Power cabinet fault prediction self-healing controller and deep learning method

Abstract

The application discloses a power cabinet fault prediction self-healing controller and a deep learning method, and particularly relates to the field of power distribution cabinet monitoring, and is used for solving the misjudgment and misoperation problems caused by the non-synchronization of multi-mode signals. A phase consistency constraint mechanism is constructed through multi-channel fusion caching, the coupling relationship between the current, partial discharge and vibration signals in the fault sample expansion stage is strengthened, and the generation of samples deviating from the true synchronous characteristics is avoided. In the online reasoning stage, a probability fluctuation resolution amplitude and confidence imbalance measurement value extraction strategy based on a sliding window is introduced, a synchronous robustness coefficient is comprehensively generated, and the synchronous robustness coefficient is used as the trigger basis for adaptive resampling and bypass switching, so that a linkage closed structure from sample generation to prediction verification is formed. The dynamic response capability to boundary drift and synchronization mismatch is realized, and the fault recognition precision and the controller action stability in a complex disturbance scene are effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of power distribution cabinet monitoring, more particularly, to a power cabinet fault prediction self-healing controller and a deep learning method. BACKGROUND

[0002] Under the compound working condition of photovoltaic intermittent grid connection and random parallel connection of charging piles, the outdoor power distribution cabinet needs a fault prediction self-healing controller to aggregate multiple-mode transient signals such as current, partial discharge and vibration; in the offline training stage, a generative adversarial network is introduced to expand samples to alleviate the shortage of fault waveforms, but the generator only overwrites the single-channel current track and does not reconstruct the subtle phase coupling and energy complementary relationship between channels. This unbalanced supplement causes the deep model to form an idealized same-frequency feature boundary during training, and when encountering nanosecond-level asynchronous disturbances in the mutual inductor current and partial discharge channels during online operation, the healthy waveform is identified as a high-confidence fault and triggers the solid-state bypass action. Mis-isolation causes the bus to be unloaded instantaneously, generating a reverse voltage spike, and the contact arc quickly erodes the copper bar. Since the log only traces the single-channel prediction results, the mismatch logic is difficult to expose in routine playback, reflecting that there is a coupling blind area between the data supplement strategy and the multi-mode synchronous monitoring link, which is difficult to meet the precision requirements of power cabinet fault prediction and self-healing control.

[0003] To solve the above problems, a technical solution is provided. SUMMARY

[0004] To overcome the above-mentioned defects of the prior art, embodiments of the present application provide a power cabinet fault prediction self-healing controller and a deep learning method. A phase consistency constraint mechanism is constructed through multi-channel fusion caching, which strengthens the maintenance of the coupling relationship between the current, partial discharge and vibration signals in the fault sample expansion stage, avoiding the deviation of generated samples from the true synchronous features. In the online inference stage, a probability fluctuation resolution amplitude and confidence imbalance measurement value extraction strategy based on a sliding window is introduced, and a synchronous robustness coefficient is synthesized as the trigger basis for adaptive resampling and bypass switching, forming a linked closed structure from sample generation to prediction verification. The dynamic response capability to boundary drift and synchronization mismatch is realized, effectively improving the fault identification accuracy and controller action stability in complex disturbance scenarios, to solve the problems raised in the background art.

[0005] To achieve the above-mentioned purposes, the present application provides the following technical solutions:

[0006] S1: A time synchronization window reads multiple-mode sensing streams and writes them into a fusion cache after adding a phase identifier, which is used to lock the transient alignment starting point of the current-partial discharge-vibration three channels;

[0007] S2: Calculate the cross-channel coupling graph according to the fusion cache and extract the phase consistency vector, which is updated with the frame stream as a generative constraint descriptor;

[0008] S3: Call the coupled constraint-based adversarial generation network to expand the synchronous fault sample based on the descriptor, and then write the qualified samples that pass the phase consistency threshold to the enhanced library;

[0009] S4: Train the dilated convolution-attention network with the enhanced library and the original library, and freeze the coupled graph weight as the inference benchmark to stabilize the decision boundary;

[0010] S5: The edge end checks each frame of fault prediction probability online with the inference benchmark as the ruler, extracts the fault prediction probability fluctuation resolution and confidence imbalance metric value through a double-layer sliding window, and generates a self-healing trigger threshold coefficient, which immediately adapts to re-sampling and writing the enhanced library and driving bypass switching once it leaves the safe interval.

[0011] In a preferred embodiment, step S1 includes the following:

[0012] The power cabinet fault prediction self-healing controller reads the current signal, partial discharge signal and vibration signal through the time synchronization window, writes them into the fusion cache after adding the phase identifier, locks the transient alignment starting point of the three signals, sets the time synchronization window, extracts synchronized signal segments from the current sensor, partial discharge sensor and vibration sensor, applies Hilbert transform to each signal segment to calculate the phase identifier, integrates the signal segment and the phase identifier into a record, stores it in the fusion cache in time sequence, and realizes the alignment of the signals in time and phase through the time synchronization window and the phase identifier, locking the transient alignment starting point.

[0013] In a preferred embodiment, step S2 includes the following:

[0014] Read the records in time sequence from the fusion cache, calculate the mutual information between the current signal and the partial discharge signal, the mutual information between the current signal and the vibration signal, and the mutual information between the partial discharge signal and the vibration signal, respectively as the edge weight of the cross-channel coupled graph, calculate the phase difference based on the phase identifier and normalize it to form a phase consistency vector, and combine the mutual information and the phase consistency vector into a generation constraint descriptor, which is dynamically updated with the time synchronization window.

[0015] In a preferred embodiment, step S3 includes the following:

[0016] The coupled constraint-based adversarial generation network is used to expand the synchronous fault sample, and the generator receives random noise and generates a multi-channel fault sample containing current signal, partial discharge signal and vibration signal.

[0017] In a preferred embodiment, step S3 further includes the following:

[0018] The generated constraint descriptor is composed of the cross-channel coupling graph and the phase consistency vector, and the optimization generation process using the coupling constraint loss and the phase consistency loss ensures that the coupling relationship and the phase characteristics of the generated sample are consistent with the real sample.

[0019] In a preferred embodiment, step S3 further comprises the following content:

[0020] The discriminator evaluates the difference between the generated sample and the real sample to optimize the network parameters, calculates the phase consistency vector of the generated sample, and compares the Euclidean distance with the real sample to screen qualified samples and write them into the enhanced library to improve the sample quality.

