Intelligent air switch control monitoring method based on lora communication

Abstract

The invention discloses an intelligent air switch control monitoring method based on lora communication, and relates to the technical field of air switch control monitoring, and the method comprises the following steps: carrying out the continuous collection of a voltage sampling sequence and a current sampling sequence of an air switch, generating a variable-amplitude density curve, a reversal interval period value and a waveform pole delay track, forming a continuous feature sequence, and carrying out the inversion of the interval period value and the waveform pole delay track; whether the sampling signal has regular distortion under the influence of the electromagnetic interference is judged; and under the condition that the sampling signal has regular distortion, extracting a distortion confirmation point set from the continuous feature sequence, and calculating a feature deviation factor, a trend residual vector and a stable window convergence degree based on each distortion confirmation point to form an abnormal form mapping group. According to the invention, the problem that the structural distortion of the sampled data under electromagnetic interference is difficult to identify is solved, distortion perception and state closed-loop regulation and control based on the feature sequence are realized, and the monitoring accuracy and the system safety are improved.

Description

Technical Field

[0001] This invention relates to the field of circuit breaker control and monitoring technology, specifically to an intelligent circuit breaker control and monitoring method based on LoRa communication. Background Technology

[0002] Intelligent circuit breaker control and monitoring based on LoRa communication refers to upgrading traditional circuit breakers into intelligent devices with remote control and status monitoring capabilities using LoRa (Long Range) low-power wide-area network communication technology, thereby achieving intelligent management and fault early warning of power lines. In existing technologies, such systems typically include multiple components such as intelligent circuit breaker devices, embedded LoRa modules, LoRa gateways, backend cloud platforms, and user terminals. The workflow is generally as follows: the backend platform or user terminal issues control commands (such as power off or power on), which are forwarded to the LoRa gateway via the cloud platform and then transmitted to the target circuit breaker via the LoRa wireless channel. While executing the command, the circuit breaker collects real-time operating status data such as current, voltage, and temperature and transmits it back to the gateway via LoRa. The gateway then reports this data to the cloud platform for storage, analysis, and display, enabling remote monitoring and control of the equipment's operating status. In addition, some systems integrate overload / short circuit identification, automatic power-off protection, alarm push notifications, data statistics, and strategy control, forming a closed-loop intelligent control and monitoring system covering multiple complete links such as command issuance, execution response, status acquisition, data feedback, and intelligent analysis. The advantages of this type of system lie in its long communication distance, strong penetration capability, and low power consumption, making it particularly suitable for scenarios involving wide-area deployment of power equipment, distributed control, and low-power operation.

[0003] The existing technology has the following shortcomings: In the process of implementing intelligent circuit breaker control and monitoring based on LoRa communication, when a high-intensity electromagnetic interference source (such as industrial welding equipment) is temporarily connected in the environment where the circuit breaker is located, it may cause instantaneous interference to the local electrical signal sampling process of the circuit breaker, resulting in regular "instantaneous zeroing" or "periodic shift" of the collected operating parameters such as voltage and current. Since this interference occurs during the data sampling stage rather than the communication link itself, and the sampling results are still within the numerical threshold range, the data uploaded via LoRa communication appears to be in a "superficially stable" operating state on the platform. Existing intelligent circuit breaker control and monitoring technologies based on LoRa communication typically assume that a normal communication link means the data is true and reliable. They do not effectively identify and verify interference factors during the data acquisition process, nor do they establish a authenticity judgment mechanism based on data structure characteristics. Therefore, they cannot identify whether the operating state of the circuit breaker has been distorted based on the abnormal electrical parameters when the sampled signal shows regular distortion under the influence of electromagnetic interference. This leads to the platform misjudging the circuit breaker as being in a normal state during operation evaluation. This problem will cause the system to rely on structurally distorted data for status judgment in the long term, affecting the accuracy of trend identification and risk assessment, easily leading to monitoring failure and alarm lag, and ultimately may result in the failure to provide timely warnings of load failures, seriously affecting the system's intelligent monitoring capabilities and power safety operation.

[0004] The information disclosed in the background section is only intended to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this invention is to provide an intelligent circuit breaker control and monitoring method based on LoRa communication to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides the following technical solution: a smart circuit breaker control and monitoring method based on LoRa communication, specifically including the following steps: S1. Continuously acquire the voltage and current sampling sequences of the circuit breaker to generate the amplitude density curve, the inversion interval period value, and the waveform pole delay trajectory, forming a continuous feature sequence to determine whether there is a regular distortion of the sampling signal under the influence of electromagnetic interference. S2. When the sampled signal has regular distortion, extract the set of distortion confirmation points from the continuous feature sequence, and calculate the feature deviation factor, trend residual vector and stability window convergence based on each distortion confirmation point to form an abnormal shape mapping group. S3. Combine and process the various indicators in the abnormal morphology mapping group to generate the amplitude response offset rate, morphological coupling score and total stability deviation, and calculate the credibility score of the operating status to identify whether the operating status of the circuit breaker is distorted. S4. Based on the relationship between the credibility score and the dynamic threshold mapping coordinate, generate state classification results and divide the running state into the data credibility zone, the data critical zone and the data distortion zone. S5. Perform dynamic control operations based on the state classification results. In the data trust zone, maintain the original sampling and communication frequency. In the data critical zone, increase the sampling frequency and pause the remote control response. In the data distortion zone, trigger signal screening and mark the current data as an abnormal source, and feed it back to the continuous feature sequence acquisition process to form a closed loop.

[0007] Preferably, S1 specifically includes the following steps: S101. The voltage sampling sequence and current sampling sequence of the circuit breaker are continuously acquired according to a unified time base, so that each voltage sampling value and the current sampling value at the corresponding time are time-aligned, and a basic sampling sequence is formed in the order of sampling time. S102. Based on the amplitude variation relationship between adjacent sampled values ​​in the basic sampling sequence, calculate the amplitude variation density in the continuous sampling interval to generate a variable amplitude density curve; calculate the time interval between adjacent reversal points based on the position where the sampled value change direction reverses to generate a reversal interval period value; calculate the extreme value time offset based on the position change of the amplitude extreme point in the sampling sequence relative to the time base to generate a waveform extreme point delay trajectory. S103. The amplitude density curve, the inversion interval period value, and the waveform pole delay trajectory are arranged in chronological order to form a continuous feature sequence. The periodic repetition degree of the amplitude density curve, the time consistency degree of the inversion interval period value, and the offset stability degree of the waveform pole delay trajectory are compared in the continuous feature sequence. When the amplitude density curve shows a fixed periodic repetition relationship, the inversion interval period value maintains a time consistency relationship, and the waveform pole delay trajectory maintains a stable offset relationship, it is determined that the sampled signal has a regular distortion under the influence of electromagnetic interference.

[0008] Preferably, S102 is as follows: In the basic sampling sequence, the amplitude difference is calculated for adjacent sampled values, and the amplitude difference is arranged in the order of sampling time. By statistically analyzing the distribution density of amplitude differences within continuous sampling intervals, an amplitude density curve reflecting the frequency of amplitude changes is formed. In the basic sampling sequence, the position where the amplitude change direction changes from rising to falling or from falling to rising is identified, the position where the change direction reverses is determined as the reversal point, and the reversal interval period value is calculated based on the sampling time interval corresponding to adjacent reversal points. In the basic sampling sequence, the positions where the amplitude reaches a local maximum or local minimum are identified as extreme points. Using a unified time base as a reference, the time offset of each extreme point relative to the time base is calculated. By arranging the time offsets of consecutive extreme points, the waveform extreme point delay trajectory is generated.

[0009] Preferably, S2 specifically includes the following steps: S201. When the sampled signal has regular distortion, perform point-by-point analysis along the time sequence of the continuous feature sequence. When the amplitude density curve corresponding to the same time position shows a repeated change rhythm in multiple adjacent time intervals, the reversal interval period value maintains the same time interval distribution in multiple adjacent time intervals, and the waveform pole delay trajectory shows a continuous and consistent time offset trend in multiple adjacent time intervals, the data at the corresponding time position is determined as the distortion confirmation point, and multiple distortion confirmation points are arranged in time sequence to form a distortion confirmation point set. S202. For each twist confirmation point in the twist confirmation point set, based on the feature value of the continuous feature sequence corresponding to that time position, calculate the offset of the feature value relative to the overall distribution position of the continuous feature sequence to obtain the feature deviation factor. At the same time, calculate the difference distribution between the trend of continuous feature change before and after that time position and the overall trend of continuous feature sequence change to obtain the trend residual vector. And within the continuous time window centered on the twist confirmation point, calculate the concentration of feature change to obtain the stability window convergence. S203. The feature deviation factor, trend residual vector and stability window convergence corresponding to each distortion confirmation point are combined and arranged according to the time order of the distortion confirmation point in the continuous feature sequence to form an abnormal morphology mapping group, which is used to characterize the abnormal electrical parameters when the sampled signal exhibits regular distortion under the influence of electromagnetic interference.

