A multi-modal wireless synchronized grid-connected control system and method

Abstract

The application discloses a kind of multimodal wireless synchronization grid-connected control system and method, it is related to grid-connected control technical field.This kind of multimodal wireless synchronization grid-connected control system and method, including S1, to in-situ electrical data, real-time link data and historical criterion data are preprocessed;S2, fusion circuit breaker type identification, operating condition perception and communication link discrimination, construct grid operation decision;S3, to in-situ electrical data is distributed in-situ multimodal perception, constructs in-situ adaptive health perception system;S4, by cross-link information fusion, constructs self-healing type decision link cooperation, enters regulation approximation degree criterion and closing control module;S5, fusion high-order convergence trend analysis, carries out closing window self-determination and end side closed-loop drive.Solve the existing scheme cannot realize terminal side in-situ electrical state adaptive perception and local criterion fusion, resulting in criterion completely dependent on remote link, system synchronization and closing are susceptible to link delay and step-out influence, overall reliability is insufficient problem.

Description

Technical Field

[0001] This invention relates to the field of grid-connected control technology, specifically to a multimodal wireless synchronous grid-connected control system and method. Background Technology

[0002] With the rapid development of distributed energy, emergency power, and smart grid technologies, the scale of power equipment connected to distribution networks continues to expand, and the automation and digitalization levels of distribution network operation have significantly improved. In actual operation and maintenance and dispatching, the industry has widely applied various technologies such as on-site monitoring, status awareness, remote communication, and equipment control to meet the operation management and safety requirements when power generation equipment is connected to the main grid, promoting the intelligent and refined operation of the power system. At the same time, with the diversification of equipment types and the increasing complexity of the field environment, the industry's demand for information collection, operation status determination, and remote control capabilities is constantly increasing. Distribution network grid connection and switching operations are continuously evolving towards greater efficiency, intelligence, and safety, laying a solid foundation for ensuring the safe, stable, and efficient operation of the power system.

[0003] For example, utility model patent CN217984554U discloses a mobile generator grid-connected system, including a central control unit, a high-voltage line, at least one low-voltage line, at least one high-voltage mobile generator set, and at least one low-voltage mobile generator set. The central control unit is connected to either the high-voltage line or the low-voltage line. The high-voltage mobile generator sets are connected to the high-voltage line, and the low-voltage mobile generator sets are connected to the low-voltage line. The low-voltage lines are connected to the high-voltage line via transformers. The central control unit is equipped with a grid-connected control cabinet that wirelessly communicates with the output control cabinets on the high-voltage and low-voltage generator sets. The system is internally equipped with a controller, a communicator, and a synchronization detection module, enabling flexible grid connection and synchronization detection of high- and low-voltage mobile generator sets, and providing centralized management of the output control cabinets, effectively improving the grid-connection efficiency and management convenience of the mobile generators.

[0004] For example, invention patent CN117293879A discloses a grid-connected control method and device for an energy storage grid-connected system. The method includes: first, determining the initial voltage value of the DC voltage bus, and then adding voltage variables to obtain a new bus voltage value; then, determining the control voltage value of each energy storage unit. By judging the changes in the conversion efficiency of the energy storage units and the overall conversion efficiency of the system, dynamic control of the power converter of the energy storage units is achieved. The system can adjust the bus voltage and control voltage according to the output parameters and efficiency thresholds of each energy storage unit, optimize the power flow and energy efficiency distribution during grid connection, and improve the stability and adaptability of the energy storage grid-connected system.

[0005] While existing generator grid-connected systems and energy storage grid-connected control methods can enable the access and grid connection of various power generation devices and energy storage units, they still have limitations in terms of field adaptability, communication link stability, real-time criterion fusion, and local intelligent decision-making capabilities. Existing systems largely rely on centralized remote criterion calculations, lacking real-time adaptive perception of electrical status and redundant criterion coordination at the terminal side. This makes it difficult to effectively address issues such as network latency, link fluctuations, and device heterogeneity in complex scenarios, and the overall grid-connected reliability and security need further improvement.

