Multi-mode cooperative control system and method for cable branch box

Abstract

The invention discloses a cable branch box multi-mode cooperative control system and method, and relates to the technical field of power equipment intelligent control, and the method comprises the following steps: injecting a low-amplitude synchronous detection pulse in an operation environment of a cable branch box, collecting a feedback response signal of the detection pulse, inverting an arrival sequence of an interference signal according to the feedback response signal, and carrying out the inversion of the arrival sequence of the interference signal; and constructing a micro time slot interference portrait, and generating a time sequence baseline for an interference identification reference. According to the method, the time sequence base line is established through micro time slot detection, and accurate identification of microsecond interference signals is realized in combination with phase consistency analysis, multi-channel sampling and a cross-cycle evolution model. The false triggering probability is reduced by adopting a phase-locked suppression and delay cache mechanism, and the safe release of peak energy is realized through a phase turn-back amplitude limiting and energy guiding mode, so that the stability and the anti-interference capability of the cable branch box in a complex electromagnetic environment are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent control of power equipment, and particularly relates to a cable branch box multi-modal collaborative control system and method. BACKGROUND

[0002] The "cable branch box multi-modal collaborative control" refers to, in the operation and management process of the cable branch box, no longer relying on a single sensor or a single control logic, but introducing multiple perception dimensions (such as electrical parameters, environmental parameters, partial discharge signals, temperature and humidity, video images, vibration acoustic information, etc.) at the same time, fusing and processing these heterogeneous and multi-source data, and forming a multi-modal information collaborative judgment mechanism. On this basis, the control system can realize more accurate state recognition, fault warning and action decision according to the complementarity and correlation between different modal data, so as to dynamically optimize the operation condition of the branch box and improve the safety, stability and intelligent level of the power distribution network. In other words, the multi-modal collaborative control is a closed-loop mode of multi-dimensional perception + fusion analysis + linkage execution, which upgrades the cable branch box from "single-point monitoring" to an active control unit of "comprehensive perception + intelligent decision".

[0003] The prior art has the following disadvantages: In the prior art, the cable branch box usually relies on a fast switching mode to complete load switching, protection action switching and state self-checking and other dynamic operations in the operation process. However, in such a fast switching operation mode, the electromagnetic interference signals existing in the external environment are easily continuously superimposed in a micro-time slot and collapsed into a sharp pulse with an amplitude far exceeding the normal threshold in a very short time scale. Since the appearance and disappearance of the sharp peak signal occur in the microsecond level, and the controller in the prior art usually takes the millisecond level as the response and judgment time window, the system cannot timely identify such abnormal interference signals in the collection and judgment process. As a result, the controller is prone to misjudge the sharp peak signal as a real overcurrent, short circuit or partial discharge fault, thereby triggering high-risk actions such as tripping and isolation, which not only causes abnormal interruption of the operation of the cable branch box, but also may cause large-scale power supply instability and equipment damage.

[0004] The above information disclosed in the background section is only used to enhance the understanding of the background of the present disclosure, and therefore it can include information that does not constitute prior art known to those of ordinary skill in the art. SUMMARY

[0005] The purpose of the present application is to provide a cable branch box multi-modal collaborative control system and method, which establishes a timing baseline through micro-slot detection, combines phase consistency analysis, multi-channel sampling and cross-cycle evolution model to realize accurate identification of microsecond-level interference signals. The use of phase-locked suppression and delay buffer mechanism reduces the probability of false triggering, and the use of phase foldback limiting and energy guiding methods realizes the safe release of peak energy, significantly improves the stability and anti-interference ability of the cable branch box in complex electromagnetic environments, to solve the problems in the above background technology.

[0006] To achieve the above purpose, the present application provides the following technical scheme: a cable branch box multi-modal collaborative control method, comprising the following steps: Inject a low-amplitude synchronous detection pulse into the operating environment of the cable branch box, collect the feedback response signal of the detection pulse, inverse the arrival sequence of the interference signal according to the feedback response signal, construct a micro-slot interference image, and generate a timing baseline for interference identification reference; On the basis of the timing baseline, calculate the phase consistency index between multiple interference signals, lock the trigger time window of interference signal superposition collapse according to the phase consistency index, and extract a candidate peak signal set within the trigger time window; For the candidate peak signal set, simultaneously perform first channel high-fidelity data acquisition and second channel phase disturbance data acquisition, and by comparing the amplitude difference of the two channels at the same time point, eliminate the pseudo-peak signals with inconsistent amplitudes, and obtain the effective peak signal set for subsequent analysis; Based on the historical distribution characteristics of the effective peak signal set, a cross-cycle peak signal evolution model is established to predict the peak risk signal that will appear in the next microcycle, and according to the prediction result, a phase-locked signal suppression time window is set, and an energy release channel is configured to release the risk energy in advance; Within the phase-locked signal suppression time window, an adaptive delay micro-buffer mechanism is enabled to extend the peak signal in the time domain to reduce the instantaneous slope of the peak signal and interrupt the cumulative trigger path of false action; Based on the output signal of the adaptive delay micro-buffer mechanism, a phase foldback limiting process is started to guide the residual peak signal to the anti-signal convergence path, and the limiting process parameters are dynamically adjusted in combination with the real-time residual signal to realize the continuous dissipation and safe transfer of peak energy.

[0007] Preferably, the timing baseline generation step is as follows: Under the condition that the cable branch box is in a steady state, inject a synchronous detection pulse with an amplitude of 10 to 30 volts, a pulse width of less than 5 microseconds and a repetition period of 20 milliseconds into the cable line; By laying four voltage acquisition points in the incoming terminal, the adapter joint, the outgoing terminal end and the load lead-out terminal, the response time, amplitude and waveform trajectory of the detection pulse are collected respectively, and the response time of the incoming terminal is taken as zero point to establish a signal propagation sequence map; By normalizing the response time of multiple acquisition points, amplitude normalization and waveform interpolation smoothing, a standardized time sequence baseline is constructed with the incoming terminal as the starting node. The established time sequence baseline is used as the time discrimination reference for interference recognition. In the subsequent steps, it is used to compare the offset degree of the actual running signal in response time and propagation sequence to identify the trigger precursor of the sharp peak interference.

[0008] Preferably, the candidate sharp peak signal set extraction step is as follows: Collect the detection pulse response data in multiple cycles at the incoming terminal, outgoing joint, middle section copper bar and load lead-out terminal, and construct the phase offset sequence of each sampling point with the first response time of the incoming terminal as the reference. Perform periodic difference processing on the phase offset sequence to identify whether multiple sampling points show phase synchronization convergence trend in a given time period, and detect the aggregation degree of response time of each point. When the response time offset value of multiple sampling points in the same time interval is less than 300 nanoseconds, the amplitude change rate is more than 50 volts per microsecond, and the main waveform energy density is significantly higher than five times the background noise, mark this time interval as the trigger time window of interference signal superposition collapse. In the trigger time window, based on the voltage signal rising rate, amplitude mutation degree, duration and periodicity characteristics of each sampling point, filter out the sharp peak signal segments that meet the high mutation and high consistency conditions to form the candidate sharp peak signal set, and mark the corresponding time, amplitude and position parameters.