[0010] Preferably, S202 specifically refers to: For each twist confirmation point in the twist confirmation point set, the feature value of the continuous feature sequence corresponding to that time position is extracted, and the position of the feature value is compared with the feature value distribution formed by the continuous feature sequence over the complete time range. By calculating the distance magnitude of the feature value relative to the distribution center, the feature deviation factor used to characterize the degree of local feature deviation is obtained. Around the same twist confirmation point, the continuous feature change trajectory within the continuous time interval before and after the time position is extracted, and the trend of the change trajectory is compared with the change trajectory of the continuous feature sequence over the entire time range. By describing the distribution of the differences between the two types of change trajectories in terms of directionality and change amplitude, the trend residual vector is obtained. Centered on the time position corresponding to the same twist confirmation point, the feature values ​​of the continuous feature sequence within a continuous time window are selected. The concentration and dispersion of the feature value changes within the time window are statistically analyzed, and the statistical results are used as the stability window convergence to characterize the clustering state of feature changes in the neighborhood of the twist confirmation point.

[0011] Preferably, S3 specifically includes the following steps: S301. Extract the feature deviation factor, trend residual vector and stable window convergence corresponding to each distortion confirmation point in the abnormal morphology mapping group in chronological order, and construct the corresponding index arrangement sequence while maintaining time alignment, as the basic feature data source for subsequent combination processing. S302. Based on the change trajectory of different types of indicators in the indicator arrangement sequence, calculate the offset ratio between the maximum value and the average value of the feature deviation factor to generate the amplitude response offset rate; perform vector angle aggregation analysis on the consistency of the residual direction at each time in the trend residual vector to generate the morphological coupling score; and perform fluctuation amplitude statistics within the sliding interval on the stability window convergence to generate the total stability deviation. S303. Normalize the amplitude response offset rate, morphological coupling score and total stability deviation, and perform weighted accumulation based on the preset multi-dimensional index fusion weight to generate a credibility score value for the operating status; when the credibility score value is lower than the distortion identification threshold, it is identified that the operating status of the circuit breaker has been distorted.

[0012] Preferably, S302 is as follows: Extract the index arrangement sequence composed of feature deviation factors, calculate the ratio of the difference between the maximum feature deviation factor value and the average feature deviation factor value in the index arrangement sequence, and use this ratio as the amplitude response offset rate to measure the difference in response intensity of local electrical parameter anomalies in the overall sequence. The index arrangement sequence composed of trend residual vectors is extracted. For each trend residual vector, the angle between it and the main trend direction of the whole sequence is calculated. Cosine aggregation is performed on all angles to obtain the morphological coupling score that reflects the consistency of residual direction, which is used to characterize the consistency of trend deviation at multiple points. The index sequence composed of the convergence of the stability window is extracted, the fluctuation amplitude between the convergence values ​​is calculated in adjacent time intervals, and the total stability deviation is constructed based on the maximum fluctuation amplitude within the statistical range to evaluate the influence of local stability changes on the overall feature sequence.

[0013] Preferably, S4 is as follows: Based on the calibrated operating states and corresponding credibility scores in historical samples, a mapping coordinate system between credibility scores and dynamic thresholds is constructed. By extracting the correspondence between credibility scores and distortion levels under different working conditions in multiple time periods, a segmented fitting method is used to generate multiple dynamic threshold curves, forming a mapping coordinate system that reflects the relationship between score changes and state division boundaries. Map the current credibility score to the constructed mapping coordinate system. Based on the corresponding interval position of the credibility score on the dynamic threshold curve, determine its relative position in the mapping coordinate system and obtain the numerical relationship between the current credibility score and the dynamic threshold. The status is classified based on the position of the current credibility score in the mapped coordinate system. If the credibility score is in the high score range of the dynamic threshold curve, the running status is classified as the data credibility zone; if the credibility score is in the fluctuating transition range of the dynamic threshold curve, the running status is classified as the data critical zone; if the credibility score is below the lower boundary of the dynamic threshold curve, the running status is classified as the data distortion zone, and the corresponding status classification result is generated.

[0014] Preferably, S5 is as follows: When the state classification result is in the data confidence zone, the control sampling system maintains the current sampling frequency of the voltage sampling sequence and the current sampling sequence, and keeps the current LoRa communication frequency unchanged, so that the data continues to enter the continuous feature sequence acquisition process under stable conditions, ensuring the time consistency and signal integrity in the feature generation process. When the state classification result is a data critical zone, adjust the sampling control unit parameters, increase the sampling frequency of the voltage sampling sequence and the current sampling sequence to a high-precision sampling level, and shield the response mechanism of the remote control signal, suspend the transmission and execution of remote control commands, and ensure that the integrity of the sampling signal is not affected by control intervention, so as to enhance the accuracy of anomaly identification. When the state classification result is a data distortion area, the signal processing unit is activated to filter the sampled data in the current time period, identify data points with feature distribution offset in the continuous feature sequence, mark such data as an abnormal source, and feed the marking result back to the continuous feature sequence acquisition process to drive the isolation management of abnormal data in subsequent sampling cycles, thus constructing a dynamic control mechanism for state closed-loop control.

[0015] The technical effects and advantages provided by the present invention in the above technical solution are as follows: 1. This invention achieves accurate identification of the regular distortion characteristics of voltage and current sampling sequences under electromagnetic interference by constructing a continuous feature sequence centered on amplitude density curves, reversal interval period values, and waveform pole delay trajectories. By extracting feature deviation factors, trend residual vectors, and stability window convergence from the distorted data, an abnormal morphology mapping group is further constructed. Combined with the index arrangement sequence and distribution relationship, a reliability score is generated and correlated with dynamic threshold mapping coordinates to achieve refined classification and judgment of operating status. Compared to traditional methods that rely solely on communication link status or numerical thresholds, this solution can identify potentially distorted data that appears stable on the surface but is structurally abnormal, solving the technical problem of existing intelligent circuit breaker monitoring systems being unable to detect sampling process interference and prone to misjudging system status.

[0016] 2. This invention constructs a dynamic state control mechanism based on a reliability score, which can automatically adjust the sampling frequency, control the remote signal response, filter abnormal signals, and feed them back to the acquisition process according to the classification results of the operating state, thereby forming a dynamic control system oriented towards state-based closed-loop control. It maintains the original sampling efficiency in the data reliability zone, enhances feature monitoring accuracy in the critical zone, and actively manages abnormal data isolation in the distortion zone, ensuring that the entire monitoring system possesses both real-time performance and the ability to perceive and adaptively control structured data anomalies in the LoRa communication environment. This scheme not only improves the accuracy and robustness of operating state identification but also provides key technical support for building intelligent protection capabilities for power monitoring systems in high-interference environments. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this invention. For those skilled in the art, other drawings can be obtained based on these drawings.

[0018] Figure 1 This is a flowchart illustrating the intelligent circuit breaker control and monitoring method based on LoRa communication according to the present invention. Detailed Implementation

[0019] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, they are provided so that the description of this disclosure will be more complete and fully convey the concept of the exemplary embodiments to those skilled in the art.

[0020] This invention provides, for example Figure 1 The intelligent circuit breaker control and monitoring method based on LoRa communication shown includes the following steps: S1. Continuously acquire the voltage and current sampling sequences of the circuit breaker to generate the amplitude density curve, the inversion interval period value, and the waveform pole delay trajectory, forming a continuous feature sequence to determine whether there is a regular distortion of the sampling signal under the influence of electromagnetic interference. In this embodiment, S1 specifically includes the following steps: S101. The voltage sampling sequence and current sampling sequence of the circuit breaker are continuously acquired according to a unified time base, so that each voltage sampling value and the current sampling value at the corresponding time are time-aligned, and a basic sampling sequence is formed in the order of sampling time. When monitoring the electrical status of circuit breakers, voltage and current signals can be acquired in real time using a synchronous sampler. In practice, the voltage and current sampling terminals must be connected to the same time control clock to ensure that sampling actions are performed simultaneously under a unified time reference. A high-precision synchronous sampling chip or a unified timing trigger mechanism can be used to ensure a one-to-one correspondence between voltage and current sampling values ​​at each moment. After acquisition, each pair of voltage and current sampling values ​​is arranged in chronological order, forming a time-incrementing sampled data stream. For example, if 10 samples are taken within one cycle, and the voltage is 220 volts and the current is 5 amps at the 5th sample, these two values ​​are written into the basic sampling sequence as a feature combination at the same time point. This method of synchronously sampling voltage and current using a unified clock and constructing a time series effectively ensures data temporal consistency, facilitating subsequent feature calculations and anomaly detection without distortion.

[0021] A circuit breaker voltage sampling sequence refers to an ordered dataset of voltage values ​​continuously acquired from the circuit breaker's voltage detection points, while a current sampling sequence corresponds to a set of current values ​​acquired within the same sampling period. A unified time reference means that all sampling actions are controlled by the same clock, ensuring that each set of voltage and current samples precisely corresponds to the same sampling moment. Time alignment refers to the one-to-one pairing of voltage and current data at the sampling level, ensuring strict synchronization in the time domain. The basic sampling sequence is a data set assembled in chronological order from each set of voltage and current samples under unified time control. It not only preserves the signal's inherent time-domain dynamic characteristics but also provides a complete, continuous, and comparable raw data foundation for subsequent continuous feature extraction. The core significance of constructing a basic sampling sequence lies in preventing analytical errors caused by data misalignment, thereby improving the accuracy of interpreting electrical operating conditions, especially suitable for anomaly detection in high-frequency interference environments.