[0006] Therefore, in order to address the above problems, there is an urgent need for a multimodal wireless synchronous grid-connected control system and method. Summary of the Invention

[0007] Technical problems to be solved

[0008] To address the shortcomings of existing technologies, this invention provides a multimodal wireless synchronous grid-connected control system and method, which solves the problem that existing solutions cannot achieve adaptive sensing and local criterion fusion of the in-situ electrical status of the terminal side, resulting in the criterion being completely dependent on the remote link, and the system synchronization and closing being easily affected by link delay and loss of synchronization, leading to insufficient overall reliability.

[0009] To achieve the above objectives, the present invention provides the following technical solution: a multimodal wireless synchronous grid-connected control method, comprising: S1, collecting in-situ electrical data and real-time link data, and obtaining historical criterion data; preprocessing the in-situ electrical data, real-time link data, and historical criterion data; S2, integrating circuit breaker type identification, operating condition perception, and communication link discrimination to construct a reliable grid-connected operation intelligent decision-making system; S3, performing distributed in-situ multimodal perception and heterogeneous signal sliding statistics on the in-situ electrical data to construct a smart terminal-side time-series signal fusion and in-situ adaptive health perception system; S4, constructing a self-healing decision link collaboration through cross-link information fusion and primary / backup emergency multi-channel criterion optimization, and entering the adjustment approximation criterion and closing control module; S5, based on the in-situ electrical data, real-time link data, and historical criterion data, integrating high-order convergence trend analysis to perform closing window self-judgment and end-side closed-loop drive.

[0010] Furthermore, in-situ electrical data and real-time link data are collected to obtain historical criterion data; the specific process of preprocessing the in-situ electrical data, real-time link data, and historical criterion data is as follows: In-situ electrical data is collected, including AC voltage signals from the mains, AC voltage signals from the generator, AC current signals from the mains, and AC current signals from the generator. Real-time link data is also collected, including round-trip time of the mains link heartbeat packet, round-trip time of the generator link heartbeat packet, temperature of the mains equipment, temperature of the generator equipment, signal-to-noise ratio of the mains link, and signal-to-noise ratio of the generator link. Historical criterion data is acquired and a historical criterion database is established, including historical mains communication link delay sequences, historical generator communication link delay sequences, historical instantaneous phase angle difference at the center of the sliding window, historical phase angle difference change rate, health sliding window, redundant sliding window, and approximation sliding window. Missing values ​​are repaired and outliers are identified in the real-time link data through interpolation completion and extreme value removal algorithms. Data synchronization and window correspondence are performed on the historical criterion data through multi-step sampling alignment and time-series window mapping. The in-situ electrical data, real-time link data, and historical criterion data are standardized and normalized through distribution standardization and linear normalization algorithms.

[0011] Furthermore, by integrating circuit breaker type identification, operating condition perception, and communication link discrimination, the specific process for constructing a reliable intelligent decision-making system for grid connection operations is as follows: It supports three working modes: adaptive wireless intelligent grid connection, human-machine interactive visual-assisted grid connection criterion, and wired redundancy protection for extreme scenarios. In the adaptive wireless grid connection mode, the status of the electric circuit breaker is monitored in real time, and the adjustment proximity criterion and closing control are completed. Closing commands are sent in real time via the wireless network, achieving unattended grid connection throughout the entire process. The human-machine interactive visual-assisted grid connection criterion mode is suitable for non-electric circuit breakers and minimally sized wiring scenarios. It performs real-time synchronous adjustment proximity value criterion calculation and displays the closing window visually on the terminal interface. The operator manually completes the circuit breaker closing based on the parameter instructions. The wired redundancy protection mode for extreme scenarios is used for wireless communication anomalies and extreme scenarios. It switches to wired control and completes the safe closed loop of the closing action through a physical wired link.