[0009] Preferably, the effective sharp peak signal set generation step is as follows: High-fidelity acquisition channels and phase disturbance acquisition channels are set at the incoming terminal, outgoing copper bar, middle section adapter copper bar and load lead terminal respectively, and the voltage waveforms corresponding to the candidate sharp peak signals are collected synchronously by the two channels. The candidate sharp peak signals collected in the two channels are subjected to time domain alignment and amplitude normalization processing, the amplitude difference of each sampling point is calculated, and the waveform consistency characteristics are analyzed. When the amplitude difference of more than 70% of the total sampling points in the two channels is less than 5% of the maximum amplitude, and the amplitude difference at the peak point is less than 3% of the maximum amplitude, and all the above conditions are met in three consecutive detection cycles, the candidate sharp peak signal is identified as an effective sharp peak signal and is retained. All candidate spike signals that meet the consistency requirements are combined into a valid spike signal set, and their occurrence time, sampling point number, peak voltage and stability level parameters are recorded.

[0010] Preferably, the phase-locked signal suppression time window is set according to the prediction results, and an energy release channel is configured to mitigate the risk energy in advance. The specific steps are as follows: Extract the time position, amplitude, rise rate and propagation path of effective spike signals within multiple consecutive working cycles, construct a time density map of spike signals and analyze the high-frequency clustering area and path repetition characteristics within the cycle; Based on the time position offset, amplitude change trend and path stability, a peak signal evolution feature group is constructed to predict the time interval of the peak risk signal in the next micro-cycle; Set the phase-locked signal suppression time window according to the predicted time interval, pause the high-sensitivity fault identification process, and start the instantaneous waveform buffering mechanism; Within the phase-locked signal suppression time window, a parallel resistor network, a varistor, and a surge absorption inductor are connected to construct an energy dissipation channel, guiding the energy of the spike signal to dissipate along a low-resistance path and reducing its impact on the main line.

[0011] Preferably, within the phase-locked signal suppression time window, an adaptive delay micro-buffer mechanism is enabled to extend the peak signal in the time domain. The steps are as follows: After the phase-locked signal suppression time window is activated, the spike signal collected within the window is buffered. The buffer unit has a capacity of 1,024 bytes, supports nanosecond-level read and write response capability, and continuously records voltage values, timestamps and rising edge change rates. The buffer period is consistent with the phase-locked window. The buffered signal is subjected to a delay expansion operation, which extends the sampling interval to three to five times for signal segments with voltage changes exceeding 80 volts within one microsecond, reduces the slope of the rising segment to one-third to one-fifth of the original, and reconstructs the waveform of the falling segment with an amplitude drop of more than 60 volts within one microsecond into a gradual descent platform to ensure a smooth signal shape. The extended signal is re-injected into the protection discrimination path, and logical judgment is made based on the action threshold and action duration to verify that the extended signal no longer triggers malfunctions, thereby cutting off the trigger chain formed by the continuous superposition of spike signals.

[0012] Preferably, the output signal based on the adaptive delay micro-buffering mechanism initiates the phase foldback limiting process as follows: Using the extended output signal as input, the residual spike signal in the waveform with an amplitude greater than the preset warning threshold is identified, and an inverse signal with the same amplitude, consistent frequency, time synchronization and opposite voltage direction is generated in the physical circuit. The signal is then guided to the energy absorption end for dissipation through the parallel branch path. While the inverted signal path is started, the limiting voltage suppression circuit of the main channel dynamically adjusts the action threshold and response time according to the amplitude, duration and slope characteristics of the residual signal to achieve voltage spike clipping and delay suppression. The temperature rise, current change rate, and impedance drift of the reverse signal guide path are monitored. When the temperature rise exceeds the set value or the current fluctuates in the opposite direction, the control unit switches to the backup attenuation path to ensure that the residual energy is completely absorbed and safely transferred.

[0013] A multimodal collaborative control system for cable branch boxes includes an interference profiling module, an interference focusing identification module, a spike verification and cleaning module, a spike prediction and suppression planning module, a delay-expansion peak reduction module, and a residual energy absorption and limiting control module. The interference profile construction module injects low-amplitude synchronous detection pulses into the operating environment of the cable branch box, collects the feedback response signal of the detection pulses, inverts the arrival sequence of the interference signal based on the feedback response signal, constructs a micro-timeslot interference profile, and generates a time-series baseline for interference identification. The interference focusing identification module calculates the phase consistency index between multiple interference signals based on the time-series baseline, locks the trigger time window of interference signal superposition and collapse according to the phase consistency index, and extracts a set of candidate spike signals within the trigger time window. The spike verification and cleaning module simultaneously acquires high-fidelity data from the first channel and phase perturbation data from the second channel for the candidate spike signal set. By comparing the amplitude differences of the two channels at the same time point, it removes false spike signals with inconsistent amplitudes and obtains a set of effective spike signals for subsequent analysis. The peak prediction and suppression planning module establishes a cross-cycle peak signal evolution model based on the historical distribution characteristics of the effective peak signal set, predicts the peak risk signal that will appear in the next micro-cycle, sets the phase-locked signal suppression time window according to the prediction results, and configures the energy release channel to release the risk energy in advance. The delayed expansion peak clipping module enables an adaptive delay micro-buffer mechanism within the phase-locked signal suppression time window to extend the peak signal in the time domain, thereby reducing the instantaneous slope of the peak signal and interrupting the cumulative triggering path of malfunctions. The residual energy absorption and limiting control module, based on the output signal of the adaptive delay micro-buffer mechanism, initiates the phase foldback limiting process, guides the residual peak signal to the inverted signal confluence path, and dynamically adjusts the limiting processing parameters in combination with the real-time residual signal to achieve continuous dissipation and safe transfer of peak energy.

[0014] The technical effects and advantages provided by the present invention in the above technical solution are as follows: This invention starts with micro-timeslot detection, establishes a timing baseline, and integrates phase consistency analysis, multi-channel comparative sampling, and cross-cycle evolution models. This not only improves the identification accuracy of microsecond-level interference signals but also effectively reduces the probability of false triggering of interference signals through mechanisms such as phase-locked loop suppression and delay buffering. Simultaneously, by employing phase foldback limiting and reverse energy guidance, it achieves physical mitigation and safe transfer of peak energy, significantly improving the stable operation of cable branch boxes in complex electromagnetic environments, avoiding high-risk actions such as false tripping and false isolation, and enhancing the overall anti-interference capability and intelligence level of the power distribution system. Attached Figure Description

[0015] 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.

[0016] Figure 1 This is a flowchart of a multi-modal collaborative control method for a cable branch box according to the present invention.

[0017] Figure 2 This is a schematic diagram of a multimodal collaborative control system for a cable branch box according to the present invention. Detailed Implementation

[0018] 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.

[0019] This invention provides, for example Figure 1 The multimodal cooperative control method for a cable branch box, as shown, includes the following steps: Low-amplitude synchronous detection pulses are injected into the operating environment of the cable branch box, and the feedback response signal of the detection pulses is collected. The arrival sequence of the interference signal is inverted based on the feedback response signal to construct a micro-timeslot interference profile and generate a time-series baseline for interference identification. In the initial phase of operation, an interference propagation image is constructed and a temporal reference baseline is generated through active detection to provide a basis for subsequent multimodal judgment. The specific process of this step is as follows: Under steady-state operation of the cable branch box, inject low-amplitude synchronous detection pulses into the cable line at an appropriate time. To ensure the effectiveness and safety of the injected signal, an initial assessment of the current operating status of the branch box must be performed before pulse injection. Assessment indicators include: the internal bus voltage of the branch box should be maintained within ±2% of the rated value; the load current variation rate of the branch cable should not exceed 5% / second; the internal temperature of the cable should be within the allowable operating range of the equipment, for example, between 15℃ and 55℃; the grounding resistance should be less than 4 ohms to ensure the stability of the electromagnetic coupling path; and the spatial electromagnetic background noise intensity within the cable channel should not exceed 50 microvolts / meter to avoid the pulse signal being overwhelmed by environmental interference. After completing the above assessment, start the detection pulse generator. The pulse is a unipolar, narrow-width, constant-amplitude waveform with an amplitude range of 10-30 volts, a duration of less than 5 microseconds, and a pulse repetition period set to 20 milliseconds. The injection method is as follows: inject the detection pulse signal from the incoming terminal, allowing it to propagate along the cable transmission path to the outgoing terminal and the end contact, ensuring that the signal can cover the main conductive structure of the entire branch cable.