[0022] S102. Based on the amplitude variation relationship between adjacent sampled values ​​in the basic sampling sequence, calculate the amplitude variation density in the continuous sampling interval to generate a variable amplitude density curve; calculate the time interval between adjacent reversal points based on the position where the sampled value change direction reverses to generate a reversal interval period value; calculate the extreme value time offset based on the position change of the amplitude extreme point in the sampling sequence relative to the time base to generate a waveform extreme point delay trajectory. S103. The amplitude density curve, the inversion interval period value, and the waveform pole delay trajectory are arranged in chronological order to form a continuous feature sequence. The periodic repetition degree of the amplitude density curve, the time consistency degree of the inversion interval period value, and the offset stability degree of the waveform pole delay trajectory are compared in the continuous feature sequence. When the amplitude density curve shows a fixed periodic repetition relationship, the inversion interval period value maintains a time consistency relationship, and the waveform pole delay trajectory maintains a stable offset relationship, it is determined that the sampled signal has a regular distortion under the influence of electromagnetic interference.

[0023] In determining whether a sampled signal exhibits regular distortion, the amplitude density curve, the reversal interval period value, and the waveform pole delay trajectory can be sequentially mapped and concatenated to generate a continuous feature sequence containing multiple time segments. Each time segment contains three data points that comprehensively describe the frequency of amplitude changes, the rhythm of waveform reversal, and the delay trend of extreme points within that time period. Next, a sliding window technique is used to analyze each time segment of the continuous feature sequence segment by segment. The dominant frequency repetition pattern in the amplitude density curve is extracted, its repetition period is calculated, and its persistence is assessed. Simultaneously, the consistency of the time span of the reversal interval period value is calculated. If the difference between adjacent period values ​​is minimal, it indicates strong temporal consistency. For the waveform pole delay trajectory, it is analyzed whether the time offset change of the extreme points maintains a constant slope or exhibits a regular linear offset. If all three indicators show high stability and repeatability across multiple time periods, it can be determined that the current signal exhibits significant regular distortion, indicating potential influence from stable external electromagnetic interference. This type of interference often does not alter the communication link but continuously distorts data sampling.

[0024] Continuous feature sequences are composite data streams that synchronously arrange different types of analytical features using time as an index, used for multi-dimensional detection of electrical parameter change trends. The periodic repetition of the amplitude density curve refers to whether the curve exhibits the same fluctuation frequency and waveform distribution pattern over a continuous time period. Regular repetition usually indicates that the sampled signal has been subjected to periodic interference. The temporal consistency of the reversal interval period value refers to whether the time difference between different reversal points tends to be constant over multiple periods. This can be used to determine whether signal changes are controlled by a disturbance source of a fixed frequency. The stability of the waveform pole delay trajectory offset describes the overall displacement trend of the extreme point's time position. If it continuously increases or decreases linearly, it indicates that the interference effect is stable and has a clear direction. When all three features simultaneously meet the regularity condition—that is, the amplitude density curve has a fixed periodic repetition relationship, the reversal interval period value maintains a temporal consistency relationship, and the waveform pole delay trajectory maintains a stable offset relationship—it can be determined with high confidence that the sampled signal has suffered non-random, continuous interference, thus identifying that the signal has been regularly distorted. This analysis process, through cross-verification of composite indicators, significantly improves the accuracy of anomaly identification.

[0025] In this embodiment, S102 specifically refers to: In the basic sampling sequence, the amplitude difference is calculated for adjacent sampled values, and the amplitude difference is arranged in the order of sampling time. By statistically analyzing the distribution density of amplitude differences within continuous sampling intervals, an amplitude density curve reflecting the frequency of amplitude changes is formed. In the basic sampling sequence, each pair of immediately preceding and following voltage or current samples can be considered as adjacent sample values. By calculating the difference between the values ​​of the two samples, the amplitude difference corresponding to the current sampling point is obtained. These amplitude differences are then arranged sequentially according to the sampling time to form an amplitude variation sequence. To analyze the activity level of the sampled data over a period of time, the frequency distribution of amplitude differences can be statistically analyzed within a preset continuous sampling interval. This dense distribution is then mapped to a density scale to generate a variable amplitude density curve. This curve appears on the image as a trend of varying peak density. Steep and frequently fluctuating sections of the curve represent strong and frequent voltage or current fluctuations, while flat sections reflect relative signal stability during that time period. For example, within a 10-millisecond sampling interval, if the difference between adjacent sample values ​​frequently approaches the threshold, the density curve exhibits a dense peak state, indicating the presence of short-term strong fluctuations in that region. Such fluctuations are likely associated with anomalies triggered by electromagnetic interference. Adjacent sampled values ​​are the basic units for constructing fluctuation characteristics. The amplitude difference provides information on the instantaneous rate of change, the distribution density depicts the overall trend of change, and the amplitude density curve integrates this information to visualize and quantify the dynamic characteristics of the signal in local segments, thus providing a basis for subsequent judgment on whether there is a regular distortion in the sampled signal.

[0026] In the basic sampling sequence, the position where the amplitude change direction changes from rising to falling or from falling to rising is identified, the position where the change direction reverses is determined as the reversal point, and the reversal interval period value is calculated based on the sampling time interval corresponding to adjacent reversal points. In a basic sampling sequence, the direction of amplitude change of the sampled signal can be determined by continuously comparing the numerical relationship between each sampled value and its predecessor. When the amplitude of a certain sampling point is found to be larger than the previous value and smaller than the next value, or smaller than the previous value but larger than the next value, it indicates that the direction of amplitude change changes from rising to falling or from falling to rising around this point. The position where this trend reverses can be defined as a reversal point. The appearance of a reversal point reflects the turning characteristics of the signal waveform and is an important basis for identifying the fluctuation period and oscillation frequency. After identifying multiple reversal points, the sampling time interval between any pair of adjacent reversal points can be extracted and used as the reversal interval period value to characterize the time length required for the signal waveform to change from one direction to another. For example, in a current signal, if the sampling frequency is a fixed value, and a change from rising to falling is detected between the 100th and 120th sampling points, then the interval between this pair of reversal points is the time corresponding to 20 sampling points, which is the reversal interval period value of this segment of the signal. Frequent and regular reversal intervals often indicate that the signal is affected by periodic interference. Identifying the direction of amplitude change reveals the dynamic trend of the waveform, reversal points capture the inflection point characteristics of the signal, and reversal intervals quantify the rhythm of these inflection points, providing a crucial time scale for subsequent judgment on whether there is regular distortion in the sampled signal.

[0027] In the basic sampling sequence, the positions where the amplitude reaches a local maximum or local minimum are identified as extreme points. Using a unified time base as a reference, the time offset of each extreme point relative to the time base is calculated. By arranging the time offsets of consecutive extreme points, the waveform extreme point delay trajectory is generated.

[0028] In the basic sampling sequence, each sampling point can be traversed sequentially using a sliding window, and the amplitude of the current point is compared with the amplitudes of its several neighboring sampling points. When the amplitude of the current sampling point is simultaneously higher than its two adjacent sampling points, it can be identified as a local maximum; conversely, when the amplitude of the current sampling point is simultaneously lower than its two adjacent sampling points, it can be identified as a local minimum. The locations of these local extrema are called extreme points. To accurately measure the distribution characteristics of extreme points on the time axis, a unified time reference needs to be established, usually the starting time of the sampling sequence. Then, the sampling time of each extreme point can be compared with the time reference, and the offset of the extreme point relative to the reference time can be calculated, thus obtaining a set of extreme time offset data. These time offsets are then arranged sequentially according to the chronological order of the extreme points to form the waveform extreme point delay trajectory, which is used to characterize the distribution pattern of extreme points in the waveform along the time dimension. For example, if multiple extreme points appear sequentially in a sampling sequence, and their time offsets maintain a constant difference, it indicates that the waveform has strong temporal regularity and may be affected by a stable interference source. Conversely, if the time offset fluctuates significantly, it suggests that the waveform is disturbed and unstable. Local maxima and minima reflect the peak-valley variations of the signal. A unified time reference ensures the consistency of the time offset, and the time offset provides information on the rhythm of the poles. The waveform pole delay trajectory integrates this information to analyze the correlation between waveform regularity and external disturbances, which is an important basis for identifying regular distortion phenomena.