[0021] In a preferred embodiment, step S4 comprises the following content:

[0022] The original library and the enhanced library are merged to form a joint training set, which contains real-acquired current signals, partial discharge signals and vibration signals and their fault labels, as well as the expanded synchronous fault samples generated by the generative adversarial network; the network architecture is designed to consist of dilated convolution layers, attention mechanisms and coupling graph embedding layers, the dilated convolution layers extract long-range temporal features, the attention mechanisms enhance the attention to key coupling features, and the coupling graph embedding layers embed the cross-channel coupling graph into the network to guide feature fusion.

[0023] In a preferred embodiment, step S4 comprises the following content:

[0024] During training, the goal is to minimize the cross-entropy loss, the coupling graph embedding layer weights are frozen, and only the parameters of the dilated convolution layers, attention mechanisms and classification layers are updated to ensure that the model follows the real physical relationship between signals, while the decision boundary is stabilized by freezing the coupling graph embedding layer weights.

[0025] In a preferred embodiment, step S4 comprises the following content:

[0026] The double-layer sliding window structure is composed of an inner sliding window and an outer sliding window, the inner sliding window calculates the instantaneous difference of fault prediction probability to capture short-term fluctuations, the outer sliding window accumulates the peak intensity of the difference to evaluate long-term trends, the fluctuation resolution amplitude is calculated by piecewise integration of the peak intensity of the difference sequence to reflect the fluctuation activity of the fault prediction probability, and the confidence imbalance metric value is calculated by logarithmic entropy normalization of the confidence contrast array to reflect the imbalance degree of the confidence distribution.

[0027] In a preferred embodiment, step S5 comprises the following content:

[0028] The self-healing trigger threshold coefficient is calculated by the minimum circumscribed ellipse method to calculate the vector module length from the joint scatter density center of the fluctuation resolution amplitude and the confidence imbalance metric value to the origin, and when the self-healing trigger threshold coefficient exceeds the safety threshold, the adaptive resampling and writing of the enhanced library are performed and the bypass switching is driven.

[0029] The technical effects and advantages of the power cabinet fault prediction self-healing controller and deep learning method of the present application are as follows:

[0030] The present application realizes the collaborative modeling of the current, partial discharge and vibration three channels in the fault sample expansion process by constructing the multi-channel synchronous constraint mechanism fused with the cache drive and the generation control strategy based on the phase consistency, effectively avoiding the misjudgment boundary misplacement problem caused by single-channel training. In the online stage, the fault prediction probability output by reasoning is further included in the sliding window dynamic monitoring framework to reveal the sensitive amplitude of the model response with the fluctuation resolution amplitude, and the prediction skewness risk is reflected with the confidence imbalance value, and the synchronous robustness coefficient is generated in combination, which is the core basis for triggering adaptive resampling and bypass switching. Thus, a complete prediction self-healing channel is constructed, which is driven by coupled perception, generates enhancement and online verification closed-loop linkage, not only improves the resistance of abnormal identification to instantaneous disturbance, but also enables the model to have real-time adaptation ability to boundary drift and data heterogeneity in actual deployment, significantly enhances the discriminant stability and self-healing execution precision under complex working conditions. BRIEF DESCRIPTION OF DRAWINGS

[0031] Figure 1 The flowchart of the power cabinet fault prediction self-healing controller and deep learning method of the present application is shown. DETAILED DESCRIPTION

[0032] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0033] Embodiment 1: Figure 1 The power cabinet fault prediction self-healing controller and deep learning method of the present application is given, including:

[0034] S1: The time synchronization window reads the multi-mode sensing stream and writes it into the fusion cache after adding the phase identifier, so as to lock the transient alignment starting point of the current-partial discharge-vibration three channels.

[0035] S2: According to the fusion cache, the cross-channel coupling graph is calculated and the phase consistency vector is extracted, and the vector is updated as a generation constraint descriptor with the frame stream.

[0036] S3: The adversarial generation network with coupling constraints is called to expand the synchronous fault sample based on the descriptor, and the qualified samples are written into the enhanced library after being filtered according to the phase consistency threshold limit.

[0037] S4: The dilated convolution-attention network is trained with the enhanced library and the original library, and the coupling graph weight is frozen as the reasoning benchmark to stabilize the decision boundary.

[0038] S5: The edge end checks the fault prediction probability of each frame online with the inference benchmark as the ruler, extracts the fluctuation resolution amplitude of the fault prediction probability and the confidence imbalance metric value through a double-layer sliding window, and generates a self-healing trigger threshold coefficient. Once it deviates from the safe interval, it immediately adapts to the enhanced library and drives the bypass switch.

[0039] As the core equipment of power distribution and protection in the power system, power cabinets need to cope with the complex working conditions of intermittent photovoltaic grid connection and random parallel charging piles in outdoor environments. In this scenario, power cabinets face the challenges of severe load fluctuations and complex electrical environments, which may lead to power outages, equipment damage, and even safety accidents due to overload, short circuit, or insulation failure. Traditional fault detection relies on single signal (such as current or voltage) monitoring, which is difficult to adapt to the complexity of multi-modal transient signals, and the scarcity of fault data limits the training effect of deep learning models. Therefore, a method based on power cabinet fault prediction self-healing controller and deep learning is proposed, which aggregates multi-modal transient signals such as current, partial discharge (partial discharge), and vibration, and uses deep learning technology to realize fault prediction and self-healing control.

[0040] This method improves fault prediction accuracy through time synchronization and coupling analysis of multi-modal signals, and triggers self-healing actions in real time on the edge. Its operation includes time-synchronized signal acquisition, cross-channel coupling feature extraction, generative adversarial network expansion of fault samples, deep learning model training, and edge real-time decision-making. Compared with traditional methods, this scheme not only can predict potential faults, but also can take self-healing measures such as load adjustment or fault isolation before the fault occurs, thereby improving power supply reliability and equipment safety.

[0041] In the power cabinet fault prediction self-healing controller and deep learning method, step S1 is the basic link of multi-modal signal processing. In view of the complex transient characteristics of outdoor power cabinets under the combined working conditions of photovoltaic and charging piles, step S1 ensures the alignment of current, partial discharge, and vibration signals in time and phase through time synchronization window and phase identification, providing accurate input data for subsequent step S2 to calculate cross-channel coupling graph and extract phase consistency vector.