[0020] During the pulse injection detection process, feedback signals are received and sampled through voltage acquisition nodes located at key positions inside the cable branch box. The specific node arrangement is as follows: the first acquisition point is located inside the incoming terminal to record the initial injection time and signal emission reference time; the second acquisition point is located near the cable fixing support structure adjacent to the adapter joint to record the signal response characteristics at the winding path turning point; the third acquisition point is located at the end of the transition shielding layer before the outgoing terminal to observe the signal attenuation amplitude and propagation delay in the later stages of transmission; the fourth acquisition point is located between the load lead-out terminal and the connecting bolt to record the amplitude change and waveform distortion of the signal at the end. Each acquisition point uses a high-speed sampling device with a sampling rate of 10 MHz or higher, and the data recording format is timestamp + amplitude + waveform trajectory. The signal transmission line uses a shielded twisted-pair structure, and the time sampling accuracy of each node is controlled within 100 nanoseconds. To ensure the accuracy of the detection results, the system is set to only acquire data once within the same pulse cycle to avoid data overlap errors caused by multiple samplings. The acquired feedback signals will be saved in time series form for subsequent interference path inversion.

[0021] Based on the probe pulse response data acquired from multiple sampling points, a complete signal propagation sequence map is established by comparing the relative order of response times and amplitude variation characteristics of each node. Using the incoming terminal as the zero point, the response delay time of each other sampling point relative to this zero point is calculated one by one. For example, if the pulse response time at the incoming terminal is T0, at the adapter joint is T1, at the outgoing terminal is T2, and at the end load is T3, then the constructed propagation sequence is T0→T1→T2→T3. This propagation sequence not only reflects the physical transmission characteristics of electrical signals in the conductor structure but also implicitly includes the combined effects of factors such as the dielectric structure of the cable path, inductance and capacitance distribution, contact impedance, and connection integrity. To enhance the map's analytical capabilities, further quantification of the signal amplitude variation rate is required. For instance, if the amplitude decreases by more than 30% at the adapter joint, it indicates a possible impedance mismatch or local coupling interference, requiring a correction factor to be introduced in subsequent judgments. For secondary reflections or ringing waveforms appearing in the response signal, the principle of "the first main response wavefront being valid" is adopted to filter out secondary responses in the high-frequency wave tail, ensuring that only propagation trajectories with main path significance are retained in the propagation spectrum. To achieve the mapping between spatial structure and time series, cable layout drawings are referenced during the spectrum construction process, and the signal propagation paths are marked on the cable structure diagram accordingly to form a complete spatial-temporal two-dimensional interference image.

[0022] After constructing the signal propagation map, a standardized time-series baseline is established based on the first response time of each response point. The time-series baseline is used as the time discrimination criterion in the subsequent peak interference identification process, and its generation process includes three sub-steps.

[0023] The first step is to perform uniform normalization on all response timestamps to eliminate random offset errors between pulses; The second step is to standardize the amplitude of the waveform at each response point by using a fixed normalization ratio to map the amplitude to the range of [0,1], which facilitates joint modeling of time axis and amplitude changes. The third step is to perform smooth interpolation on the normalized data to eliminate instantaneous high-frequency jitter caused by electromagnetic fluctuations at the scene.

[0024] The resulting timing baseline is a set of time data that starts from the incoming terminal and increments node by node. It has a clear structure, stable waveform, and uniform amplitude distribution. In the subsequent interference identification process, if the instantaneous signal acquired during actual operation deviates significantly from this timing baseline in terms of response time or propagation sequence, it can be used as a precursor to spike interference, providing a precise time anchor point for predicting and suppressing abnormal signals.

[0025] This step lays the foundation for a time reference and response characteristics for the intelligent identification and precise control of cable branch boxes in complex electromagnetic interference environments. By actively injecting low-amplitude, short-duration synchronous detection pulses under stable operating conditions of the cable branch box, and simultaneously acquiring response signals during pulse propagation at multiple key nodes, the time delay, amplitude changes, and waveform morphology of the signal's actual propagation in the cable path can be comprehensively understood. Then, through inversion processing of these feedback signals, the actual arrival sequence of the interference signal is established, revealing the true trajectory of interference propagation within the structure. The resulting micro-timeslot interference profile and standardized timing baseline serve as reference standards for determining whether spike-like abnormal interference signals exist during subsequent operation. This not only improves the timing accuracy of subsequent identification processes but also provides precise time anchors and spatial correlation basis for tracing interference sources, analyzing evolution trends, and regulating interference response mechanisms. This active sensing step significantly enhances the system's ability to identify and judge microsecond-level interference behavior, providing fundamental support for constructing a multimodal collaborative control mechanism.

[0026] Based on the timing baseline, the phase consistency index among multiple interference signals is calculated. The trigger time window for the superposition and collapse of interference signals is locked according to the phase consistency index, and a set of candidate spike signals is extracted within the trigger time window. After constructing the time-series baseline for interference identification, a phase consistency analysis-based method is proposed to identify and extract spike interference signals caused by electromagnetic disturbances during operation. This method identifies the synchronous superposition behavior of multiple interference signals in the time series, extracting a set of signals with typical spike characteristics to provide input for subsequent discrimination and response. The process includes the following steps: Based on the timing baseline established in the previous stage, low-amplitude synchronous probe pulses are repeatedly injected at fixed intervals during the operation of the cable branch box. Response data from multiple sampling points inside the branch box are collected after each injection. Sampling points are located inside the incoming terminal, at the center connection point of the outgoing connector, on the copper busbar connecting the intermediate section, and on the load-side terminal. Each sampling point is equipped with a high-speed voltage acquisition unit with a sampling rate of no less than ten million times per second and has an internal time reference for unified time marking. After completing more than fifty consecutive probe pulse injection operations, the main response wavefront time of each response process is collected and organized, i.e., the time of the first voltage change at each sampling point in each pulse, forming a multi-cycle, same-node response time series.

[0027] The aforementioned multi-cycle response time series is time-synchronized to establish the trend trajectory of response time changes at the same sampling point across multiple cycles, and then compared horizontally. In this process, the response time of the incoming terminal is used as a reference zero point, and the response time offset values ​​of the remaining sampling points are uniformly adjusted to relative time differences to construct the phase change trajectory of each node. For example, if the response time of the outgoing connector is 2.3 microseconds in the first cycle, 2.1 microseconds in the second cycle, and 2.4 microseconds in the third cycle, then the phase offset sequence for that point is +0.2 microseconds, 0 microseconds, and +0.3 microseconds. After similar calculations are performed at all sampling points, the phase offset values ​​of each point are integrated along the time dimension to form a cluster of periodic phase change curves. To enhance the accuracy of trend representation, each curve undergoes first-order differencing to calculate the change in response time between adjacent cycles, identifying the trend of phase convergence or divergence within a time period.