[0029] S2. When the sampled signal has regular distortion, extract the distortion confirmation point set from the continuous feature sequence, and calculate the feature deviation factor, trend residual vector and stability window convergence based on each distortion confirmation point to form an abnormal morphology mapping group, which is used to characterize the abnormal electrical parameters when the sampled signal has regular distortion under the influence of electromagnetic interference. In this embodiment, S2 specifically includes the following steps: S201. When the sampled signal has regular distortion, perform point-by-point analysis along the time sequence of the continuous feature sequence. When the amplitude density curve corresponding to the same time position shows a repeated change rhythm in multiple adjacent time intervals, the reversal interval period value maintains the same time interval distribution in multiple adjacent time intervals, and the waveform pole delay trajectory shows a continuous and consistent time offset trend in multiple adjacent time intervals, the data at the corresponding time position is determined as the distortion confirmation point, and multiple distortion confirmation points are arranged in time sequence to form a distortion confirmation point set. When the sampled signal exhibits regular distortions, point-by-point analysis along the temporal sequence of continuous feature signals can be achieved using a sliding time window approach. This involves sliding a window with a fixed time width from front to back, extracting the changes in the feature curve frame by frame. Within each window, the amplitude density curve, the inversion interval period value, and the waveform pole delay trajectory corresponding to the current time point are extracted, and their trends within the current and preceding / following time intervals are analyzed. When the amplitude density curve exhibits an approximately periodic fluctuation rhythm across multiple consecutive time windows, the inversion interval period value maintains a consistent time interval distribution across multiple adjacent time periods, and the waveform pole delay trajectory continuously shifts in the same direction with a relatively stable shift rate within these windows, it can be determined that the three features corresponding to that time point possess stable repeatability, directional consistency, and rhythmic characteristics. This time point can then be identified as a distortion confirmation point. By performing a complete scan of the entire continuous feature sequence, multiple time points satisfying the above three characteristic behaviors are extracted sequentially according to the sampling time order and organized into a distortion confirmation point set for subsequent characterization of electrical parameter anomalies.

[0030] The amplitude density curve exhibits a repetitive rhythmic variation across multiple adjacent time intervals, indicating that at the same time point, this feature manifests as alternating high-density and low-density regions with a fixed period, reflecting a regular and repetitive structure in the voltage and current amplitude changes. The reversal interval period value maintains the same time interval distribution across multiple time intervals, indicating that the frequency of current or voltage waveform reversal is stable and does not fluctuate randomly with external disturbances, potentially representing a stable waveform distortion formed under controlled interference. The waveform pole delay trajectory exhibits a continuous and consistent time offset trend across multiple time intervals, indicating that extreme points are systematically stretched or compressed on the time axis, characterizing a consistent drift caused by a fixed phase influence on the waveform. The distortion confirmation point is the time point that simultaneously satisfies these three conditions, possessing high representativeness and capable of capturing structural distortion behavior in data under interference. The distortion confirmation point set arranges these time points chronologically, forming a set of key nodes for subsequent analysis. Through this set, targeted data extraction and distortion characterization paths can be constructed, thereby enabling the identification of deep structural anomalies in the sampled signal.

[0031] S202. For each twist confirmation point in the twist confirmation point set, based on the feature value of the continuous feature sequence corresponding to that time position, calculate the offset of the feature value relative to the overall distribution position of the continuous feature sequence to obtain the feature deviation factor. At the same time, calculate the difference distribution between the trend of continuous feature change before and after that time position and the overall trend of continuous feature sequence change to obtain the trend residual vector. And within the continuous time window centered on the twist confirmation point, calculate the concentration of feature change to obtain the stability window convergence. S203. The feature deviation factor, trend residual vector and stability window convergence corresponding to each distortion confirmation point are combined and arranged according to the time order of the distortion confirmation point in the continuous feature sequence to form an abnormal morphology mapping group, which is used to characterize the abnormal electrical parameters when the sampled signal exhibits regular distortion under the influence of electromagnetic interference.

[0032] At each distortion confirmation point, the obtained feature deviation factor, trend residual vector, and stability window convergence can be considered as characteristic behaviors specific to that time location. By mapping and combining these three types of indicators from multiple distortion confirmation points in chronological order, a complete and continuous chain of anomaly representations can be constructed, forming a feature set with a sequential structure. This set not only presents the evolutionary trajectory of the disturbance's influence over time but also reflects the numerical shifts, trend differences, and local stability changes caused by the disturbance in the feature dimension. This process can be represented using a three-dimensional tensor, with different dimensions corresponding to the time index, feature type, and numerical distribution, respectively. For example, if the feature deviation factor of multiple distortion confirmation points continues to increase over a certain period, while the trend residual vector exhibits directional fluctuations and the stability window convergence gradually decreases, a typical disturbance evolution process can be formed, and the three types of indicators together constitute the "behavioral fingerprint" of this process.

[0033] Anomaly morphology mapping sets are time-series-based feature fusion collections used to characterize the regular distortion features of sampled signals under the influence of electromagnetic interference from multiple dimensions. Essentially, it is a data representation form with multi-index joint encoding. Feature deviation factors are used to measure the magnitude of point value shifts, trend residual vectors are used to measure the degree of trend disruption caused by interference on the trajectory, and stability window convergence is used to reflect the local stability degradation characteristics under the influence of interference. This mapping set connects the local anomaly features of each distortion confirmation point in time sequence, forming a complete anomaly morphology path, which can be regarded as the "feature trace" left by electrical parameter disturbances in the sampled signal. In subsequent processing, this mapping set can serve as a core reference for identifying anomaly types, judging interference intensity, and tracing the evolution of interference sources.

[0034] In this embodiment, S202 specifically refers to: For each twist confirmation point in the twist confirmation point set, the feature value of the continuous feature sequence corresponding to that time position is extracted, and the position of the feature value is compared with the feature value distribution formed by the continuous feature sequence over the complete time range. By calculating the distance magnitude of the feature value relative to the distribution center, the feature deviation factor used to characterize the degree of local feature deviation is obtained. At each distortion confirmation point in the distortion confirmation point set, the feature values ​​of the continuous feature sequence corresponding to that time position are first extracted. These feature values ​​can be the amplitude density value, the reversal interval period value, or the waveform pole delay value at that moment. These values ​​collectively reflect the instantaneous characteristics of the voltage and current waveforms at that moment. Subsequently, all similar feature values ​​of the continuous feature sequence over the complete time range are statistically analyzed to form the numerical distribution interval of the feature value in the global range. For example, for the amplitude density curve, all amplitude density values ​​over the entire time period can be taken to form a distribution dataset, and the mean, mode, or median of the distribution can be calculated as a representative distribution center. Then, the feature value of the current distortion confirmation point is compared with the position of this global distribution center, and the deviation distance of the value from the distribution center in the numerical space is calculated. This distance magnitude is the feature deviation factor, used to characterize the degree of anomalousness of the current feature value in the global statistical features. The larger the deviation factor, the more the feature behavior of the distortion confirmation point deviates from the overall operating trend, which may represent an abnormal signal structure under the influence of interference. In this way, the dispersion of local distortion behavior in the global feature environment can be quantified, thus providing a mathematical basis and judgment scale for subsequent distortion identification.

[0035] Around the same twist confirmation point, the continuous feature change trajectory within the continuous time interval before and after the time position is extracted, and the trend of the change trajectory is compared with the change trajectory of the continuous feature sequence over the entire time range. By describing the distribution of the differences between the two types of change trajectories in terms of directionality and change amplitude, the trend residual vector is obtained. To analyze whether the changes in characteristics at a specific distortion confirmation point are abnormal, it is necessary to extract the continuous characteristic change trajectory within a continuous time interval before and after that point. This trajectory consists of the trend of characteristic values ​​on the time axis, reflecting the dynamic evolution of characteristics before and after that point. Correspondingly, it is necessary to extract the global change trajectory from the continuous characteristic sequence over a complete time range as a reference benchmark for characteristic changes under normal system operation. The trend comparison process involves synchronously comparing the local and global change trajectories within the same time span, focusing on whether the direction of their numerical increases or decreases is consistent and whether there are significant differences in the rate of change. This comparison is not based on single-point values ​​but on the continuous trend in the time series. Directional differences are manifested as a local increase in a characteristic curve while the global curve decreases, or vice versa; differences in amplitude are manifested as the amplitude of local characteristic values ​​being significantly higher or lower than the average bandwidth of the overall trend. By quantifying these directional and amplitude deviations within the time interval and using distributional statistics to form a numerical vector, the trend residual vector can be obtained. This trend residual vector is a multi-dimensional expression of anomalies that can reveal the degree of deviation of the distortion confirmation point from the overall pattern in continuous evolution, which helps to further identify the systematic deformation characteristics caused by disturbances.

[0036] Centered on the time position corresponding to the same twist confirmation point, the feature values ​​of the continuous feature sequence within a continuous time window are selected. The concentration and dispersion of the feature value changes within the time window are statistically analyzed, and the statistical results are used as the stability window convergence to characterize the clustering state of feature changes in the neighborhood of the twist confirmation point.