[0042] When outdoor power distribution cabinets face the complex working conditions of photovoltaic intermittent grid connection and random parallel charging piles, the load fluctuates dramatically and the electrical environment is complex, and the fault features are often hidden in the subtle coupling of multi-modal signals. The reason for collecting current signals, partial discharge signals and vibration signals is that they can comprehensively represent the running state of the power distribution cabinet: current signals directly reflect load changes and electrical faults such as short circuits or overloads; partial discharge signals are sensitive to insulation degradation and potential breakdown risks, and can provide early warning of insulation faults; vibration signals capture mechanical looseness or structural abnormalities to avoid failures caused by physical damage. The combination of the three creates a multi-dimensional feature space that covers multiple main fault modes such as electrical, insulation and mechanical. Compared with other signals, they have higher sensitivity and specificity, significantly improving the accuracy and robustness of fault prediction. In addition, these three types of signal acquisition technology are mature, with low sensor cost and low deployment difficulty, suitable for long-term stable operation in outdoor environments. Other signals such as temperature or sound, while having some reference value, have lower sensitivity or are easily disturbed by the environment, and are less complementary and practical than the above three. Therefore, based on the importance of fault diagnosis, the complementarity between signals and engineering practicality, it is reasonable and efficient to prioritize collecting current signals, partial discharge signals and vibration signals.

[0043] S1.1, multi-modal sensor stream reading:

[0044] The power cabinet fault prediction self-healing controller first performs a multi-modal sensor stream reading step.

[0045] The current signal, partial discharge signal and vibration signal of the power cabinet are collected in real time by the current sensor, partial discharge sensor and vibration sensor respectively. These signals are continuous time series data, which reflect the electrical and mechanical state of the power cabinet under different operating conditions. In order to ensure the accuracy of subsequent analysis, the controller extracts synchronized signal segments from the current sensor, partial discharge sensor and vibration sensor channels. The extraction process requires that the signal segments be aligned in time to provide consistent data input for subsequent signal coupling analysis and fault prediction.

[0046] S1.2, time synchronization window setting:

[0047] After the multi-modal sensor stream reading is completed, the time synchronization window setting step is entered.

[0048] The controller sets a fixed time period as the time synchronization window, for example, the time length is 100 milliseconds. Within this time synchronization window, the same number of sampling points are extracted from the current signal, partial discharge signal and vibration signal respectively. Assuming the sampling frequency is 10000 times per second, each time synchronization window contains 1000 sampling points. In this way, the controller ensures the time consistency of the current signal, partial discharge signal and vibration signal within each time synchronization window. The setting of the time synchronization window aims to provide a unified time reference for subsequent phase analysis and inter-signal coupling calculation.

[0049] S1.3, Phase identifier calculation:

[0050] After completing the time synchronization window setting, the phase identifier calculation step is performed.

[0051] For each current signal segment, partial discharge signal segment and vibration signal segment within each time synchronization window, the respective phase identifier is calculated to represent the phase relationship between the signals. The calculation process first applies Hilbert transform to each signal segment to generate the corresponding complex signal. Then, the instantaneous phase value of each sampling point is extracted from the complex signal. Then, all the instantaneous phase values are averaged to obtain the phase identifier of the signal segment. The phase identifier reflects the overall phase characteristics of the signal segment, which is used to analyze the phase consistency between the current signal, partial discharge signal and vibration signal.

[0052] S1.4, Write fusion cache:

[0053] After the phase identifier calculation is completed, the write fusion cache step is entered.

[0054] The controller integrates the signal data and phase identifier within each time synchronization window into a record and stores it in the fusion cache in chronological order. Each record contains the start timestamp of the time synchronization window, the current signal segment, the partial discharge signal segment, the vibration signal segment, and the corresponding phase identifier. Through this structured storage method, the fusion cache retains the original current signal, partial discharge signal and vibration signal, while adding phase information. This data organization method facilitates subsequent steps to retrieve and process records in chronological order, providing complete data support for cross-channel signal coupling analysis.

[0055] S1.5, Lock transient alignment starting point:

[0056] After completing the write fusion cache, the lock transient alignment starting point step is performed.

[0057] The time synchronization window and phase identifier are used to align the current signal, partial discharge signal, and vibration signal in time and phase. The time synchronization window ensures the synchronization of the three signals in time, while the phase identifier provides a reference standard for the phase relationship between the signals. By aligning the time and phase, the controller locks the transient alignment starting point of the three signals. This alignment method can accurately capture the specific phase and amplitude relationship between signals, thereby improving the accuracy and reliability of fault prediction.

[0058] Since the current signal, partial discharge signal, and vibration signal are key data reflecting the state of the power cabinet, they need to be consistent in time and phase to accurately analyze the coupling relationship between signals. The time synchronization window and phase identifier are used to achieve high-precision synchronization and alignment of multi-modal signals. This method significantly improves the accuracy of signal analysis and provides high-quality input data for deep learning models. At the same time, the integration of phase information enhances the ability to capture subtle relationships between signals, thereby improving the robustness and adaptability of fault prediction. Ultimately, the application of this technology logic reduces the risk of power cabinet misoperation and operation and maintenance costs, and improves the reliability and safety of the power system.

[0059] Step S1 achieves transient alignment of multi-modal signals through a time synchronization window and stores the data with an additional phase identifier in a fusion cache, laying a data foundation for subsequent processing. However, relying solely on synchronized signals cannot fully represent the phase coupling and energy complementary characteristics between signals, especially under nanosecond-level non-synchronous disturbances. Single-channel sample generation can easily lead to model misjudgment. Therefore, step S2 focuses on extracting cross-channel coupling graphs and phase consistency vectors from the fusion cache to quantify these relationships and provide constraint conditions for fault sample generation.

[0060] Step S2 generates cross-channel coupling graphs and phase consistency vectors through mutual information calculation and phase difference extraction, successfully representing the coupling characteristics of current, partial discharge, and vibration signals under complex working conditions. The dynamic update of the generated constraint descriptors provides precise constraints for fault sample generation in step S3, improving sample quality and effectively alleviating the problem of model misjudgment caused by single-channel generation.

[0061] S2.1, read fusion cache data:

[0062] The records are read in time sequence from the fusion cache, each record containing a start timestamp of a time synchronization window, a sampling sequence of the current signal, a phase identifier of the current signal, a sampling sequence of the partial discharge signal, a phase identifier of the partial discharge signal, a sampling sequence of the vibration signal, and a phase identifier of the vibration signal. The reading process ensures that all signal sampling sequences and their corresponding phase identifiers are arranged in the order of time synchronization windows. Reading the records in time sequence can ensure the consistency of signal sampling sequences and phase identifiers in time sequence, providing accurate data sources for subsequent calculation of cross-channel coupling graphs and extraction of phase consistency vectors.