[0028] Based on the cluster of phase change curves, the degree of aggregation of phase shift values ​​within a specific time window is detected to pinpoint the trigger time window where potential spike interference may occur. To achieve this identification, the time aggregation peaks of the phase shift curve at each sampling point are first extracted, identifying the time period where the phase shift approaches zero. When multiple sampling points have phase shift values ​​approaching zero within the same time interval, and the response time difference does not exceed 300 nanoseconds, the amplitude change rate is greater than 50 volts per microsecond, the rise time of the response waveform is less than 1 microsecond, and the energy density of the wavefront is significantly higher than 5 times the background noise, this time period can be determined as a region of highly concentrated phase consistency. By observing the recurrence in fifty injection cycles, if the frequency of phase convergence within the same time period exceeds 80%, it is identified as a trigger time window where interference signals may superimpose and collapse. The start and end time of this window is defined from the first occurrence of phase convergence at all sampling points to the end of the response at the last set of sampling points, typically controlled within 3 to 6 microseconds. This time window establishes the boundary range for subsequent spike feature extraction.

[0029] Within the locked interference signal trigger time window, the response signal waveform at each sampling point is meticulously analyzed to extract waveform segments with high amplitude abrupt changes, extremely steep rise times, and extremely short durations, forming a candidate spike signal set. Specifically, within the trigger time window, each sampling point is searched for a signal segment with a voltage signal rise slope greater than 100 volts per microsecond and a continuous holding time of less than 2 microseconds. Then, it is determined whether this waveform segment represents the maximum amplitude signal in the corresponding period, and whether its peak value exceeds 1.5 times the average of historical response signals at the same point. Finally, it is further verified whether the signal repeats in a similar manner across multiple periods, with its peak time shift not exceeding 200 nanoseconds. After meeting these three conditions, the waveform segment is defined as a candidate spike signal, and its occurrence time, peak amplitude, rise time, duration, and corresponding sampling point position are recorded. Finally, all waveform segments meeting the conditions are combined into a candidate spike signal set in structured data format, sorted, and saved for subsequent steps of identification, verification, and prediction.

[0030] The purpose of this step is to identify the time periods during which the interference signals exhibit synchronous superposition during propagation by analyzing the time response performance of the interference signals at various sampling points across multiple cycles, based on the established time-series baseline. This allows for the precise extraction of signal sets with spike characteristics. By comparing the relative response times of different sampling points in multiple probe pulse responses, the phase change trend is calculated, identifying a concentrated feature where multiple sampling points exhibit phase convergence and minimal time offset within a specific time period. When this phase consistency is highly concentrated, it often indicates that multiple weak interference signals have accumulated and superimposed within that time period, collapsing to form a spike signal with a sudden increase in amplitude and an extremely short duration. By locking these trigger time windows and performing high-precision waveform analysis within them, signal samples suspected to be genuine spikes can be extracted, eliminating the interference from ordinary noise and asynchronous interference. This provides a data foundation and time anchor for subsequent identification of false spikes, high-fidelity verification, and dynamic prediction, significantly improving the accuracy and reliability of the identification.

[0031] For the candidate spike signal set, high-fidelity data acquisition of the first channel and phase perturbation data acquisition of the second channel are carried out simultaneously. By comparing the amplitude difference of the two channels at the same time point, false spike signals with inconsistent amplitudes are eliminated, and the effective spike signal set for subsequent analysis is obtained. Before further identifying the candidate spike signal set extracted in the previous step, a cross-validation method based on dual-channel data acquisition is proposed to eliminate false spike signals caused by non-real interference sources, environmental electromagnetic disturbances, equipment contact oscillations, or partial discharges, ensuring the authenticity and stability of the input data. This method accurately compares the consistency of the response amplitude of the same spike signal in two physical acquisition channels to identify valid spike signals that truly exist in the cable's physical structure and have stable propagation paths. This allows for the elimination of falsely identified signals from the candidate spike signal set, constructing a reliable set of valid spike signals. The process includes the following steps: Two independent data acquisition paths with identical structure and similar performance but different acquisition mechanisms are configured simultaneously at key sampling locations in the cable branch box. The first acquisition path is a high-fidelity channel, which uses a high-precision analog voltage acquisition unit with a sampling frequency of 50 million times per second, a bandwidth greater than 20 MHz, and an input impedance of not less than 10 megohms. This unit is installed inside the incoming terminal, at the outgoing copper busbar connection, at the intermediate busbar transition, and at the load output lead terminal. These acquisition points are connected to the acquisition equipment via shielded twisted-pair cables to ensure complete signal transmission even in high-frequency interference environments. The second acquisition path is a phase perturbation channel. Based on the same hardware structure as the first channel, this channel introduces a small amount of controlled jitter into the sampling clock, randomly shifting its sampling time point within each cycle. The shift amplitude is within ±300 nanoseconds, simulating the effect of slight phase drift on the signal acquisition results. This sampling perturbation does not affect the true characteristics of the signal itself, but is sufficient to reveal the signal's sensitivity to sampling phase shifts. Both acquisition paths are connected to a unified trigger control unit to ensure that data is started synchronously under the same injection pulse event.

[0032] The peak signal waveforms acquired from the two acquisition paths are standardized and time-domain aligned. To ensure alignment accuracy, the starting point of each candidate peak signal in the high-fidelity channel is selected as a reference. The waveform segment corresponding to the time period in the phase perturbation channel is extracted and reconstructed using equal-interval interpolation to ensure consistency with the high-fidelity channel in time resolution. Within the time interval of each candidate peak signal, the voltage amplitude of each sampling point in both channels is extracted to form a one-to-one data pair. Special attention is paid to characteristic regions such as the slope change of the voltage rise edge, the peak voltage point, and the descent rate of the post-peak attenuation segment. In the real peak signal, since the signal is caused by the release of electromagnetic energy coupled within the cable structure, it has a definite propagation path and physical consistency. Therefore, even if there is a sampling phase shift in the two channels, its amplitude, slope, and overall waveform shape will remain highly consistent. In pseudo-peak signals, since the signal is often caused by external noise or local contact oscillation, its waveform is greatly affected by phase disturbance. Phenomena such as amplitude abrupt change and asynchronous, peak time deviation and signal contour reconstruction failure will occur in the two channels.

[0033] Based on the amplitude difference between the two channels, a judgment criterion is constructed, and all candidate spike signals are cross-checked one by one. During the judgment process, the voltage values ​​of corresponding sampling points in the two channels are compared, and the amplitude difference ratio across the entire band is calculated, with a focus on analyzing the amplitude consistency at the rising edge and peak point. When the candidate signal exhibits high waveform overlap in the two channels (i.e., the voltage difference is less than 5% of the maximum amplitude at more than 70% of the total sampling points, and the voltage difference at the peak point is less than 3% of the maximum amplitude), the signal is determined to be a genuine and valid signal and is retained. Conversely, if any of the above conditions are not met, or if a severe mismatch occurs in the waveform (e.g., a peak exists in one channel while the other only shows a gradually changing curve), the signal is determined to be a false spike signal and is discarded. To further improve the reliability of the verification, each candidate spike signal must undergo comparative verification within at least three consecutive probe pulse cycles. Only when the comparison results in all cycles meet the consistency requirements is the signal officially included in the valid signal set.

[0034] All spike signals that passed the dual-channel amplitude consistency test were collected to construct a valid spike signal set. Each spike signal in the set was accompanied by key parameters such as its occurrence time, sampling point number, waveform rise time, peak voltage, dual-channel overlap ratio, maximum amplitude difference, and stability level within the sampling period, forming a structured data table for subsequent spike behavior prediction, interference trend modeling, and response control strategy formulation. This method eliminates false positives from the candidate spike signal set, improving the authenticity and robustness of the input data used in the entire control process.