[0037] Centered on the time position corresponding to a certain twist confirmation point, several sampling points can be extended forward and backward to construct a continuous time window of fixed length. Each moment within the window corresponds to a continuous feature sequence value, and these feature values ​​constitute the feature change sequence within the time window. To measure whether the feature values ​​within this window tend to converge or diverge under some pattern or disturbance, it is necessary to statistically analyze the similarity between all feature values ​​in the sequence. Specifically, this can be measured by calculating indicators such as the value range, local variance, and moving average bandwidth to assess the concentration and dispersion of changes. If most feature values ​​are closely clustered within a certain numerical range, it indicates that the change trend within the window is relatively concentrated; conversely, it indicates that the feature value distribution within the window is relatively dispersed. The statistical result that simultaneously expresses concentration and dispersion is defined as the stability window convergence, which reflects the dynamic change stability characteristics within the neighborhood of the twist confirmation point. For example, if the feature values ​​before and after a twist confirmation point show continuous oscillations but are concentrated within a certain numerical range, it indicates that the feature values ​​tend to stabilize under disturbance, and the stability window convergence is high; conversely, if the feature values ​​fluctuate frequently and drastically, the value is low. By calculating the convergence of the stable window, we can more accurately identify whether the changes after disturbance have entered a specific stable distortion mode, thus providing refined numerical support for the subsequent characterization of electrical parameter anomalies.

[0038] S3. Combine and process the various indicators in the abnormal morphology mapping group to generate the amplitude response offset rate, morphological coupling score and total stability deviation, and calculate the credibility score of the operating status to identify whether the operating status of the circuit breaker is distorted. In this embodiment, S3 specifically includes the following steps: S301. Extract the feature deviation factor, trend residual vector and stable window convergence corresponding to each distortion confirmation point in the abnormal morphology mapping group in chronological order, and construct the corresponding index arrangement sequence while maintaining time alignment, as the basic feature data source for subsequent combination processing. In the abnormal morphology mapping group, for each distortion confirmation point, the characteristic deviation factor, trend residual vector, and stability window convergence corresponding to that time position are extracted sequentially. These three indicators are arranged in chronological order to ensure consistency of indicator values ​​across time points. By setting a unified time index, the characteristic deviation factor, trend residual vector, and stability window convergence are aligned within the same time dimension, constructing a multi-dimensional, time-consistent indicator arrangement structure. This structure can be implemented as a two-dimensional matrix, where each column represents an indicator type and each row represents the time position of a distortion confirmation point, ensuring data integrity and processing synchronization. The purpose of this operation is to fully preserve temporal characteristics, facilitating subsequent statistical processing, trend analysis, and indicator fusion calculations in the time domain, thereby ensuring that the temporal correlation between feature quantities is not disrupted. For example, when there are five distortion confirmation points within a sampling interval, the system will extract three indicator values ​​from each distortion confirmation point, forming a five-row, three-column arrangement structure as the basic data input for calculating parameters such as amplitude response offset.

[0039] The feature deviation factor is used to characterize the degree of anomaly of the feature value at the current time point relative to the distribution over the entire time period. The trend residual vector is used to quantify the offset angle between the feature trend at the current point and the global trend. The stability window convergence degree is used to describe the stability of feature fluctuations within a time window near the current point. The indicator arrangement sequence refers to a data sequence set formed by organizing three indicators at multiple distortion confirmation points in chronological order. This set maintains both the internal temporal evolution trajectory of each indicator and the numerical correspondence between them at the same time position. Its core technical features lie in the time alignment of the data and the multi-dimensional joint organization capability of the indicators, giving the originally scattered time-series indicators structural integrity and computability, providing a reliable foundation for subsequent calculation of multi-indicator fusion scores. This technical arrangement breaks through the traditional method of analyzing only a single feature sequence, enhancing the ability to interpret the linkage between indicators, and is one of the core supporting means for judging the distortion characteristics of the circuit breaker's operating status.

[0040] S302. Based on the change trajectory of different types of indicators in the indicator arrangement sequence, calculate the offset ratio between the maximum value and the average value of the feature deviation factor to generate the amplitude response offset rate; perform vector angle aggregation analysis on the consistency of the residual direction at each time in the trend residual vector to generate the morphological coupling score; and perform fluctuation amplitude statistics within the sliding interval on the stability window convergence to generate the total stability deviation. S303. Normalize the amplitude response offset rate, morphological coupling score and total stability deviation, and perform weighted accumulation based on the preset multi-dimensional index fusion weight to generate a credibility score value for the operating status; when the credibility score value is lower than the distortion identification threshold, it is identified that the operating status of the circuit breaker has been distorted.

[0041] When calculating the reliability score of the operating status, the amplitude response offset rate, morphological coupling score, and total stability deviation are first normalized to a uniform scale. This is typically achieved using max-min scaling or Z-score normalization to map the values ​​of each indicator to the same range, eliminating interference from differences in units and numerical ranges in subsequent calculations. After normalization, the three indicators are weighted and summed according to the system's preset multi-dimensional indicator fusion weights. Each weight is configured based on its importance to the operating status; for example, the morphological coupling score has a higher weight in interference identification, followed by the total stability deviation, and then the amplitude response offset rate. The final reliability score is generated by summing the weighted normalized indicators. This score reflects the overall reliability of the current operating status of the circuit breaker. A lower reliability score indicates a greater deviation between the feature data and the stable state. Combined with a distortion identification threshold set in the system, when the score falls below this threshold, it is considered that the current sampled data of the circuit breaker can no longer accurately reflect the actual operating status, thus determining that the operating status has been distorted.

[0042] Indicator normalization involves numerically mapping the amplitude response offset rate, morphological coupling score, and total stability deviation. Common methods include linearly scaling the values ​​to between 0 and 1 to achieve equal weighting between different indicators. Preset multi-dimensional indicator fusion weights are a set of numerical proportions reflecting the contribution of each indicator to the overall evaluation, guiding the data fusion direction in the weighted processing stage. Weighted accumulation processing involves multiplying each of the three normalized indicators by its corresponding fusion weight and summing the results to obtain the final credibility score. The credibility score for the operating status is a numerical expression representing the comprehensive deviation of electrical parameter characteristics in amplitude, trend, and stability dimensions within the current time window. The distortion identification threshold is an empirically set boundary value used to distinguish between credible and distorted data. When the credibility score is below this threshold, it indicates a significant difference between the sampled characteristics and the stable operating status, requiring further control strategies or alarm mechanisms. This structured scoring mechanism effectively supports the objective identification and risk perception of circuit breaker status under electromagnetic interference conditions.

[0043] In this embodiment, S302 specifically refers to: Extract the index arrangement sequence composed of feature deviation factors, calculate the ratio of the difference between the maximum feature deviation factor value and the average feature deviation factor value in the index arrangement sequence, and use this ratio as the amplitude response offset rate to measure the difference in response intensity of local electrical parameter anomalies in the overall sequence. In the index sequence composed of characteristic deviation factors, the characteristic deviation factors corresponding to each distortion confirmation point are first extracted and arranged in chronological order to form a numerical sequence with temporal characteristics. Then, the maximum characteristic deviation factor value in the entire sequence is calculated, representing the value corresponding to the point with the most significant deviation across all time points. Simultaneously, the arithmetic mean of all characteristic deviation factor values ​​is calculated to characterize the overall level of characteristic deviation over the entire time period. Next, the difference between the maximum and the average value is proportionally calculated to obtain a ratio reflecting the degree of local anomaly amplification; this ratio is the amplitude response offset rate. This index measures whether there is a significant deviation in local electrical parameters at a certain distortion confirmation point that is far greater than the overall distribution, accurately indicating whether the local response possesses significant abnormal characteristics. For example, if most characteristic deviation factor values ​​in a sequence are distributed between 1.0 and 1.5, and the characteristic deviation factor reaches 3.0 at a certain moment, the ratio of its difference from the average value will be significantly higher, resulting in a large amplitude response offset rate, meaning that there may be a strong local distortion response in the circuit breaker's operating state at that moment. This processing method not only possesses data sensitivity but also enhances the expression of abnormal electrical parameters by quantifying differences in amplitude, avoiding ambiguous judgments of the overall trend. It is an important technical means for identifying local abnormal shocks.

[0044] The index arrangement sequence composed of trend residual vectors is extracted. For each trend residual vector, the angle between it and the main trend direction of the whole sequence is calculated. Cosine aggregation is performed on all angles to obtain the morphological coupling score that reflects the consistency of residual direction, which is used to characterize the consistency of trend deviation at multiple points. In the index sequence composed of trend residual vectors, each trend residual vector represents the deviation direction and degree of the continuous characteristic change trajectory of a certain distortion confirmation point within a specific time period compared to the entire sequence. To assess the consistency of the changing trends of multiple distortion points, it is necessary to first determine the main direction of the changing trend in the entire continuous characteristic sequence. This can be achieved by extracting the direction vector as the main trend direction of the entire sequence through linear fitting or principal component analysis of all trend residual vectors. Subsequently, for each trend residual vector, the angle between it and the main trend direction is calculated to represent the degree of difference between the characteristic change direction at that moment and the overall trend. When the angle is small, it indicates that the trend at that point is relatively consistent with the main trend; when the angle is large, it indicates that the trend at that point may deviate significantly from the main trend. After extracting the cosine values ​​of all angles and performing aggregation, the overall consistency of all trend directions can be reflected by the average cosine value, and the result obtained is the morphological coupling score. The closer the score is to 1, the more consistent the distortion confirmation points are in the trend direction, indicating that they may be synchronously affected by the same interference source; the lower the score, the greater the trend difference, indicating that the interference effect has randomness or locality. By using this vector angle analysis and cosine convergence processing, we can accurately identify whether abnormal trends constitute systematic morphological distortions, avoid misjudgments caused by relying solely on single-point judgments, and thus more effectively characterize the consistency of multi-point trend deviations.