[0063] S2.2, calculating the cross-channel coupling graph:

[0064] The mutual information method is used to quantify the dependency between the current signal and the partial discharge signal, the current signal and the vibration signal, and the partial discharge signal and the vibration signal. The specific process is as follows: first, the mutual information of the current signal sampling sequence and the partial discharge signal sampling sequence is calculated. The calculation method is based on the joint probability distribution of the two signal sampling sequences and their respective marginal probability distributions. The joint probability distribution is estimated by counting the discrete values of the two signal sampling sequences, and the marginal probability distributions are calculated from each signal sampling sequence, respectively, to determine the dependency between the two signals. Then, the mutual information of the current signal sampling sequence and the vibration signal sampling sequence, and the mutual information of the partial discharge signal sampling sequence and the vibration signal sampling sequence are calculated in the same way. The three mutual information values calculated are used as the weights of the corresponding edges in the cross-channel coupling graph, forming a fully connected graph containing three nodes, which correspond to the current signal, the partial discharge signal, and the vibration signal, respectively.

[0065] The mutual information method can capture the nonlinear dependency between signals, and is more suitable for representing the complex signal characteristics of power cabinets under complex working conditions compared to traditional linear correlation analysis. Building a cross-channel coupling graph can intuitively reflect the coupling strength between multi-modal signals, providing important constraint information for subsequent fault sample generation, and ensuring that the generated samples meet the true signal relationship.

[0066] S2.3, extracting the phase consistency vector:

[0067] Based on the phase identification of the current signal, the phase identification of the partial discharge signal, and the phase identification of the vibration signal, the phase difference between each pair of signals is calculated. The specific process is as follows: first, the absolute difference between the current signal phase identification and the partial discharge signal phase identification is calculated to obtain the phase difference between the current signal and the partial discharge signal; then, the absolute difference between the current signal phase identification and the vibration signal phase identification is calculated to obtain the phase difference between the current signal and the vibration signal; and then, the absolute difference between the partial discharge signal phase identification and the vibration signal phase identification is calculated to obtain the phase difference between the partial discharge signal and the vibration signal. The above three phase differences are divided by the maximum possible value of the phase difference, i.e. the circumference, to complete the normalization processing, so that the value of each phase difference falls within the interval of 0 to 1. The three normalized phase differences are arranged in the order of current signal and partial discharge signal, current signal and vibration signal, and partial discharge signal and vibration signal to form a phase consistency vector.

[0068] The phase consistency vector can quantify the consistency degree of the signals in phase, which is of great significance for power cabinet fault prediction, because the fault features are often closely related to the phase relationship between signals. Extracting the phase consistency vector provides phase constraints for fault sample generation, ensuring that the generated samples remain consistent with the real signals in terms of phase characteristics, thereby improving the accuracy of fault prediction.

[0069] S2.4, update the generation constraint descriptor:

[0070] The edge weights of the cross-channel coupling graph, i.e. the three calculated mutual information values, and the phase consistency vector, i.e. the three normalized phase differences, are combined into a generation constraint descriptor. The specific process is as follows: arrange the three mutual information values and the three normalized phase differences in a predetermined order to form a descriptor containing six items of data, which flows dynamically with the frame stream of the time synchronization window, and each update is based on the signal data in the current time synchronization window to recalculate the mutual information and phase difference. The generation constraint descriptor integrates the coupling strength and phase consistency information between signals, providing comprehensive constraint conditions for fault sample generation in step S3. Dynamically updating the generation constraint descriptor can adapt to the time-varying characteristics of the signals, ensuring that the generated samples always reflect the real relationship between the current signals, thereby enhancing the robustness and adaptability of fault prediction.

[0071] Through the above processing logic, the coupling relationship and phase consistency between the current signal, the partial discharge signal, and the vibration signal are accurately characterized, providing precise constraint conditions for fault sample generation. Compared with traditional single-channel sample generation methods, this multi-modal signal analysis can effectively reduce the possibility of model misjudgment, improving the accuracy and reliability of fault prediction. At the same time, the dynamically updated generation constraint descriptor adapts to the signal variation characteristics of the power cabinet under complex working conditions, ensuring the adaptability of the prediction process to complex environments.

[0072] The step S2 generates the cross-channel coupling graph and the phase consistency vector by mutual information calculation and phase difference extraction, successfully representing the coupling characteristics of the current signal, the partial discharge signal and the vibration signal under the composite working condition. The dynamic update of the generated constraint descriptor provides accurate constraints for the generation of the fault sample in step S3, improves the sample quality, effectively alleviates the model misjudgment problem caused by single-channel generation, and provides reliable data support for subsequent training and reasoning.

[0073] The step S2 generates the cross-channel coupling graph and the phase consistency vector by mutual information calculation and phase difference extraction, successfully representing the coupling characteristics of the current signal, the partial discharge signal and the vibration signal under the composite working condition. The dynamic update of the generated constraint descriptor provides accurate constraints for the generation of the fault sample in step S3, improves the sample quality, effectively alleviates the model misjudgment problem caused by single-channel generation, and provides reliable data support for subsequent training and reasoning.

[0074] The step S3 successfully expands the synchronous fault sample by the adversarial generative network with coupling constraints and the phase consistency threshold screening, alleviating the problem of insufficient fault waveforms of the outdoor distribution cabinet under the composite working condition. The qualified samples generated are written into the enhanced library, not only preserving the subtle coupling relationship among the current, partial discharge and vibration signals, but also improving the diversity and representativeness of the samples, laying a solid foundation for the training of the subsequent dilated convolution-attention network, and effectively enhancing the accuracy and robustness of fault prediction.

[0075] S3.1, adversarial generative network with coupling constraints:

[0076] The adversarial generative network consists of a generator and a discriminator. The generator receives random noise and the generated constraint descriptor as input, processes them through an internal neural network, and outputs a multi-channel fault sample containing current signal, partial discharge signal and vibration signal.

[0077] Random noise is derived from a normal distribution and is used to introduce randomness, while the generated constraint descriptor contains the cross-channel coupling graph and the phase consistency vector, which are used to guide the generation process.

[0078] The discriminator receives real fault samples or generated fault samples and outputs a discrimination result representing the probability that the input sample is real.

[0079] The generator must ensure that the generated current signal, partial discharge signal and vibration signal meet the requirements of the cross-channel coupling graph and the phase consistency vector in the generated constraint descriptor when generating the multi-channel fault sample.

[0080] The loss function consists of three parts: adversarial loss, coupling constraint loss, and phase consistency loss. The adversarial loss measures the degree to which the generated samples approximate the real sample distribution by comparing the discriminator's judgments of real and generated samples.

[0081] The coupling constraint loss quantifies the consistency of the coupling strength between the generated current signal, partial discharge signal, and vibration signal with the real sample by calculating the difference in mutual information between the generated and real samples across channels.

[0082] The phase consistency loss ensures that the phase characteristics of the generated samples remain consistent with the real samples by comparing the differences in phase between the generated and real samples.