[0035] The purpose of this step is to identify and eliminate spurious spike signals of unknown origin or unstable physical characteristics from the candidate spike signal set through dual-channel acquisition and verification, ensuring the authenticity and stability of the data upon which subsequent analysis relies. During the operation of cable branch boxes, due to the complex electromagnetic environment, transient waveforms often appear caused by external interference, loose contacts, electrical noise, and other factors. Although these signals may exhibit high-amplitude abrupt changes in a single channel, they do not possess the electromagnetic propagation consistency expected of genuine spike signals. This step sets up two acquisition paths: one acquires the original high-fidelity waveform, and the other introduces a small disturbance in the sampling clock, comparing the amplitude and waveform profile of signals within the same time window. If a signal exhibits highly consistent amplitude response and morphological characteristics in both channels, it can be identified as a genuine spike; conversely, if there is a significant amplitude difference or waveform mismatch, it is considered a spurious spike and eliminated. This step effectively improves the accuracy of signal identification, avoiding erroneous judgments or control commands triggered by false signals in subsequent stages, and providing a reliable data foundation for evolution prediction, energy regulation, and other applications.

[0036] Based on the historical distribution characteristics of the effective spike signal set, a cross-cycle spike signal evolution model is established to predict the spike risk signal that will appear in the next micro-cycle. According to the prediction results, a phase-locked signal suppression time window is set, and an energy release channel is configured to release the risk energy in advance. After screening and verifying the effective spike signals, the resulting signal set possesses clear temporal characteristics, electrical amplitude characteristics, and structural propagation path consistency. To achieve trend prediction of spike interference behavior in future operating cycles and avoid the risk of malfunctions caused by concentrated spike energy bursts, cross-cycle evolution characteristics are established based on the distribution patterns of effective spike signals within historical cycles. A phase-locked loop (PLL) signal suppression window is then set accordingly. Simultaneously, channels for energy guidance and mitigation are configured in the physical circuit to achieve early detection and proactive response to high-risk spike interference. This process includes the following steps: Effective spike signal distribution data are extracted from multiple consecutive working cycles to construct a time series feature set. The spike signal record for each cycle includes the absolute time point of spike occurrence (referenced to the system start time), duration, peak amplitude, waveform rise rate, fall time, occurrence location, and propagation path identifier. Over thirty or more consecutive cycles, the above data is summarized, and all signal events are arranged along a time axis to form a spike signal temporal density map. After normalization of each signal, the time position is mapped to a relative timescale relative to the cycle start point to analyze the occurrence trend of spike events in each cycle. Subsequently, high-frequency clustering areas of spike events within a cycle are statistically analyzed. For example, if signal segments with consecutive spike time differences less than one microsecond frequently concentrate in a specific time period of the cycle (e.g., the 9-12 microsecond interval), this period can be preliminarily identified as a high-risk area for spike evolution. Simultaneously, by observing the consistency of propagation paths of adjacent spikes across multiple cycles—for example, whether they all originate at the incoming line end or are concentrated at the busbar node—the structural repeatability of the spike formation mechanism can be further determined, providing a foundation for subsequent modeling.

[0037] Based on the aforementioned peak timing characteristics, evolutionary trend elements are extracted, and an evolutionary behavior model of multi-period peak events is constructed. To establish this model, the changes in the time position, amplitude fluctuation, and rise rate gradient of the peak signal in different periods need to be quantified. For example, if, within periods 15 to 30, a peak signal at a certain sampling point consistently appears earlier, with a time offset decreasing by 0.3 microseconds per period and a continuously increasing amplitude, averaging an increase of 6 volts per period, this phenomenon indicates that the peak signal is accelerating its evolution and exhibits a strong energy superposition trend. By comparing these peaks laterally, it is possible to identify whether multiple sampling points simultaneously exhibit similar changing characteristics. If so, it indicates that the cable branch box has entered a peak coupling evolution period, potentially forming a significant peak impact in future periods. During the modeling process, signal timescale, amplitude trend, slope change, and path repeatability are jointly constructed as evolutionary trend identification criteria, forming a feature set for predicting the next period.

[0038] Based on the aforementioned evolutionary trend characteristics, the time interval most likely to experience a spike impact in the next operating microcycle is predicted, and a phase-locked signal suppression time window is set. This time window is set based on the high-frequency clustering period of the aforementioned spike events in historical cycles and the time offset of the evolutionary trend's forward progression. For example, if a high-risk spike occurs around the 10th microsecond in the last three cycles, and each cycle advances by 0.2 microseconds, the spike in the next cycle can be predicted to occur between 9.4 and 9.8 microseconds. Therefore, the phase-locked window is set to 9.2 to 10.0 microseconds. To avoid false interception due to an excessively wide window, the window width should not exceed 2 microseconds, and a safety buffer zone should be reserved at the edges to prevent signal boundary penetration. During the activation of the phase-locked window, high-sensitivity fault judgment processing is suspended to prevent false tripping caused by spike signals. Simultaneously, the protection mode is activated within this window, entering a state of instantaneous waveform buffering and post-processing. By precisely setting the phase-locked window, proactive time avoidance can be achieved, improving anti-interference capabilities.

[0039] Simultaneously with the activation of the phase-locked loop (PLL) suppression window, an energy discharge channel is configured to guide and dissipate potential high-amplitude energy spikes, preventing them from concentrating on critical nodes or relay protection devices. The energy discharge channel is constructed as follows: a parallel bypass resistor network is added inside the cable branch box, connected between the neutral point and ground, with a resistance of 5 to 10 ohms, forming a low-resistance absorption path when a spike signal arrives; simultaneously, a varistor is installed at the busbar connection end, with its initial operating voltage set to a threshold 10% higher than the normal operating maximum. Once the spike amplitude exceeds this value, it immediately activates, guiding the spike energy to be released to the ground end; furthermore, a surge absorption inductor is installed at the cable terminal to suppress the spike rise rate and delay voltage ramp-up. This channel is in a conducting state when the PLL window is open, effectively diffusing spike energy into a low-energy path, reducing its impact on the main line. After the PLL window closes, the channel exits the conducting state and returns to normal operation.

[0040] The purpose of this step is to extract the distribution characteristics of the effective spike signal set obtained in the previous stage through cross-cycle historical analysis, thereby identifying the evolution trend of the spike signal during the operation of the cable branch box, and then predicting the time period during which a spike impact may occur in the next micro-cycle. This prediction not only improves the ability to anticipate electromagnetic interference events, but also provides a precise time anchor for proactive prevention and control. Based on this, a phase-locked signal suppression time window is set, allowing the protection strategy to enter a buffer state during high-risk periods, avoiding misjudging spikes as actual faults and causing tripping. At the same time, by configuring an energy dissipation channel composed of resistors, varistor devices, and surge absorption elements, a controllable energy transfer path is provided before the predicted spike occurs, effectively diverting the spike energy to non-critical branches and dissipating it in advance. This step realizes the transformation from "passive identification" to "proactive prediction and mitigation," significantly improving the system's ability to cope with high-frequency spike interference and its operational stability.