[0045] The index sequence composed of the convergence of the stability window is extracted, the fluctuation amplitude between the convergence values ​​is calculated in adjacent time intervals, and the total stability deviation is constructed based on the maximum fluctuation amplitude within the statistical range to evaluate the influence of local stability changes on the overall feature sequence.

[0046] In constructing the total stability deviation, the convergence values ​​of the stability windows corresponding to all torsion confirmation points are first extracted and arranged in chronological order to form an index sequence composed of stability window convergence values. Each convergence value represents the degree of clustering of characteristic change values ​​within a continuous time window centered on a certain torsion confirmation point, which can usually be quantified by the coefficient of variation or local standard deviation of the characteristic values. Subsequently, for the convergence values ​​in adjacent time intervals, their numerical differences are calculated one by one to obtain the fluctuation amplitude reflecting the degree of stability fluctuation. These fluctuation amplitudes reflect the stability fluctuation of characteristic changes on the time axis; the larger the fluctuation amplitude, the more unstable the electrical parameter state of the system is within that interval. To quantify the concentrated manifestation of instability in the overall sequence, all fluctuation amplitude values ​​can be statistically analyzed over the entire time range, and the maximum value can be selected as the evaluation benchmark to construct the total stability deviation reflecting extreme stability deviations. The total stability deviation is used to measure the intensity of the impact of local disturbances on the global scale. If its value is significantly high, it indicates that there is a violent disturbance in the characteristic sequence, which may be related to nonlinear disturbances caused by strong electromagnetic interference. This method can effectively identify the imbalance trend of stability characteristics, thereby supporting the accurate identification and abnormal early warning of the circuit breaker's operating status.

[0047] S4. Based on the relationship between the credibility score and the dynamic threshold mapping coordinate, generate state classification results and divide the running state into the data credibility zone, the data critical zone and the data distortion zone. In this embodiment, S4 specifically refers to: Based on the calibrated operating states and corresponding credibility scores in historical samples, a mapping coordinate system between credibility scores and dynamic thresholds is constructed. By extracting the correspondence between credibility scores and distortion levels under different working conditions in multiple time periods, a segmented fitting method is used to generate multiple dynamic threshold curves, forming a mapping coordinate system that reflects the relationship between score changes and state division boundaries. To construct the mapping coordinates between credibility scores and dynamic thresholds, it is first necessary to obtain calibrated operating status labels and their corresponding credibility scores from a large number of historical samples. These samples typically originate from electrical parameter signals collected under different operating environments and electromagnetic interference intensities. By statistically analyzing the distribution characteristics of credibility scores for each operating status, and combining this with manually or rule-based labeling of distortion levels, a quantitative correspondence between the scores and the actual degree of distortion is established. Subsequently, over multiple time periods with significant differences in operating conditions, these correspondences are extracted and organized. Piecewise linear fitting or polynomial fitting methods are used to model different state boundaries, constructing dynamic threshold curves suitable for different operating conditions. All curves are combined into a unified mapping coordinate system using the boundary points of state division as the connection basis. This enables state classification mapping queries for any credibility score, possessing dynamic adaptation and time-varying adjustment capabilities, thereby meeting the state identification requirements under multi-source interference.

[0048] In this process, the historically calibrated operating states and corresponding credibility scores refer to the electrical parameter data records whose true operating nature has been clarified through offline analysis, along with the credibility scores calculated from them. These data constitute the basic input for mapping modeling. The mapping coordinates between credibility scores and dynamic thresholds are used to reflect the reference positions of the scores at different state classification boundaries, for subsequent classification judgment. The correspondence between credibility scores and distortion levels under different operating conditions refers to the mapping distribution characteristics formed by the same score potentially corresponding to different distortion levels in scenarios with different external parameters such as temperature, voltage, and load. The piecewise fitting method divides the score value interval into partitions, making the mapping relationship within each segment more consistent with the actual distribution trend, and improving modeling accuracy through piecewise modeling. Multi-segment dynamic threshold curves refer to the score value boundary model expressed in the form of multiple piecewise functions, used to reflect the critical range between each distortion level. The mapping coordinate system refers to a multi-dimensional data structure that integrates multiple dynamic threshold curves and score distribution data, used to achieve automatic correspondence and matching between score values ​​and state boundaries, supporting real-time classification judgment of operating states.

[0049] Map the current credibility score to the constructed mapping coordinate system. Based on the corresponding interval position of the credibility score on the dynamic threshold curve, determine its relative position in the mapping coordinate system and obtain the numerical relationship between the current credibility score and the dynamic threshold. To classify and determine the credibility score of the current operating state, the score needs to be input into a pre-constructed mapping coordinate system for position matching. This system contains multiple dynamic threshold curves, each representing a state classification boundary under a given operating condition. Projecting the current credibility score onto this coordinate system first identifies the dynamic threshold curve segment to which the current operating condition belongs, then compares the score with the curve to pinpoint its specific interval on the curve. If the score is above the upper threshold, the state is considered stable and reliable; if it is below the lower threshold, significant distortion exists; and if it is in the transition zone between the two thresholds, the state is in a critical fluctuation state. This mapping process can be automated using index lookup or interval matching algorithms. By searching for the corresponding segment in the mapping coordinate system, real-time linkage between the score and the state level is achieved, offering high efficiency and versatility.

[0050] In this process, the corresponding interval position of the credibility score on the dynamic threshold curve refers to the numerical landing point of the score between the state boundaries represented by the multi-segment function curves, which is used as the initial classification basis for determining its state level. The relative relationship refers to the distance distribution, relative position, and numerical offset between the score and the upper and lower limits, as well as the inflection point of the curve; it reflects the current score's position within the entire scoring system. The numerical relationship between the current credibility score and the dynamic threshold is based on quantitative relationships such as ratios, differences, or normalized weights after coordinate mapping and interval identification are completed, and is used for subsequent logical judgments of state division and classification strategy formulation.

[0051] The status is classified based on the position of the current credibility score in the mapped coordinate system. If the credibility score is in the high score range of the dynamic threshold curve, the running status is classified as the data credibility zone; if the credibility score is in the fluctuating transition range of the dynamic threshold curve, the running status is classified as the data critical zone; if the credibility score is below the lower boundary of the dynamic threshold curve, the running status is classified as the data distortion zone, and the corresponding status classification result is generated.

[0052] To accurately identify the operating status of the circuit breaker, the specific position of the current reliability score in the mapped coordinate system needs to be compared with a preset dynamic threshold curve. The score interval is determined based on the score's location, and the status is classified accordingly. When the reliability score is higher than the upper boundary of the dynamic threshold curve corresponding to the current operating conditions, i.e., in the high-score interval of the curve, it indicates that the current operating status has high stability and reliability, and the result is determined as the data reliability zone. If the reliability score falls in the fluctuating transition interval in the middle of the curve, it indicates that the current status has some disturbance or uncertainty, but has not yet formed a serious anomaly, and the result is determined as the data critical zone. When the score is lower than the lower boundary of the curve, it means that the current status exhibits characteristics highly similar to historical distorted samples, and the circuit breaker's operating status has entered the data distortion zone. The status classification process can be completed through an interval comparison algorithm, combined with logical judgment conditions to automatically output the final status result label, which serves as the basis for subsequent alarm processing or operational adjustments of the system.

[0053] State classification is a multi-level partitioning decision-making process based on the numerical relationship between confidence scores and dynamic threshold curves. It identifies whether the currently sampled features belong to a stable, critical, or distorted state. The high-score range of the dynamic threshold curve typically corresponds to the numerical region where the score value is above a specific confidence interval, representing the high stability of the system under multiple operating conditions; this is classified as the data confidence zone. The fluctuation transition range, where the score value lies between the upper and lower boundaries, contains some uncertainty and is classified as the data critical zone, indicating potential interference or trend shifts. The region below the lower boundary is defined as the data distortion zone, a marker area indicating significant electrical parameter distortion under electromagnetic interference. The criteria for these zone divisions are based on statistical analysis of a large amount of historical sample data and model fitting results, and are adapted to different operating environments through a dynamic threshold adjustment mechanism, effectively improving the accuracy and robustness of state identification.

[0054] S5. Perform dynamic control operations based on the state classification results. In the data trust zone, maintain the original sampling and communication frequency. In the data critical zone, increase the sampling frequency and pause the remote control response. In the data distortion zone, trigger signal screening and mark the current data as an abnormal source, and feed it back to the continuous feature sequence acquisition process to form a closed loop.