[0083] The total loss is defined as the weighted sum of the adversarial loss, coupling constraint loss, and phase consistency loss. By optimizing the total loss, the generator and discriminator are jointly trained to generate realistic fault samples that conform to the multi-channel coupling characteristics.

[0084] S3.2, based on the generated constraint descriptor, the synchronous fault sample is expanded:

[0085] The generator generates multi-channel fault samples based on random noise and generated constraint descriptors. Random noise is sampled from a normal distribution as the initial input for the generation process, and the generated constraint descriptor is extracted from the fusion cache of step S2, containing mutual information and phase consistency vectors, used to constrain the generation results.

[0086] The generation process maps random noise and generated constraint descriptors to multi-channel fault samples containing current signals, partial discharge signals, and vibration signals through a neural network. During training, the coupling constraint loss and phase consistency loss are used to ensure that the coupling relationship and phase characteristics of the generated current signal, partial discharge signal, and vibration signal are consistent with the real samples.

[0087] The training process uses an alternating optimization approach. First, fix the generator and optimize the discriminator to enable the discriminator to accurately distinguish between real and generated samples. Then, fix the discriminator and optimize the generator to make the generated samples closer to the real sample distribution. The discriminator evaluates the differences between the generated and real samples, calculates the total loss, and updates the network parameters of the generator and discriminator through backpropagation. Training continues until the total loss converges and the distribution of the generated samples approaches that of the real samples.

[0088] S3.3, screen qualified samples according to the phase consistency threshold:

[0089] Calculate the phase consistency vector for the generated multi-channel fault samples, which includes the phase difference between the current signal and the partial discharge signal, the phase difference between the current signal and the vibration signal, and the phase difference between the partial discharge signal and the vibration signal in the generated samples.

[0090] The phase consistency vector of the generated sample is compared with the phase consistency vector of the real fault sample, and the Euclidean distance between the two is calculated. The Euclidean distance represents the degree of deviation of the generated sample from the real sample in phase characteristics. If the Euclidean distance between the phase consistency vector of the generated sample and the phase consistency vector of the real sample is less than the preset phase consistency threshold, the generated sample is determined to be a qualified sample. The qualified sample is written into the enhanced library as the basis for using the training data in step S4. Through the phase consistency threshold screening, it is ensured that the generated sample is highly consistent with the real sample in phase characteristics, thereby improving the quality and diversity of the sample.

[0091] The adversarial generation network with coupling constraints guides the generation process by generating constraint descriptors, ensuring that the generated multi-channel fault samples reflect the true coupling characteristics between current signals, partial discharge signals, and vibration signals. This method can effectively alleviate the problem of insufficient fault waveforms of outdoor distribution cabinets under complex working conditions, providing more available data support for fault prediction. The phase consistency threshold screening mechanism further ensures the consistency of the generated sample with the real sample in phase characteristics by comparing the Euclidean distance of the phase consistency vector, thereby improving the authenticity and reliability of the sample. The generated high-quality fault samples increase the diversity and representativeness of the training data, enabling the subsequent fault prediction model to better adapt to complex signal characteristics and improve the accuracy and robustness of the prediction. This technical logic fully utilizes the internal relationship of multi-modal signals and solves the contradiction between data scarcity and model performance.

[0092] Step S1 aligns the transient states of the multi-modal signals through a time synchronization window and stores the aligned signals in a fusion cache. Step S2 calculates the cross-channel coupling graph and extracts the phase consistency vector based on the data in the fusion cache, forming the generation constraint descriptor to provide guidance information for subsequent steps. Step S3 uses the adversarial generation network with coupling constraints to expand the synchronous fault samples with the generation constraint descriptor as the input condition, and selects qualified samples through the phase consistency threshold screening to ensure that the generated samples meet the coupling characteristics of the real signals. The generated high-quality samples are written into the enhanced library to provide diversified and reliable data support for model training in step S4. The entire process forms a complete technical chain from signal alignment, feature extraction to sample generation and screening, gradually solving the problem of insufficient fault waveforms and providing a solid data foundation for the optimization of the fault prediction model.

[0093] Step S2 calculates the cross-channel coupling graph and extracts the phase consistency vector according to the fusion cache, as a generation constraint descriptor, to ensure the multimodal consistency during sample generation. Step S3 expands the synchronous fault samples by using the adversarial generation network with coupling constraints, and filters qualified samples to write into the enhanced library, solving the problem of insufficient rare fault samples. However, relying only on high-quality enhanced samples is still insufficient to cope with the dynamic coupling relationship between multimodal signals under complex working conditions, and a fault prediction model with long-range dependence capture ability and decision boundary stability needs to be further constructed. Therefore, step S4 focuses on training the dilated convolution-attention network jointly with the enhanced library and the original library, and freezes the coupling graph weight as the inference benchmark to improve the robustness and prediction accuracy of the model under nanosecond-level asynchronous disturbances.

[0094] Step S4 successfully constructs a fault prediction model that can capture the complex coupling relationship between current, partial discharge and vibration by jointly training the dilated convolution-attention network using multimodal signal samples from the enhanced library and the original library. Freezing the cross-channel coupling graph weight as the inference benchmark ensures the stability of the decision boundary of the model under the photovoltaic intermittent grid connection and charging pile parallel composite working conditions, effectively improving the accuracy and robustness of fault prediction, and providing a reliable model basis for step S5 edge real-time inference.

[0095] S4.1, the construction process of the joint training set:

[0096] By merging the samples in the original library with the samples in the enhanced library, a joint training set is formed. The original library contains current signals, partial discharge signals and vibration signals collected from actual working conditions, as well as fault labels corresponding to these signals. The enhanced library contains synchronous fault samples generated by step S3 and their corresponding fault labels. The purpose of constructing the joint training set is to integrate the advantages of real data and generated data for jointly training the model to improve the model's ability to identify fault features. The merged joint training set not only contains fault samples under real working conditions, but also contains diversified fault samples expanded by the generative adversarial network, thereby increasing the diversity and representativeness of the data. This approach can effectively solve the problem of insufficient rare fault samples and provide comprehensive data support for the model, thereby improving the accuracy and robustness of fault prediction.

