuit breaker's fault in each operating phase is updated again. If the state transition probability of any of the updated operating stages is greater than the preset safety threshold, continue to perform game-theoretic optimization of the control parameters of each operating stage by backtracking step by step, in combination with Nash equilibrium constraints, until the state transition probability of all the updated operating stages is greater than the preset safety threshold, and then stop optimizing the control parameters of each operating stage.

7. The circuit breaker fault control method according to claim 6, characterized in that, Starting from the stage of the precursor fault occurrence, the process involves backtracking step by step along the directed edge topology of the directed fault propagation graph. For each operational stage reached through this backtracking, the control parameters of each operational stage are optimized using a game-theoretic approach, taking into account Nash equilibrium constraints. This includes: Starting from the stage of the precursor fault occurrence, the system backtracks step by step along the reverse propagation path of the directed edge topology of the directed fault propagation graph, traversing each backtracked operational stage. For each backtracked operation phase, the optimization objective is to minimize the fault confidence after the control parameters are optimized in the operation phase, and a control parameter optimization model is constructed with Nash equilibrium constraints and control parameter range constraints as constraints. For each of the operation stages, the optimization model for the control parameters is optimized and solved, and the optimized control parameters corresponding to the operation stage are determined based on the optimal solution.

8. A circuit breaker fault control system, characterized in that, include: The data acquisition unit is used to acquire multi-source sensor data of the disconnector of the target circuit breaker during each operating stage of the opening and closing process; wherein, the operating stages include the preparation stage, energy storage stage, opening and closing execution stage and lockout reset stage; The coupling strength determination unit is used to extract the master data features of the multi-source sensing data of each of the operation stages, and determine the coupling strength of each operation stage based on each operation stage and the master data features of each operation stage. A state transition unit is used to construct a state transition model for each operating stage based on the master data characteristics and coupling strength of each operating stage; wherein the state transition model for each operating stage is used to characterize the propagation path and probability distribution of the fault between each operating stage. The fault occurrence stage identification unit is used to update the state transition probability of the fault of the target circuit breaker in each of the operating stages based on the fault state transition model of the operating stage and the multi-source sensor data of the current operating stage of the target circuit breaker, and to identify the precursor fault occurrence stage according to the state transition probability of each of the operating stages. The control parameter optimization unit is used to optimize the control parameters of the target circuit breaker in each of the operating stages, taking the early fault occurrence stage as the starting point for fault suppression, according to the propagation path of the fault between each of the operating stages, until the state transition probability of each of the operating stages meets the preset safety threshold, and then stop optimizing the control parameters of each of the operating stages.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor. The memory stores a computer program, which, when executed by the processor, causes the processor to perform the circuit breaker fault control method as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer-readable storage medium stores computer-executable instructions for performing the circuit breaker fault control method as described in any one of claims 1-7.

Images