[0055] In this embodiment, S5 specifically refers to: When the state classification result is in the data confidence zone, the control sampling system maintains the current sampling frequency of the voltage sampling sequence and the current sampling sequence, and keeps the current LoRa communication frequency unchanged, so that the data continues to enter the continuous feature sequence acquisition process under stable conditions, ensuring the time consistency and signal integrity in the feature generation process. When the state classification result indicates a reliable data region, the sampling system maintains the sampling frequencies of the current voltage and current sampling sequences to ensure temporal uniformity of the acquired data across the continuous time axis. This operation can be performed by the local controller, based on the judgment results output by the state classification module, by configuring internal parameters to maintain the original period settings of the sampling timing module, ensuring that subsequent sampling actions are executed strictly according to the predetermined frequency. Simultaneously, the communication management module maintains the operating frequency and transmission interval of the LoRa communication channel stably, without triggering any communication rate adjustments or renegotiation operations. Because the data variation is small and the feature trajectory is stable at this stage, data continues to be collected at the default pace and sent to the continuous feature sequence generation channel, ensuring system continuity and temporal consistency of feature extraction while ensuring computational resource efficiency. For example, at a sampling frequency of 250Hz and a data upload frequency of once per minute, the system does not need to adjust the acquisition and transmission strategies; it directly appends newly acquired data to the existing sequence in chronological order to continuously construct feature results such as amplitude density curves, inversion interval period values, and waveform pole delay trajectories.

[0056] Voltage and current sampling sequences are time-series datasets formed by sampling based on the same time base, representing the electrical parameter status of circuit breakers at different points in time. Sampling frequency refers to the number of samples collected per unit time, directly determining data resolution and feature parsing capability. LoRa communication frequency corresponds to the data upload frequency or bandwidth allocation strategy of the wireless communication module, affecting the remote transmission rhythm of data. Time consistency refers to the uniformity of continuous sampling data along the time axis, avoiding time gaps caused by frequency jumps. Signal integrity refers to the undistorted and uninterrupted nature of the sampled signal during transmission and processing, ensuring the accurate and effective calculation of feature values. The continuous feature sequence acquisition process is the starting module of the entire anomaly detection process, relying on continuous and stable signal input for dynamic feature extraction. Therefore, maintaining the original frequency and communication parameters under reliable data conditions is crucial for ensuring the stable operation of the overall system.

[0057] When the state classification result is a data critical zone, adjust the sampling control unit parameters, increase the sampling frequency of the voltage sampling sequence and the current sampling sequence to a high-precision sampling level, and shield the response mechanism of the remote control signal, suspend the transmission and execution of remote control commands, and ensure that the integrity of the sampling signal is not affected by control intervention, so as to enhance the accuracy of anomaly identification. When the state classification result indicates a data critical region, the system adjusts the internal sampling period parameters of the sampling control unit to increase the sampling frequency of the voltage and current sampling sequences to a high-precision sampling level, for example, from the conventional 250Hz to 500Hz or higher. This allows for more detailed capture of waveform changes within each cycle, thereby enhancing the response capability to minute electrical parameter fluctuations. Simultaneously, to prevent remote control signals from interfering with electrical parameter acquisition or causing sudden changes in the power grid state during anomaly identification, the system disables the remote control signal parsing module or locks the control interruption processing channel, effectively shielding the remote command transmission and execution process. This ensures that the sampling process is not interrupted by external operations in the data critical state, helping to stably acquire key data segments from the continuous feature sequence used for subsequent analysis. For example, in a typical application, when minor arcing anomalies or transient interference fluctuations occur in power equipment, temporarily increasing the sampling accuracy and pausing the control response can significantly improve the clarity of identifying electrical parameter change trends and avoid misjudgments.

[0058] Voltage and current sampling sequences are high-frequency signal streams that record the changes of various electrical parameters in the circuit breaker over time, used to analyze whether the electrical state is abnormal. The sampling frequency determines how many valid data points are collected per second; higher precision sampling levels correspond to shorter sampling periods and finer time scales. The sampling control unit is a dedicated control module in the system for managing sampling timing and channel scheduling; its parameter adjustments directly affect the data acquisition strategy. Remote control signals refer to operation commands issued by external systems or higher-level control platforms through the communication module, such as switching operations and mode switching. If these signals are executed during anomaly detection, they may disrupt the local electrical parameter state, thereby interfering with the stability of feature generation and detection results. The response mechanism shielding operation temporarily sets the control signal as an invalid input to ensure that the data stream is not interrupted by external events during sampling. This dynamic control design not only improves sampling integrity but also provides a high-quality data foundation for subsequent anomaly feature analysis.

[0059] When the state classification result is a data distortion area, the signal processing unit is activated to filter the sampled data in the current time period, identify data points with feature distribution offset in the continuous feature sequence, mark such data as an abnormal source, and feed the marking result back to the continuous feature sequence acquisition process to drive the isolation management of abnormal data in subsequent sampling cycles, thus constructing a dynamic control mechanism for state closed-loop control.

[0060] When the state classification result indicates a data distortion zone, the system activates the signal processing unit to perform real-time analysis of the sampled data for the current period, focusing on filtering out data points in the continuous feature sequence that exhibit significant anomalies in statistical distribution, trend continuity, or periodic consistency. The filtering method can involve identifying data points whose feature deviation factors exceed the distribution threshold, or calibrating the location of directional abrupt changes in the trend residual vector, while simultaneously determining the boundaries of abnormal distributions by considering the fluctuation range of the stability window convergence. After identifying suspected abnormal data points, the system marks these data with anomaly tags and forms label information, which is then fed back to the continuous feature sequence acquisition process. Upon receiving the label feedback, the acquisition process dynamically updates the sampling buffer and queue management rules, performing isolation operations on content falling into similar time periods or data feature neighborhoods in the next sampling cycle, thereby preventing abnormal data from affecting the overall data structure and subsequent state discrimination results. For example, in a specific scenario, when electromagnetic interference causes non-physical periodic repetition misalignment of the sampling signal over several cycles, the above processing mechanism can effectively eliminate erroneous segments and correct the acquisition strategy in real time, constructing an adaptive closed-loop control mechanism.

[0061] The signal processing unit refers to a real-time analysis module deployed in a local terminal or edge computing module, capable of judging data quality based on feature trajectories. Sampling data consists of a basic data stream composed of voltage and current sampling sequences, which requires integrity judgment before entering the continuous feature sequence. The continuous feature sequence reflects the feature mapping results of sampled values ​​at different time points. If the feature distribution of its data points deviates from the normal trajectory, such as a sudden interruption of periodic regularity or a drastic jump in the degree of feature deviation, it can be considered an anomaly. Data marked as an anomaly source is not only used for screening in the current time period but also needs to be fed back to the acquisition process to form a "collection-judgment-control-collection" feedback loop. This drives subsequent cycles to achieve diversionary acquisition, segmented processing, and difference modeling for anomaly features, thereby constructing a dynamic control mechanism that supports stable operation of state recognition. This mechanism improves the system's response accuracy and recovery capability to distorted states, ensuring the entire monitoring process has robustness and real-time closed-loop adjustment capabilities.

[0062] The above embodiments can be implemented, in whole or in part, by software, hardware, firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions according to the embodiments of this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. Computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired or wireless means (e.g., infrared, wireless, microwave, etc.). A computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. Available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media. Semiconductor media can be solid-state drives.

[0063] It should be understood that in the various embodiments of this application, the order of the above-mentioned processes does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0064] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0065] In the several embodiments provided in this application, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be an indirect coupling or communication connection through some interfaces, devices, or units, and may be electrical, mechanical, or other forms.

[0066] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0067] In addition, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0068] The above are merely specific embodiments of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for intelligent circuit breaker control and monitoring based on LoRa communication, characterized in that, Specifically, the following steps are included: S1. Continuously acquire the voltage and current sampling sequences of the circuit breaker to generate the amplitude density curve, the inversion interval period value, and the waveform pole delay trajectory, forming a continuous feature sequence to determine whether there is a regular distortion of the sampling signal under the influence of electromagnetic interference. S2. When the sampled signal has regular distortion, extract the set of distortion confirmation points from the continuous feature sequence, and calculate the feature deviation factor, trend residual vector and stability window convergence based on each distortion confirmation point to form an abnormal shape mapping group. S3. Combine and process the various indicators in the abnormal morphology mapping group to generate the amplitude response offset rate, morphological coupling score and total stability deviation, and calculate the credibility score of the operating status to identify whether the operating status of the circuit breaker is distorted. S4. Based on the relationship between the credibility score and the dynamic threshold mapping coordinate, generate state classification results and divide the running state into the data credibility zone, the data critical zone and the data distortion zone. S5. Perform dynamic control operations based on the state classification results. In the data trust zone, maintain the original sampling and communication frequency. In the data critical zone, increase the sampling frequency and pause the remote control response. In the data distortion zone, trigger signal screening and mark the current data as an abnormal source, and feed it back to the continuous feature sequence acquisition process to form a closed loop.