[0097] S4.2, the architecture design process of the dilated convolution-attention network:

[0098] The network architecture is designed by an expanded convolutional layer, an attention mechanism, and a coupled graph embedding layer. The expanded convolutional layer expands the receptive field by increasing the dilation rate of the convolution kernel, which is used to extract long-range temporal features of the current signal, partial discharge signal, and vibration signal. The attention mechanism uses a self-attention mechanism to enhance the attention to key coupling features by calculating the correlation between features in different channels. The coupled graph embedding layer uses the cross-channel coupled graph embedding network generated in step S2 as prior knowledge to guide feature fusion (such as weighted summation), ensuring that the model can learn the coupling relationship between the current signal, partial discharge signal, and vibration signal. The reason for designing this architecture is that there are complex temporal and coupling characteristics between multi-modal signals, and separate feature extraction cannot fully capture these relationships. The expanded convolutional layer expands the receptive field, the attention mechanism highlights key features, and the coupled graph embedding layer introduces prior knowledge, all of which work together to significantly improve the model's recognition accuracy of fault features. This structured design fully utilizes the complementarity of multi-modal signals and enhances the accuracy and adaptability of fault prediction.

[0099] S4.3, training target and loss function calculation process:

[0100] The training target is set to minimize the cross-entropy loss. Cross-entropy loss quantifies the prediction accuracy of the model by comparing the difference between the model's predicted fault probability and the true fault label. During training, a batch of sample data is randomly selected from the joint training set and input into the expanded convolutional-attention network to calculate the fault prediction probability of the model output. Then, according to the difference between the predicted probability and the true label, the cross-entropy loss value is calculated, and the parameters of the expanded convolutional layer, attention mechanism, and classification layer are updated using this loss value. The calculation logic is as follows: first, based on the predicted probability and the true label, determine the probability distribution deviation of each class of faults, then sum all classes to get the total loss value, and finally adjust the network parameters through backpropagation to reduce this loss value. Using cross-entropy loss can effectively measure the prediction error in the classification task and guide the model to learn clear fault classification boundaries. This method can improve the model's ability to distinguish fault types and ensure that the prediction results are highly consistent with the true situation.

[0101] S4.4, processing process of freezing coupled graph weights:

[0102] During the training process, the weights of the coupled graph embedding layer are kept fixed, and only the parameters of the dilated convolution layer, the attention mechanism, and the classification layer are updated. The weights of the coupled graph embedding layer come from the cross-channel coupled graph generated in step S2, representing the mutual information relationship between the current signal, the partial discharge signal, and the vibration signal. The processing logic is to embed the weights of the cross-channel coupled graph as prior knowledge into the network, and at each parameter update, only the parameters of other layers are adjusted, while the weights of the coupled graph embedding layer remain unchanged. The reason for freezing the weights is that the cross-channel coupled graph reflects the true physical relationship between the signals. If the weights are allowed to change, it may cause the model to deviate from the correct coupling characteristics due to the distribution offset of the enhanced library samples or the noise interference in the original library. Fixing the weights can ensure that the model always follows the true relationship between the signals, improves the robustness to nanosecond-level asynchronous disturbances, reduces the probability of misjudgment, and thus guarantees the consistency and reliability of the prediction results under different working conditions.

[0103] S4.5, technical implementation process of stabilizing the decision boundary:

[0104] By freezing the weights of the coupled graph embedding layer, the stability of the decision boundary is achieved. During training, the dilated convolution-attention network learns the fault features on the joint training set, and the coupled graph embedding layer provides fixed signal coupling relationship as a constraint, so that the model maintains the consistency of the decision boundary during parameter updating. The processing logic is to use the fixed coupled graph weights to guide feature fusion, ensuring that the fault prediction results output by the model do not fluctuate greatly due to changes in data distribution. The reason for implementing this technology is that outdoor power distribution cabinets have complex working conditions, and signals are easily disturbed. If the decision boundary is unstable, it may lead to fault misjudgment or omission. A stable decision boundary enables the model to accurately identify fault features under complex working conditions, reduces prediction errors, improves the practicality and reliability of the power cabinet fault prediction self-healing controller, and reduces operation and maintenance costs and potential risks.

[0105] Step S1 aligns the transient states of the current signal, the partial discharge signal, and the vibration signal through the time synchronization window, step S2 calculates the cross-channel coupled graph and extracts the phase consistency vector, step S3 generates synchronous fault samples using the coupled constraint adversarial generation network and constructs the enhanced library. Step S4 constructs a joint training set based on the original library and the enhanced library, designs and trains the dilated convolution-attention network, and freezes the weights of the coupled graph embedding layer to ensure that the model captures the complex coupling relationship of multi-modal signals while maintaining the stability of the decision boundary. The entire process starts with signal alignment and gradually completes feature extraction, sample generation, model training, and inference, forming a complete technical chain and systematically solving the technical problem of fault prediction of outdoor power distribution cabinets under complex working conditions, providing support for efficient and reliable self-healing control.

[0106] Step S4 combines the enhanced library with the original library to jointly train the dilated convolution-attention network, and freezes the coupling graph weight as the inference benchmark, laying the foundation for fault prediction. However, the trained model is difficult to fully guarantee the accuracy of the prediction and the timeliness of the response when facing nanosecond-level asynchronous disturbances in real-time operation, so online checking and adaptive adjustment mechanisms need to be introduced at the edge to improve the robustness and self-healing ability of fault prediction. Step S5 focuses on this demand, by real-time monitoring of the synchronization drift and confidence imbalance of the input stream, triggering adaptive resampling and bypass switching to ensure the stable operation of the power distribution cabinet under complex working conditions.

[0107] The edge performs online checking on the fault prediction probability of each frame based on the inference benchmark, extracts the fault prediction probability fluctuation resolution and confidence imbalance metric value through double-layer sliding window, generates a self-healing trigger threshold coefficient, and once the threshold coefficient deviates from the safe interval, immediately performs adaptive resampling and writes back the enhanced library and drives the bypass switching to realize real-time monitoring and self-healing response.

[0108] S5.1, online checking and fault prediction probability acquisition:

[0109] The edge receives real-time multi-modal signal streams, including current signals, partial discharge signals, and vibration signals. These signal streams are divided into multiple frames according to the time synchronization window, each frame containing the sampling data of current signals, partial discharge signals, and vibration signals. Each frame of data is input into the trained dilated convolution-attention network, and the network outputs the prediction probability of the frame being faulty after processing. The prediction probability ranges from 0 to 1. The prediction probability represents the model's confidence in judging the current frame data as faulty, and the higher the value, the more confident the model is that the current frame has a fault. The reason for obtaining the fault prediction probability is to monitor the running state of the power distribution cabinet in real time and use the model prediction results to discover potential fault hazards in a timely manner. It can issue an early warning before the fault occurs, take preventive measures to avoid further expansion of the fault, thereby improving the stability and safety of the power distribution cabinet.