[0012] Furthermore, the specific process of performing distributed in-situ multimodal sensing and heterogeneous signal sliding statistics on in-situ electrical data is as follows: Acquire the AC voltage signal of the mains power supply and the AC voltage signal of the generator; use the Discrete Fourier Transform (DFT) algorithm to extract the effective value of the fundamental wave of the mains power supply and the AC voltage signal of the generator, and obtain the mains power voltage sequence and the generator voltage sequence within the health sliding window length; use zero-crossing detection and linear interpolation to obtain the mains power side frequency sequence and the generator power side frequency sequence within the health sliding window length; use the Fast Hilbert Transform and synchronous sampling clock alignment to obtain the mains power side phase angle sequence and the generator power side phase angle sequence within the health sliding window length; use the k-nearest neighbor mutual information estimation algorithm to obtain the mutual information value of the sliding window signal from the mains power voltage sequence and the generator power voltage sequence; use the health sliding window variance statistical algorithm to obtain the sliding window variance of the mains power voltage sequence and the generator power voltage sequence. The square root of the sum of squares of the differences between the mains-side frequency sequence and the generator-side frequency sequence is calculated and added to 1 to obtain the frequency offset adjustment term. The frequency offset adjustment term is multiplied by the mutual information value of the sliding window signal to obtain the numerator of the cooperative enhancement term. The root mean square difference term between the mains-side phase angle sequence and the generator-side phase angle sequence is calculated, and the root root of the mains-side voltage sliding window variance and the generator-side voltage sliding window variance is superimposed to obtain the voltage variance composite term. The root mean square difference term and the voltage variance composite term are multiplied and added to 1 to obtain the denominator correction term. Finally, the cooperative enhancement numerator is divided by the denominator correction term to obtain the signal cooperative stability value.

[0013] Furthermore, the specific process of constructing a time-series signal fusion and in-situ adaptive health perception system on the intelligent terminal side is as follows: real-time comparison of signal collaborative stability value and signal collaborative stability threshold, which includes a primary stability threshold and a secondary stability threshold: when the signal collaborative stability value is less than the secondary stability threshold, the local relay is triggered to trip through the intelligent grid-connected operation and maintenance terminal, the remote and local closing control permissions are locked simultaneously, a grid-connected anomaly database is created and the in-situ electrical data and real-time link data are recorded to the grid-connected anomaly database, and the system switches to the extreme scenario wired redundancy protection mode; when the signal collaborative stability value is greater than or equal to the secondary stability threshold and less than the primary stability threshold, the local sampling frequency is increased, the digital filtering and wavelet denoising units are activated, the current sampled signal is filtered, and the system switches to the human-machine interactive visual auxiliary grid-connected judgment mode; when the signal collaborative stability value is greater than or equal to the primary stability threshold, the main channel participates in the synchronous judgment closed loop and grid-connected decision-making, and the normal speed regulation, voltage regulation, and closing closed loop are maintained.

[0014] Furthermore, the specific process of cross-link information fusion and primary / backup emergency multi-channel criterion selection is as follows: Acquire the AC voltage signal, AC current signal, generator AC voltage signal, generator AC current signal, AC heartbeat packet round-trip time, generator AC heartbeat packet round-trip time, historical AC communication link delay sequence, historical generator communication link delay sequence, AC equipment temperature, generator equipment temperature, AC link signal-to-noise ratio, and generator link signal-to-noise ratio; extract the fundamental effective value of the AC voltage signal, AC current signal, generator AC voltage signal, and generator AC current signal using the DFT algorithm, and then obtain the average active power of the AC power and the average active power of the generator through the fundamental active power calculation method and the redundancy sliding window mean statistics; obtain the AC communication link delay by collecting the AC heartbeat packet round-trip time and generator AC heartbeat packet round-trip time in real time. The product of the signal cooperative stability value and the mutual information value of the sliding window signal is used to obtain the criterion coordination term. The absolute value of the difference between the average active power of the mains and the average active power of the generator is calculated, plus one, and the natural logarithm is taken, plus one again to obtain the active power deviation term. The absolute value of the difference between the delay of the mains communication link and the delay of the generator communication link is calculated, multiplied by the link delay sensitivity coefficient, and the negative number is used as an exponential function to obtain the link delay attenuation term. The criterion coordination term, the active power deviation term, and the link delay attenuation term are multiplied together to obtain the link compensation coordination term. The square root of the absolute value of the difference between the temperature of the mains equipment and the temperature of the generator equipment is multiplied by the sum of the reciprocals of the signal-to-noise ratio of the mains link and the signal-to-noise ratio of the generator link, and then multiplied by the temperature deviation term and the signal-to-noise ratio attenuation term, plus one, to obtain the environmental impact term. The link compensation coordination term is divided by the environmental impact term to obtain the redundancy synchronization criterion value.