2. The intelligent circuit breaker control and monitoring method based on LoRa communication according to claim 1, characterized in that, S1 specifically includes the following steps: S101. The voltage sampling sequence and current sampling sequence of the circuit breaker are continuously acquired according to a unified time base, so that each voltage sampling value and the current sampling value at the corresponding time are time-aligned, and a basic sampling sequence is formed in the order of sampling time. S102. Based on the amplitude variation relationship between adjacent sampled values ​​in the basic sampling sequence, calculate the amplitude variation density in the continuous sampling interval to generate a variable amplitude density curve; calculate the time interval between adjacent reversal points based on the position where the sampled value change direction reverses to generate a reversal interval period value; calculate the extreme value time offset based on the position change of the amplitude extreme point in the sampling sequence relative to the time base to generate a waveform extreme point delay trajectory. S103. The amplitude density curve, the inversion interval period value, and the waveform pole delay trajectory are arranged in chronological order to form a continuous feature sequence. The periodic repetition degree of the amplitude density curve, the time consistency degree of the inversion interval period value, and the offset stability degree of the waveform pole delay trajectory are compared in the continuous feature sequence. When the amplitude density curve shows a fixed periodic repetition relationship, the inversion interval period value maintains a time consistency relationship, and the waveform pole delay trajectory maintains a stable offset relationship, it is determined that the sampled signal has a regular distortion under the influence of electromagnetic interference.

3. The intelligent circuit breaker control and monitoring method based on LoRa communication according to claim 2, characterized in that, S102 specifically refers to: In the basic sampling sequence, the amplitude difference is calculated for adjacent sampled values, and the amplitude difference is arranged in the order of sampling time. By statistically analyzing the distribution density of amplitude differences within continuous sampling intervals, an amplitude density curve reflecting the frequency of amplitude changes is formed. In the basic sampling sequence, the position where the amplitude change direction changes from rising to falling or from falling to rising is identified, the position where the change direction reverses is determined as the reversal point, and the reversal interval period value is calculated based on the sampling time interval corresponding to adjacent reversal points. In the basic sampling sequence, the positions where the amplitude reaches a local maximum or local minimum are identified as extreme points. Using a unified time base as a reference, the time offset of each extreme point relative to the time base is calculated. By arranging the time offsets of consecutive extreme points, the waveform extreme point delay trajectory is generated.

4. The intelligent circuit breaker control and monitoring method based on LoRa communication according to claim 1, characterized in that, S2 specifically includes the following steps: S201. When the sampled signal has regular distortion, perform point-by-point analysis along the time sequence of the continuous feature sequence. When the amplitude density curve corresponding to the same time position shows a repeated change rhythm in multiple adjacent time intervals, the reversal interval period value maintains the same time interval distribution in multiple adjacent time intervals, and the waveform pole delay trajectory shows a continuous and consistent time offset trend in multiple adjacent time intervals, the data at the corresponding time position is determined as the distortion confirmation point, and multiple distortion confirmation points are arranged in time sequence to form a distortion confirmation point set. S202. For each twist confirmation point in the twist confirmation point set, based on the feature value of the continuous feature sequence corresponding to that time position, calculate the offset of the feature value relative to the overall distribution position of the continuous feature sequence to obtain the feature deviation factor. At the same time, calculate the difference distribution between the trend of continuous feature change before and after that time position and the overall trend of continuous feature sequence change to obtain the trend residual vector. And within the continuous time window centered on the twist confirmation point, calculate the concentration of feature change to obtain the stability window convergence. S203. The feature deviation factor, trend residual vector and stability window convergence corresponding to each distortion confirmation point are combined and arranged according to the time order of the distortion confirmation point in the continuous feature sequence to form an abnormal morphology mapping group, which is used to characterize the abnormal electrical parameters when the sampled signal exhibits regular distortion under the influence of electromagnetic interference.

5. The intelligent circuit breaker control and monitoring method based on LoRa communication according to claim 4, characterized in that, S202 specifically refers to: For each twist confirmation point in the twist confirmation point set, the feature value of the continuous feature sequence corresponding to that time position is extracted, and the position of the feature value is compared with the feature value distribution formed by the continuous feature sequence over the complete time range. By calculating the distance magnitude of the feature value relative to the distribution center, the feature deviation factor used to characterize the degree of local feature deviation is obtained. Around the same twist confirmation point, the continuous feature change trajectory within the continuous time interval before and after the time position is extracted, and the trend of the change trajectory is compared with the change trajectory of the continuous feature sequence over the entire time range. By describing the distribution of the differences between the two types of change trajectories in terms of directionality and change amplitude, the trend residual vector is obtained. Centered on the time position corresponding to the same twist confirmation point, the feature values ​​of the continuous feature sequence within a continuous time window are selected. The concentration and dispersion of the feature value changes within the time window are statistically analyzed, and the statistical results are used as the stability window convergence to characterize the clustering state of feature changes in the neighborhood of the twist confirmation point.

6. The intelligent circuit breaker control and monitoring method based on LoRa communication according to claim 1, characterized in that, S3 specifically includes the following steps: S301. Extract the feature deviation factor, trend residual vector and stable window convergence corresponding to each distortion confirmation point in the abnormal morphology mapping group in chronological order, and construct the corresponding index arrangement sequence while maintaining time alignment, as the basic feature data source for subsequent combination processing. S302. Based on the change trajectory of different types of indicators in the indicator arrangement sequence, calculate the offset ratio between the maximum value and the average value of the feature deviation factor to generate the amplitude response offset rate; perform vector angle aggregation analysis on the consistency of the residual direction at each time in the trend residual vector to generate the morphological coupling score; and perform fluctuation amplitude statistics within the sliding interval on the stability window convergence to generate the total stability deviation. S303. Normalize the amplitude response offset rate, morphological coupling score and total stability deviation, and perform weighted accumulation based on the preset multi-dimensional index fusion weight to generate a credibility score value for the operating status; when the credibility score value is lower than the distortion identification threshold, it is identified that the operating status of the circuit breaker has been distorted.

7. The intelligent circuit breaker control and monitoring method based on LoRa communication according to claim 6, characterized in that, S302 specifically refers to: Extract the index arrangement sequence composed of feature deviation factors, calculate the ratio of the difference between the maximum feature deviation factor value and the average feature deviation factor value in the index arrangement sequence, and use this ratio as the amplitude response offset rate to measure the difference in response intensity of local electrical parameter anomalies in the overall sequence. The index arrangement sequence composed of trend residual vectors is extracted. For each trend residual vector, the angle between it and the main trend direction of the whole sequence is calculated. Cosine aggregation is performed on all angles to obtain the morphological coupling score that reflects the consistency of residual direction, which is used to characterize the consistency of trend deviation at multiple points. The index sequence composed of the convergence of the stability window is extracted, the fluctuation amplitude between the convergence values ​​is calculated in adjacent time intervals, and the total stability deviation is constructed based on the maximum fluctuation amplitude within the statistical range to evaluate the influence of local stability changes on the overall feature sequence.

8. The intelligent circuit breaker control and monitoring method based on LoRa communication according to claim 1, characterized in that, S4 specifically refers to: Based on the calibrated operating states and corresponding credibility scores in historical samples, a mapping coordinate system between credibility scores and dynamic thresholds is constructed. By extracting the correspondence between credibility scores and distortion levels under different working conditions in multiple time periods, a segmented fitting method is used to generate multiple dynamic threshold curves, forming a mapping coordinate system that reflects the relationship between score changes and state division boundaries. Map the current credibility score to the constructed mapping coordinate system. Based on the corresponding interval position of the credibility score on the dynamic threshold curve, determine its relative position in the mapping coordinate system and obtain the numerical relationship between the current credibility score and the dynamic threshold. The status is classified based on the position of the current credibility score in the mapped coordinate system. If the credibility score is in the high score range of the dynamic threshold curve, the running status is classified as the data credibility zone; if the credibility score is in the fluctuating transition range of the dynamic threshold curve, the running status is classified as the data critical zone; if the credibility score is below the lower boundary of the dynamic threshold curve, the running status is classified as the data distortion zone, and the corresponding status classification result is generated.

9. The intelligent circuit breaker control and monitoring method based on LoRa communication according to claim 1, characterized in that, S5 specifically refers to: When the state classification result is in the data confidence zone, the control sampling system maintains the current sampling frequency of the voltage sampling sequence and the current sampling sequence, and keeps the current LoRa communication frequency unchanged, so that the data continues to enter the continuous feature sequence acquisition process under stable conditions, ensuring the time consistency and signal integrity in the feature generation process. When the state classification result is a data critical zone, adjust the sampling control unit parameters, increase the sampling frequency of the voltage sampling sequence and the current sampling sequence to a high-precision sampling level, and shield the response mechanism of the remote control signal, suspend the transmission and execution of remote control commands, and ensure that the integrity of the sampling signal is not affected by control intervention, so as to enhance the accuracy of anomaly identification. When the state classification result is a data distortion area, the signal processing unit is activated to filter the sampled data in the current time period, identify data points with feature distribution offset in the continuous feature sequence, mark such data as an abnormal source, and feed the marking result back to the continuous feature sequence acquisition process to drive the isolation management of abnormal data in subsequent sampling cycles, thus constructing a dynamic control mechanism for state closed-loop control.

Images