[0110] S5.2, double-layer sliding window setting:

[0111] The double-layer sliding window structure is designed, where the inner layer sliding window is used to capture the short-term fluctuation of the fault prediction probability, and the outer layer sliding window is used to analyze the long-term trend of the fault prediction probability. The length of the inner layer sliding window is shorter, reflecting the instantaneous change of the fault prediction probability in a short time; the length of the outer layer sliding window is longer, covering data in a longer time period, and is used to evaluate the overall trend of the fault prediction probability. The reason for setting double-layer sliding window is that the power distribution cabinet runs under complex working conditions, the signal characteristics are complex, and both short-term fluctuation and long-term trend have important influence on fault prediction. By considering the signal changes at different time scales, the accuracy and robustness of fault prediction can be improved, and false judgments caused by relying on a single time scale analysis can be avoided.

[0112] S5.3, calculate the fluctuation resolution of failure prediction probability:

[0113] In the outer sliding window, the instantaneous difference of consecutive failure prediction probabilities is calculated to generate a difference sequence, which reflects the variation amplitude of failure prediction probability between frames. Then, the difference sequence is segmented and integrated to accumulate the peak intensity in the difference sequence, and finally the fluctuation resolution is calculated. The fluctuation resolution represents the fluctuation activity of failure prediction probability per unit time, and its dimension is failure prediction probability per time. The reason for calculating the fluctuation resolution is that when a fault occurs, the failure prediction probability may fluctuate sharply. By quantifying the fluctuation amplitude, the signs of fault occurrence can be captured. The dynamic changes of failure prediction probability can be monitored in real time, and abnormal fluctuations can be found in time, thereby improving the sensitivity and response speed of failure prediction.

[0114] S5.4, calculate the confidence imbalance metric value:

[0115] In the outer sliding window, a warning threshold is set to divide the failure prediction probabilities into a high confidence group and a low confidence group. The high confidence group contains failure prediction probabilities higher than the warning threshold, and the low confidence group contains failure prediction probabilities lower than the warning threshold. The average probabilities of the high confidence group and the low confidence group are calculated respectively to form a confidence contrast array. Based on the confidence contrast array, the logarithmic entropy is calculated, which reflects the uniformity of the confidence distribution. The confidence imbalance metric value is defined as 1 minus the logarithmic entropy, and the larger the value, the more unbalanced the confidence distribution.

[0116] Under normal working conditions, the failure prediction probability should remain relatively stable, and the confidence distribution is relatively balanced. Under abnormal working conditions, the confidence distribution may be imbalanced. By quantifying the imbalance degree, abnormal situations of model prediction can be identified. The confidence of model prediction can be monitored, and the mismatch between model and data can be found in time, thereby improving the reliability of failure prediction.

[0117] S5.5, generate self-healing trigger threshold coefficient:

[0118] Map the fluctuation resolution and the confidence imbalance metric value to a two-dimensional orthogonal plane to form a coordinate point. Use the set of coordinate points under normal working conditions to fit the smallest circumscribed ellipse to determine the center position of the safety interval. Calculate the vector module length from the current coordinate point to the center of the safety interval to obtain the self-healing trigger threshold coefficient. The self-healing trigger threshold coefficient represents the degree of deviation from normal working conditions, and the larger the value, the more serious the deviation.

[0119] By comprehensively considering the fluctuation and confidence imbalance of failure prediction probability, a reliable decision basis is constructed to avoid misjudgment caused by relying on a single indicator. The accuracy and timeliness of self-healing action can be improved by accurately judging whether the input signal stream has double abnormalities of synchronous drift and confidence imbalance.

[0120] S5.6, Adaptive resampling and write-back enhancement library:

[0121] When the self-healing trigger threshold coefficient exceeds the preset safety threshold, high-frequency resampling is performed on the current frame and its adjacent frames to generate fine-grained data. The fine-grained data and its corresponding fault prediction probability are written back to the enhancement library for optimization of subsequent models.

[0122] The reason for adaptive resampling is that in abnormal situations, more detailed data can be obtained through high-frequency resampling, which can improve data quality. The reason for write-back enhancement library is that supplementing these data to the training data set can optimize model performance. The data acquisition strategy can be adjusted in real time to improve the adaptability of the model to abnormal working conditions, thereby enhancing the accuracy of fault prediction.

[0123] S5.7, Technical logic of driving bypass switching:

[0124] When the self-healing trigger threshold coefficient exceeds the preset safety threshold, solid-state bypass switching is triggered to isolate the fault area and prevent further expansion of the fault. The reason for driving bypass switching is that by isolating the fault area in time, the power distribution cabinet and related equipment can be protected from further damage. The system can quickly respond to faults and reduce the impact of faults on the overall system, thereby improving the self-healing ability and safety of the power distribution cabinet.

[0125] The previous steps complete the transient alignment, feature extraction, fault sample generation, and model training of multi-modal signals, providing data and model support for subsequent fault prediction. On this basis, step S5 calculates the fault prediction probability fluctuation resolution amplitude and confidence imbalance metric value through online verification and fault prediction probability acquisition, combined with double-layer sliding window settings, to generate a self-healing trigger threshold coefficient, realizing real-time monitoring and abnormality determination of the power distribution cabinet operating state. After abnormality determination, adaptive resampling and write-back enhancement library optimize data and models, and drive bypass switching to isolate faults, forming a complete technical chain from signal acquisition, fault prediction to self-healing control, ensuring the safe and stable operation of the power distribution cabinet under the complex working conditions of photovoltaic intermittent grid connection and charging pile parallel connection.

[0126] The adaptive resampling write enhancement library and driving bypass switching is a technical logic integrated in the power cabinet fault prediction self-healing controller and deep learning method, and the core is to realize the fault prediction and self-healing control of the system through real-time monitoring, data optimization and rapid response. The technical logic contains three key components: first, "adaptive resampling" means that the system dynamically adjusts the data acquisition strategy according to the abnormal conditions monitored in real time, so as to capture the fault characteristics under complex working conditions by acquiring signal data at a higher frequency or accuracy; second, "write enhancement library" involves storing the data obtained by adaptive resampling and the corresponding fault prediction probability into the enhancement library as the data set for training the fault prediction model, and continuously supplementing high-quality data to optimize the model performance; finally, "driving bypass switching" triggers the protection mechanism after detecting the fault, isolates the fault area to prevent the fault from expanding, and ensures the safe operation of the system. The meaning is that this technical logic improves the accuracy of fault prediction through adaptive data acquisition, optimizes the robustness and adaptability of the model by using write enhancement library, and realizes rapid fault isolation by means of bypass switching, which together guarantees the stability and safety of the power cabinet under complex working conditions.

[0127] The above formulas are dimensionless values calculated, and the formulas are obtained by software simulation of a large amount of data to obtain a formula of the latest real situation, and the preset parameters in the formula are set by a person skilled in the art according to the actual situation.