[0041] Within the phase-locked signal suppression time window, an adaptive delay micro-buffer mechanism is enabled to extend the spike signal in the time domain, thereby reducing the instantaneous slope of the spike signal and interrupting the cumulative triggering path of malfunctions. To prevent spike signals from causing malfunctions in the control logic, especially during periods of concentrated energy spikes that trigger protection functions and lead to equipment tripping, shutdown, or isolation faults, an adaptive delay micro-buffer mechanism is proposed to expand and process characteristic spike signals in the time domain within a pre-defined phase-locked loop (PLL) signal suppression time window. By smoothing and delaying the instantaneous steep changes in the signal, the probability of triggering protection conditions is reduced, thereby breaking the chain of malfunctions formed by the continuous superposition of spike signals. The implementation of this mechanism involves the following steps: After the phase-locked signal suppression time window is triggered and activated, the acquisition and buffering operation for the spike signal within that time period is immediately initiated. When the control circuit receives an external high-frequency signal input, it automatically invokes the buffering logic to temporarily halt the transmission of all voltage signal sampling points within the phase-locked window to the main protection judgment process, instead storing them sequentially in a ring buffer unit. In actual implementation, the buffer unit is a high-speed storage circuit with a capacity of 1024 bytes, supporting nanosecond-level read and write response capabilities. This buffer continuously records basic data such as voltage values, timestamps, and rising edge change rates in the target waveform signal. The buffering operation duration is consistent with the phase-locked window; for example, if the phase-locked window is set to 6 microseconds, the buffer acquisition period is also set to 6 microseconds. The acquired spike signal is saved with a complete waveform structure, awaiting subsequent extended processing.

[0042] A time-delay expansion operation is performed on the cached spike signal data. This operation is based on the recorded signal's rise time rate and amplitude variation characteristics for ordered processing. For signal segments with voltage jumps greater than 80 volts within 1 microsecond, an interpolation delay mechanism is implemented, artificially extending the original time interval between consecutive sampling points to 3 to 5 times during the store-and-forward phase. This is achieved by adjusting the output control clock precision, resulting in a significant stretching characteristic in the signal output process. For example, if the original time interval between two sampling points is 10 nanoseconds, it is adjusted to between 30 and 50 nanoseconds, reducing the waveform slope to 1 / 3 to 1 / 5 of the original. Without changing the overall voltage energy of the waveform, its distribution on the time axis is altered, making the signal present a gentle rise curve rather than a steep transition curve. For the signal falling segment, if the drop exceeds 60 volts within 1 microsecond, time expansion processing is also applied, reconstructing the rapid falling segment into a gently sloping plateau segment, thereby avoiding the formation of spike-shaped troughs and preventing control mis-triggering caused by bidirectional abrupt changes. The delay unfolding operation is performed in real time, and the processing cycle does not exceed the sampling delay cycle, ensuring that it will not cause signal backlog or timing drift issues.

[0043] The extended signal is re-injected into the protection discrimination path to participate in the judgment logic. By comparing the trigger response before and after the extension, the disconnection verification of the malfunction trigger chain is completed. After the extended signal returns to the main control path, the protection logic unit will re-judge the signal validity according to the set action threshold and action duration conditions. Since the instantaneous slope of the signal has been effectively reduced, even if the peak amplitude of the signal remains unchanged, its extended rise time will not meet the high steep continuous impact conditions required by traditional protection tripping actions, thus avoiding misidentification of the signal as overvoltage, short circuit, or grounding abnormality and triggering protection commands. At the same time, the time expansion process of the signal effectively blocks the chain response effect of multiple sampling points continuously identifying similar features in a short period of time, breaking the misjudgment triggering path formed by repeated peak superposition at the root. This processing mechanism can also perform memory-type feedback in multiple cycles. That is, when a certain type of peak signal is expanded and processed without triggering any substantial abnormality, the signal feature can be marked as low-risk, and the extended buffer channel can be automatically activated when it reappears in the future.

[0044] The purpose of this step is to actively regulate the temporal behavior of spike signals during high-risk trigger periods by introducing an adaptive delay micro-buffer mechanism, preventing them from triggering malfunctions due to their high-speed changing characteristics. Within the phase-locked loop (PLL) signal suppression time window, the acquired spike signals are temporarily buffered in a high-speed cache unit. The output rhythm is controlled to extend the signal in the time domain, smoothing out the originally steep rising edge and significantly reducing its instantaneous slope. This extension process does not change the overall energy and amplitude of the signal, but it slows down the speed at which the controller's trigger conditions are met, preventing the malfunction trigger threshold from being met continuously within a very short time. In this way, the synchronous response chain of spike signals across multiple sampling points is effectively broken, avoiding multi-point consistency misjudgments. Furthermore, this process also shields the response accumulation effect caused by the short-term recurrence of spike signals, buying time for subsequent energy suppression and feedback control, and improving the cable branch box's malfunction protection capability in complex disturbance environments.

[0045] Based on the output signal of the adaptive delay micro-buffer mechanism, the phase return limiting process is initiated to guide the residual peak signal to the inverted signal confluence path. The limiting processing parameters are dynamically adjusted in combination with the real-time residual signal to achieve continuous dissipation and safe transfer of peak energy. To further reduce the threat posed by high-amplitude residual spike signals that may still exist after the delay micro-buffering mechanism to the operational safety of cable branch boxes, a phase-reversal limiting process is introduced after the delay processing. This process constructs a phase-reversed signal path, guiding the remaining energy in the spike signal in reverse to a preset energy collection path. Combined with the residual signal collected in the main channel, the limiting parameters are adjusted in real time to achieve dynamic signal suppression and multi-path energy unloading. This ensures that spike disturbances no longer accumulate or concentrate on sensitive load structures, thus completing a closed-loop safety control from logic mis-triggered suppression to physical energy dissipation. The specific implementation steps are as follows: The output waveform processed by the adaptive delay micro-buffering mechanism is used as the input signal to identify any remaining amplitude anomalies. In specific implementation, a real-time voltage amplitude comparison stage is used to extract waveform segments with peak voltages exceeding a preset steady-state safety value. For example, if the system's normal maximum voltage is 220 volts, a safety warning threshold of 260 volts is set, and all signal segments exceeding this value are marked as residual spike signals. Based on this marking, the rising edge slope, duration, and start time are further extracted as parameter references for subsequent signal foldback generation. Subsequently, following the principles of identical amplitude, consistent frequency, and time synchronization with the original spike waveform, a set of inverted signals is generated in the physical circuit. The inverted signals are mirror images of the original signals in waveform shape, with a negative voltage direction. These inverted signals are no longer mixed with the main channel but are guided to a dedicated energy absorption terminal through a parallel branch path. This terminal includes structural units such as grounding leads, graphite-based energy-absorbing elements, and ceramic varistors in series attenuators. This path has low impedance and good heat dissipation, and can withstand the continuous impact of inverted signals, absorbing the peak electrical energy it carries in segments, thus realizing the transfer and diffusion of peak energy in physical structure.

[0046] Simultaneously with the activation of the inverted signal path, the voltage limiting suppression circuit on the main signal channel begins operation, dynamically adjusting the limiting strategy based on the residual signal within the current channel. The residual signal refers to the difference in amplitude, duration, and slope between the delayed, expanded signal and the original spike signal. This data is output in real-time by the main sampling unit and serves as the input for the limiting control parameters. If the amplitude of the residual signal still exceeds the system's safe operating range, for example, above 260 volts, the varistor in the limiting trigger device activates, clipping the voltage spike to below 230 volts to prevent the signal from crossing the activation threshold again. Simultaneously, if an abnormal increase in duration is detected in the residual signal, such as the waveform width increasing from the normal 2 microseconds to over 8 microseconds, the operating time of the limiting device will also be extended to match the energy release rhythm, preventing the signal from re-entering the judgment process before the limiting process ends and causing malfunctions. This limiting control employs a dynamic parameter mapping mechanism with a response time set within 100 nanoseconds, ensuring that voltage surges are suppressed immediately upon occurrence, rather than intervening after the waveform has taken shape, thus improving proactive response.