[0015] Furthermore, the specific process of constructing a self-healing decision link collaboration and entering the adjustment proximity criterion and closing control module is as follows: Real-time comparison of redundant synchronization criterion values ​​and redundant synchronization criterion thresholds, including primary criterion thresholds and secondary criterion thresholds: When the redundant synchronization criterion value is less than the secondary criterion threshold, the health status and synchronization criterion decision are mismatched, the closing control authority is locked, the risk link is disconnected, and de-rating, isolation, and soft disconnection processing are performed, switching to the human-machine interactive visual auxiliary grid connection criterion mode; When the redundant synchronization criterion value is greater than or equal to the secondary criterion threshold and less than the primary criterion threshold, the command communication increases the transmission power and reduces the data transmission rate until the redundant synchronization criterion value is greater than or equal to the primary criterion threshold; When the redundant synchronization value is greater than or equal to the primary criterion threshold, the main criterion link and sensing channel are healthy, and the adjustment proximity criterion and closing control module is entered.

[0016] Furthermore, based on in-situ electrical data, real-time link data, and historical criterion data, the specific process of fusing high-order convergence trend analysis is as follows: Obtain the instantaneous phase angle difference at the center of the historical sliding window and the rate of change of the historical phase angle difference; align the phase angle sequences of the mains side and the generator side using a synchronous sampling clock, and extract the phase angle difference at the center of the sliding window to obtain the instantaneous phase angle difference at the center of the sliding window; obtain the rate of change of the phase angle difference at the center of the sliding window using a first-order difference algorithm approximating the sliding window; obtain the mean frequency of the mains side and the mean frequency of the generator side using a mean algorithm for the mains side and the generator side; obtain the mean effective value of the mains side sliding window voltage and the mean effective value of the generator side sliding window voltage using a mean algorithm for the mains voltage sequence and the generator voltage sequence. The absolute value of the instantaneous phase angle difference at the center of the sliding window is multiplied by the phase difference sensitivity coefficient to obtain the synchronization phase offset suppression term; the absolute value of the rate of change of the phase angle difference is multiplied by the rate of change of the phase difference sensitivity coefficient to obtain the phase trend change suppression term; the synchronization phase offset suppression term and the phase trend change suppression term are added together as the negative exponent of the exponential function, and multiplied by the negative exponent with the signal cooperative stability value as the base to obtain the synchronization convergence enhancement numerator; the absolute value of the difference between the average frequency values ​​of the mains side and the generator side, and the absolute value of the difference between the average effective values ​​of the sliding window voltages of the mains side and the generator side are calculated, and the two absolute values ​​are added together and one is added to obtain the denominator of the synchronization frequency and voltage deviation; the synchronization convergence enhancement numerator is divided by the denominator of the synchronization frequency and voltage deviation to obtain the synchronization adjustment approximation value.