[0128] It should be noted that the system of the present application can be deployed on the device itself to realize embedded application, or can be run on PC or other terminal with user interface, so as to meet various hardware environments and use requirements.

[0129] The above only describes some exemplary embodiments of the present application by way of illustration, and it is needless to say that those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present application. Therefore, the above figures and descriptions are illustrative in nature and should not be understood as limiting the scope of protection of the claims of the present application.

[0130] It has to be noted that, in the present document, the terms "first", "second", etc. merely serve the purpose of differentiation and do not require or imply any kind of ordering or sequence of the entities or actions associated therewith. Furthermore, the terms "comprising", "containing", etc. are to be interpreted as non- exclusive in the sense that a process, method, article, or apparatus that comprises a list of elements is not necessarily limited to those elements, but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises... a", "contains... a", etc. does not, without further restriction, exclude the presence of additional identical elements in the process, method, article, or apparatus.

[0131] The above description is only specific embodiments of the present application, but the protection scope of the present application is not limited thereto, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A power cabinet fault prediction self-healing controller and deep learning method, characterized in that, The method comprises the steps of: S1: reading the multi-mode sensing stream in the time synchronization window and writing it into the fusion cache after adding the phase identification, so as to lock the transient alignment starting point of the current-local discharge-vibration three-channel; S2: calculating the cross-channel coupling graph and extracting the phase consistency vector according to the fusion cache, and updating the vector as a generation constraint descriptor with the frame stream; S3: calling the adversarial generation network with coupling constraints to expand the synchronous fault sample based on the descriptor, and generating the sample and writing it into the enhanced library after screening qualified samples according to the phase consistency threshold; S4: training the dilated convolution-attention network with the enhanced library and the original library, and freezing the coupling graph weight as an inference benchmark to stabilize the decision boundary; S5: the edge end checks the fault prediction probability of each frame online with the inference benchmark as the ruler, extracts the fault prediction probability fluctuation resolution and confidence imbalance metric value through a double-layer sliding window, and generates a self-healing trigger threshold coefficient, and once it deviates from the safe interval, it immediately adapts to the enhanced library and drives the bypass switch.

2. The power cabinet failure prediction self-healing controller and deep learning method according to claim 1, wherein, Step S1 includes the following contents: The power cabinet fault prediction self-healing controller reads the current signal, local discharge signal and vibration signal through the time synchronization window, and writes it into the fusion cache after adding the phase identification, so as to lock the transient alignment starting point of the three signals. In order to set the time synchronization window, the synchronous signal fragments are extracted from the current sensor, the local discharge sensor and the vibration sensor. Hilbert transform is applied to each signal fragment to calculate the phase identification. The signal fragment and the phase identification are integrated into a record, which is stored in the fusion cache in time sequence. The alignment of the signals in time and phase is realized through the time synchronization window and the phase identification, and the transient alignment starting point is locked.

3. The power cabinet fault prediction self-healing controller and deep learning method according to claim 2, wherein, Step S2 includes the following contents: The records are read in time sequence from the fusion cache, the mutual information between the current signal and the local discharge signal, the mutual information between the current signal and the vibration signal, and the mutual information between the local discharge signal and the vibration signal are calculated, which are respectively used as the edge weight of the cross-channel coupling graph. The phase difference is calculated based on the phase identification and normalized to form a phase consistency vector. The mutual information and the phase consistency vector are combined into a generation constraint descriptor, which is dynamically updated with the time synchronization window.

4. The power cabinet failure prediction self-healing controller and deep learning method of claim 3, wherein, Step S3 includes the following contents: The adversarial generation network with coupling constraints is used to expand the synchronous fault sample. The generator receives random noise and generates a multi-channel fault sample containing current signal, local discharge signal and vibration signal.

5. The power cabinet failure prediction self-healing controller and deep learning method according to claim 4, wherein, Step S3 also includes the following contents: The generation constraint descriptor is composed of the cross-channel coupling graph and the phase consistency vector. The coupling constraint loss and the phase consistency loss are used to optimize the generation process to ensure that the coupling relationship and phase characteristics of the generated sample are consistent with the real sample. Step S3 also includes the following contents:

6. The power cabinet failure prediction self-healing controller and deep learning method of claim 5, wherein, The discriminator evaluates the difference between the generated sample and the real sample to optimize the network parameters. The phase consistency vector of the generated sample is calculated and compared with the Euclidean distance of the real sample to screen qualified samples and write them into the enhanced library to improve the sample quality. Step S4 includes the following contents:

7. The power cabinet fault prediction self-healing controller and deep learning method according to claim 6, characterized in that, ​ The original library and the enhanced library are combined to form a joint training set, which contains real collected current signals, partial discharge signals and vibration signals and their fault labels, and synchronous fault samples extended by the generative adversarial network; the network architecture is designed to consist of dilated convolution layers, attention mechanisms and coupled graph embedding layers, the dilated convolution layers extract long-range temporal features, the attention mechanisms enhance the attention to key coupling features, and the coupled graph embedding layers embed the cross-channel coupling graph into the network to guide feature fusion.

8. The power cabinet fault prediction self-healing controller and deep learning method according to claim 7, characterized in that, Step S4 includes the following contents: During training, the cross-entropy loss is minimized, the coupled graph embedding layer weights are frozen, and only the dilated convolution layer, attention mechanism and classification layer parameters are updated to ensure that the model follows the real physical relationship between signals, and the decision boundary is stabilized by freezing the coupled graph embedding layer weights.

9. The power cabinet failure prediction self-healing controller and deep learning method of claim 8, wherein, Step S4 includes the following contents: The double-layer sliding window structure is composed of an inner sliding window and an outer sliding window, the inner sliding window calculates the instantaneous difference of fault prediction probability to capture short-term fluctuations, the outer sliding window accumulates the peak intensity of the difference to evaluate long-term trends, the fluctuation resolution amplitude is calculated by piecewise integration of the peak intensity of the difference sequence to reflect the fluctuation activity of the fault prediction probability, and the confidence imbalance measure is calculated by logarithmic entropy normalization of the confidence contrast array to reflect the imbalance degree of the confidence distribution.

10. The power cabinet failure prediction self-healing controller and deep learning method of claim 9, wherein, Step S5 includes the following contents: The self-healing trigger threshold coefficient is calculated by the minimum circumscribed ellipse method to calculate the vector module length from the joint scatter density center of the fluctuation resolution amplitude and the confidence imbalance measure to the origin, and when the self-healing trigger threshold coefficient exceeds the safety threshold, the adaptive resampling and writing enhancement library is executed and the bypass switch is driven.

Images