[0047] The entire operation of the reverse signal guiding path is monitored to ensure the continuity, safety, and controllability of the guiding dissipation process. In practice, three methods are used: parallel thermistors, current rate monitoring coils, and impedance drift probes, to monitor temperature rise, absorption current slope, and impedance shift in the reverse signal path. If the thermistor detects a local temperature rise exceeding 85 degrees Celsius, indicating that the energy-absorbing material is approaching saturation, the control unit will gradually reduce the amplitude of the reverse signal guiding voltage and increase the intervention ratio of the limiting action in the main channel, allowing the main channel to undertake more energy attenuation tasks. If the current rate continuously fluctuates in the opposite direction, indicating that the reverse signal has not been completely absorbed and secondary reflection has occurred, the guiding path structure is changed, for example, by switching to a redundant bypass channel, allowing the remaining energy to enter the backup attenuation branch for re-absorption. All these feedback adjustments operate independently at the physical end of the peak energy processing chain without interfering with the main control process, ensuring that the peak signal energy is completely transferred even under extreme interference environments, avoiding secondary failures caused by accumulation effects.

[0048] The purpose of this step is to further precisely control and orderly release the residual energy of the spike signal after it has been processed by the adaptive delay micro-buffering mechanism. This prevents the continued accumulation of high-amplitude signals or their impact on sensitive electrical components, which could lead to malfunctions or equipment damage. Although the delay processing has mitigated the abrupt changes in the spike signal in the time domain, residual energy exceeding the safety threshold may still exist. Therefore, this step constructs a set of mirror signals with opposite phase to the main signal, guiding this residual spike energy in reverse to an independent energy collection path, achieving physical energy transfer and absorption. Simultaneously, the residual signal change trend of the spike signal in the main channel is monitored in real time. Based on its amplitude, duration, and slope, the limiting parameters are dynamically adjusted to make the limiting action more precise and timely, ensuring that fixed parameter settings do not lead to misjudgments or insufficient suppression. Through this combined limiting and guiding mechanism, continuous dissipation and safe transfer of spike energy can be achieved.

[0049] This invention starts with micro-timeslot detection, establishes a timing baseline, and integrates phase consistency analysis, multi-channel comparative sampling, and cross-cycle evolution models. This not only improves the identification accuracy of microsecond-level interference signals but also effectively reduces the probability of false triggering of interference signals through mechanisms such as phase-locked loop suppression and delay buffering. Simultaneously, by employing phase foldback limiting and reverse energy guidance, it achieves physical mitigation and safe transfer of peak energy, significantly improving the stable operation of cable branch boxes in complex electromagnetic environments, avoiding high-risk actions such as false tripping and false isolation, and enhancing the overall anti-interference capability and intelligence level of the power distribution system.

[0050] This invention provides, for example Figure 2The multimodal collaborative control system for a cable branch box shown includes an interference profiling module, an interference focusing identification module, a spike verification and cleaning module, a spike prediction and suppression planning module, a delay expansion peak reduction module, and a residual energy absorption and limiting control module. The interference profile construction module injects low-amplitude synchronous detection pulses into the operating environment of the cable branch box, collects the feedback response signal of the detection pulses, inverts the arrival sequence of the interference signal based on the feedback response signal, constructs a micro-timeslot interference profile, and generates a time-series baseline for interference identification. The interference focusing identification module calculates the phase consistency index between multiple interference signals based on the time-series baseline, locks the trigger time window of interference signal superposition and collapse according to the phase consistency index, and extracts a set of candidate spike signals within the trigger time window. The spike verification and cleaning module simultaneously acquires high-fidelity data from the first channel and phase perturbation data from the second channel for the candidate spike signal set. By comparing the amplitude differences of the two channels at the same time point, it removes false spike signals with inconsistent amplitudes and obtains a set of effective spike signals for subsequent analysis. The peak prediction and suppression planning module establishes a cross-cycle peak signal evolution model based on the historical distribution characteristics of the effective peak signal set, predicts the peak risk signal that will appear in the next micro-cycle, sets the phase-locked signal suppression time window according to the prediction results, and configures the energy release channel to release the risk energy in advance. The delayed expansion peak clipping module enables an adaptive delay micro-buffer mechanism within the phase-locked signal suppression time window to extend the peak signal in the time domain, thereby reducing the instantaneous slope of the peak signal and interrupting the cumulative triggering path of malfunctions. The residual energy absorption and limiting control module, based on the output signal of the adaptive delay micro-buffer mechanism, initiates the phase foldback limiting process, guides the residual peak signal to the inverted signal confluence path, and dynamically adjusts the limiting processing parameters in combination with the real-time residual signal to achieve continuous dissipation and safe transfer of peak energy.

[0051] The present invention provides a multimodal collaborative control theory method for cable branch boxes, which is implemented through the above-mentioned multimodal collaborative control theory system for cable branch boxes. For details of the specific method and process of the multimodal collaborative control theory system for cable branch boxes, please refer to the above-mentioned embodiment of the multimodal collaborative control theory method for cable branch boxes, which will not be repeated here.

[0052] The foregoing has only described certain exemplary embodiments of the present invention by way of illustration. Undoubtedly, those skilled in the art can modify the described embodiments in various ways without departing from the spirit and scope of the present invention. Therefore, the foregoing drawings and descriptions are illustrative in nature and should not be construed as limiting the scope of protection of the claims of the present invention.

Claims

1. A multi-modal cooperative control method for cable branch boxes, characterized in that, Includes the following steps: Low-amplitude synchronous detection pulses are injected into the operating environment of the cable branch box, and the feedback response signal of the detection pulses is collected. The arrival sequence of the interference signal is inverted based on the feedback response signal to construct a micro-timeslot interference profile and generate a time-series baseline for interference identification. Based on the timing baseline, the phase consistency index among multiple interference signals is calculated. The trigger time window for the superposition and collapse of interference signals is locked according to the phase consistency index, and a set of candidate spike signals is extracted within the trigger time window. For the candidate spike signal set, high-fidelity data acquisition of the first channel and phase perturbation data acquisition of the second channel are carried out simultaneously. By comparing the amplitude difference of the two channels at the same time point, false spike signals with inconsistent amplitudes are eliminated, and the effective spike signal set for subsequent analysis is obtained. Based on the historical distribution characteristics of the effective spike signal set, a cross-cycle spike signal evolution model is established to predict the spike risk signal that will appear in the next micro-cycle. According to the prediction results, a phase-locked signal suppression time window is set, and an energy release channel is configured to release the risk energy in advance. Within the phase-locked signal suppression time window, an adaptive delay micro-buffer mechanism is enabled to extend the spike signal in the time domain, thereby reducing the instantaneous slope of the spike signal and interrupting the cumulative triggering path of malfunctions. Based on the output signal of the adaptive delay micro-buffer mechanism, the phase return limiting process is initiated to guide the residual peak signal to the inverted signal confluence path. The limiting processing parameters are dynamically adjusted in combination with the real-time residual signal to achieve continuous dissipation and safe transfer of peak energy.

2. The multi-modal collaborative control method for a cable branch box according to claim 1, characterized in that, The steps for generating a time series baseline are as follows: Under steady-state operation conditions of the cable branch box, a synchronous detection pulse with an amplitude of 10 volts to 30 volts, a pulse width of less than 5 microseconds, and a repetition period of 20 milliseconds is injected into the cable line. By setting up four voltage acquisition points at the incoming terminal, adapter, outgoing terminal and load lead-out terminal, the response time, amplitude and waveform trajectory of the detection pulse are collected respectively, and a signal propagation sequence spectrum is established with the response time of the incoming terminal as the zero point. By normalizing the response time of multiple acquisition points, normalizing the amplitude, and smoothing the waveform through interpolation, a standardized timing baseline is constructed with the incoming terminal as the starting node. The established time baseline is used as the time discrimination benchmark for interference identification. In subsequent steps, it is used to compare the degree of deviation of the actual operating signal in response time and propagation sequence to identify the trigger precursors of spike interference.