[0017] Furthermore, the specific process of self-judgment of the closing window and end-side closed-loop drive is as follows: real-time comparison of the synchronous adjustment approximation value and the synchronous adjustment approximation threshold, which includes a first-level approximation threshold and a second-level approximation threshold: when the synchronous adjustment approximation value is greater than the first-level approximation threshold, the generator set is driven to continuously optimize speed and excitation control, reduce the frequency difference in real time, and synchronously monitor the phase difference to dynamically determine the closing window. After the criteria are met, the closing action is triggered; when the synchronous adjustment approximation value is greater than the second-level approximation threshold and less than or equal to the first-level approximation threshold, the generator speed regulation and voltage regulation amplitude are increased. If the phase difference, frequency difference and trend indicators are met for k consecutive approximation sliding windows, the closing action is triggered; when the synchronous adjustment approximation value is less than or equal to the second-level approximation threshold, the closing control authority is locked, the original electrical data and real-time link data are recorded to the grid connection anomaly database, and the system is switched to the human-machine interactive visual auxiliary grid connection criterion mode.

[0018] The second aspect of this invention provides a multimodal wireless synchronous grid-connected control method, including an acquisition and preprocessing module for acquiring in-situ electrical data and real-time link data, and obtaining historical criterion data; preprocessing the in-situ electrical data, real-time link data, and historical criterion data; a multimodal grid-connected and switching mechanism module for integrating circuit breaker type identification, operating condition perception, and communication link discrimination to construct a reliable intelligent decision-making system for grid-connected operations; a signal health discrimination and prevention module for performing distributed in-situ multimodal perception and heterogeneous signal sliding statistics on in-situ electrical data to construct a smart terminal-side time-series signal fusion and in-situ adaptive health perception system; a redundancy criterion and synchronous health judgment module for constructing a self-healing judgment link collaboration through cross-link information fusion and primary / backup emergency multi-channel criterion optimization, and entering the adjustment approximation criterion and closing control module; and an adjustment approximation criterion and closing control module for performing closing window self-judgment and end-side closed-loop drive based on in-situ electrical data, real-time link data, and historical criterion data, integrating high-order convergence trend analysis. Beneficial effects

[0019] The present invention has the following beneficial effects:

[0020] (1) This invention improves the real-time response capability to on-site operating conditions by realizing adaptive sensing of the in-situ electrical state at the terminal side, thereby achieving a highly intelligent and flexible adaptation effect in on-site operations, and effectively solving the problem of insufficient on-site adaptability in the prior art.

[0021] (2) This invention improves the real-time response capability to on-site operating conditions by realizing adaptive sensing of the in-situ electrical state at the terminal side, thereby achieving a highly intelligent and flexible adaptation effect in on-site operations, effectively solving the problem of insufficient on-site adaptability in the prior art.

[0022] (3) The present invention introduces a primary / backup criterion and a multi-channel switching scheme, which can dynamically select the optimal grid connection mode under different working conditions, thereby achieving seamless switching and continuity of grid connection operation, effectively solving the problems of single mode and insufficient emergency support in the prior art.

[0023] (4) This invention achieves full-process localization of criterion decision-making and closing control through edge intelligent processing and local closed-loop control, thereby achieving extremely low latency and high real-time performance, effectively solving the problem of lagging remote criterion decision-making in the prior art.

[0024] Of course, any product implementing this invention does not necessarily need to achieve all of the advantages described above at the same time. Attached Figure Description

[0025] Figure 1 This is a flowchart of a multimodal wireless synchronous grid-connected control method according to the present invention;

[0026] Figure 2 This is a structural diagram of a multimodal wireless synchronous grid-connected control system according to the present invention;

[0027] Figure 3 This is a schematic diagram of the multimodal wireless synchronous network architecture of the present invention;

[0028] Figure 4 This is a line graph comparing the wireless synchronization parameters of the present invention.

[0029] Figure 5 This is a flowchart of the redundancy synchronization criterion decision-making process of the present invention. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] Please see Figures 1-5 This invention provides a technical solution: a multimodal wireless synchronous grid-connected control method, comprising: S1, collecting in-situ electrical data and real-time link data, and obtaining historical criterion data; preprocessing the in-situ electrical data, real-time link data, and historical criterion data; S2, integrating circuit breaker type identification, operating condition perception, and communication link discrimination to construct a reliable grid-connected operation intelligent decision-making system; S3, performing distributed in-situ multimodal perception and heterogeneous signal sliding statistics on the in-situ electrical data to construct a smart terminal-side time-series signal fusion and in-situ adaptive health perception system; S4, constructing a self-healing decision link collaboration through cross-link information fusion and primary / backup emergency multi-channel criterion optimization, and entering the adjustment approximation criterion and closing control module; S5, based on the in-situ electrical data, real-time link data, and historical criterion data, integrating high-order convergence trend analysis to perform closing window self-judgment and end-side closed-loop drive.