3. The multi-modal cooperative control method for a cable branch box according to claim 2, characterized in that, The steps for extracting the candidate spike signal set are as follows: Multiple cycles of detection pulse response data were collected at the incoming terminal, outgoing connector, intermediate connecting copper busbar and load lead-out terminal, and the phase offset sequence of each sampling point was constructed with the first response time of the incoming terminal as a reference. Periodic differential processing is performed on the phase shift sequence to identify whether multiple sampling points show a phase synchronization convergence trend within a given time period, and to detect the degree of aggregation of the response times of each point; When the response time offset of multiple sampling points is less than 300 nanoseconds, the amplitude change rate exceeds 50 volts per microsecond, and the energy density of the main waveform is significantly higher than five times the background noise in the same time interval, this time interval is marked as the trigger time window for the superposition and collapse of interference signals. Within the trigger time window, based on the voltage signal rise rate, amplitude change degree, duration and periodic repetition characteristics of each sampling point, peak signal segments that meet the conditions of high change and high consistency are selected to form a candidate peak signal set, and the corresponding time, amplitude and location parameters are marked.

4. The multi-modal collaborative control method for a cable branch box according to claim 3, characterized in that, The steps for generating the effective spike signal set are as follows: High-fidelity acquisition channels and phase disturbance acquisition channels are set at the incoming terminal, outgoing copper bus, intermediate transition copper bus and load lead terminal respectively. The two channels simultaneously acquire the voltage waveforms corresponding to the candidate spike signals. The candidate spike signals acquired from the two channels are time-domain aligned and amplitude normalized. The amplitude difference of each sampling point is calculated and the waveform consistency characteristics are analyzed. When the amplitude difference of more than 70% of the total number of sampling points in two channels is less than 5% of the maximum amplitude, and the amplitude difference at the peak point is less than 3% of the maximum amplitude, and the above conditions are met in three consecutive detection cycles, the candidate spike signal is identified as a valid spike signal and is retained. All candidate spike signals that meet the consistency requirements are combined into a valid spike signal set, and their occurrence time, sampling point number, peak voltage and stability level parameters are recorded.

5. The multi-modal cooperative control method for a cable branch box according to claim 4, characterized in that, The specific steps for setting a phase-locked signal suppression time window based on the prediction results and configuring an energy release channel to mitigate risk energy in advance are as follows: Extract the time position, amplitude, rise rate and propagation path of effective spike signals within multiple consecutive working cycles, construct a time density map of spike signals and analyze the high-frequency clustering area and path repetition characteristics within the cycle; Based on the time position offset, amplitude change trend and path stability, a peak signal evolution feature group is constructed to predict the time interval of the peak risk signal in the next micro-cycle; Set the phase-locked signal suppression time window according to the predicted time interval, pause the high-sensitivity fault identification process, and start the instantaneous waveform buffering mechanism; Within the phase-locked signal suppression time window, a parallel resistor network, a varistor, and a surge absorption inductor are connected to construct an energy dissipation channel, guiding the energy of the spike signal to dissipate along a low-resistance path and reducing its impact on the main line.

6. The multi-modal cooperative control method for a cable branch box according to claim 5, characterized in that, Within the phase-locked signal suppression time window, the adaptive delay micro-buffer mechanism is enabled to extend the peak signal in the time domain. The steps are as follows: After the phase-locked signal suppression time window is activated, the spike signal collected within the window is buffered. The buffer unit has a capacity of 1,024 bytes, supports nanosecond-level read and write response capability, and continuously records voltage values, timestamps and rising edge change rates. The buffer period is consistent with the phase-locked window. The buffered signal is subjected to a delay expansion operation, which extends the sampling interval to three to five times for signal segments with voltage changes exceeding 80 volts within one microsecond, reduces the slope of the rising segment to one-third to one-fifth of the original, and reconstructs the waveform of the falling segment with an amplitude drop of more than 60 volts within one microsecond into a gradual descent platform to ensure a smooth signal shape. The extended signal is re-injected into the protection discrimination path, and logical judgment is made based on the action threshold and action duration to verify that the extended signal no longer triggers malfunctions, thereby cutting off the trigger chain formed by the continuous superposition of spike signals.

7. The multi-modal cooperative control method for a cable branch box according to claim 6, characterized in that, The steps for initiating the phase foldback limiting process based on the output signal using the adaptive delay micro-buffering mechanism are as follows: Using the extended output signal as input, the residual spike signal in the waveform with an amplitude greater than the preset warning threshold is identified, and an inverse signal with the same amplitude, consistent frequency, time synchronization and opposite voltage direction is generated in the physical circuit. The signal is then guided to the energy absorption end for dissipation through the parallel branch path. While the inverted signal path is started, the limiting voltage suppression circuit of the main channel dynamically adjusts the action threshold and response time according to the amplitude, duration and slope characteristics of the residual signal to achieve voltage spike clipping and delay suppression. The temperature rise, current change rate, and impedance drift of the reverse signal guide path are monitored. When the temperature rise exceeds the set value or the current fluctuates in the opposite direction, the control unit switches to the backup attenuation path to ensure that the residual energy is completely absorbed and safely transferred.

8. A multimodal cooperative control system for a cable branch box, used to implement the multimodal cooperative control method for a cable branch box as described in any one of claims 1-7, characterized in that, It includes modules for interference profiling, interference focusing identification, peak verification and cleaning, peak prediction and suppression planning, delayed expansion and peak reduction, and residual energy absorption and limiting control. The interference profile construction module injects low-amplitude synchronous detection pulses into the operating environment of the cable branch box, collects the feedback response signal of the detection pulses, inverts the arrival sequence of the interference signal based on the feedback response signal, constructs a micro-timeslot interference profile, and generates a time-series baseline for interference identification. The interference focusing identification module calculates the phase consistency index between multiple interference signals based on the time-series baseline, locks the trigger time window of interference signal superposition and collapse according to the phase consistency index, and extracts a set of candidate spike signals within the trigger time window. The spike verification and cleaning module simultaneously acquires high-fidelity data from the first channel and phase perturbation data from the second channel for the candidate spike signal set. By comparing the amplitude differences of the two channels at the same time point, it removes false spike signals with inconsistent amplitudes and obtains a set of effective spike signals for subsequent analysis. The peak prediction and suppression planning module establishes a cross-cycle peak signal evolution model based on the historical distribution characteristics of the effective peak signal set, predicts the peak risk signal that will appear in the next micro-cycle, sets the phase-locked signal suppression time window according to the prediction results, and configures the energy release channel to release the risk energy in advance. The delayed expansion peak clipping module enables an adaptive delay micro-buffer mechanism within the phase-locked signal suppression time window to extend the peak signal in the time domain, thereby reducing the instantaneous slope of the peak signal and interrupting the cumulative triggering path of malfunctions. The residual energy absorption and limiting control module, based on the output signal of the adaptive delay micro-buffer mechanism, initiates the phase foldback limiting process, guides the residual peak signal to the inverted signal confluence path, and dynamically adjusts the limiting processing parameters in combination with the real-time residual signal to achieve continuous dissipation and safe transfer of peak energy.

Images