[0032] Specifically, in-situ electrical data and real-time link data are collected, and historical criterion data is obtained. The specific process for preprocessing the in-situ electrical data, real-time link data, and historical criterion data is as follows: Figure 3The diagram shows a multimodal wireless synchronous grid-connected architecture. It collects in-situ electrical data, including AC voltage signals from the mains, AC voltage signals from the generator, AC current signals from the mains, and AC current signals from the generator. It also collects real-time link data, including round-trip time of the mains link heartbeat, round-trip time of the generator link heartbeat, temperature of the mains equipment, temperature of the generator equipment, signal-to-noise ratio (SNR) of the mains link, and signal-to-noise ratio (SNR) of the generator link. The mains side refers to the incoming line from the public power grid to the distribution cabinet and load terminals in the transformer area. Generator-side data is collected by the generator set and intelligent controller; the generator side refers to the local distributed generation unit and its output side. The mains-side data is synchronously collected by the wireless intelligent grid-connected operation and maintenance terminal, providing the raw data foundation for subsequent edge computing. A low-power wireless communication network connects the two components and is responsible for transmitting the calculated instructions and data information.

[0033] Historical criterion data is acquired and a historical criterion database is established. The historical criterion data includes: historical mains communication link delay sequence, historical power generation communication link delay sequence, historical instantaneous phase angle difference at the center of the sliding window, historical phase angle difference change rate, health sliding window, redundant sliding window, and approximation sliding window. The database provides time-series trend analysis and benchmark reference for the adaptive criterion calculation of the intelligent operation and maintenance terminal.

[0034] By employing interpolation completion and extreme value removal algorithms, missing values ​​are repaired and outliers are identified in real-time link data. Lost heartbeat packet data is completed using linear interpolation based on a sliding time window, and abnormal outliers causing communication delays are identified and removed according to the Laida criterion. Through multi-step sampling alignment and time-series window mapping, historical criterion data is synchronized and windowed to ensure that the instantaneous phase angle difference sequence data at the center of the historical sliding window is accurately aligned with the currently acquired electrical data on the time scale. Through distribution standardization and linear normalization algorithms, in-situ electrical data, real-time link data, and historical criterion data are standardized and normalized. The Z-value standardization method is used to eliminate the influence of dimensions, and bounded data such as phase angles are scaled to the [0,1] interval using deviation normalization.

[0035] In this implementation plan, the data preprocessing stage systematically cleans, aligns, and standardizes the multi-source heterogeneous data collected by the intelligent controller on the power generation side and the maintenance terminal on the grid side. This builds a high-quality, highly consistent data foundation for subsequent intelligent decision-making, effectively eliminating noise and biases introduced by communication packet loss, signal interference, and differences in units. Furthermore, it integrates historical experience data with real-time monitoring information within a unified time-series framework, thereby transforming the raw, coarse field data into a standardized information source that can be reliably perceived, accurately judged, and intelligently decided in subsequent stages. This fundamentally improves the reliability of the perceived data and the stability of the decision-making basis.

[0036] Specifically, the process of constructing a reliable grid connection intelligent decision-making system by integrating circuit breaker type identification, operating condition perception, and communication link discrimination is as follows: It supports three working modes: adaptive wireless intelligent grid connection, human-machine interactive visual assisted grid connection criterion, and wired redundancy guarantee for extreme scenarios. These three modes together constitute the multi-modal grid connection switching mechanism of this invention, which can seamlessly switch according to on-site equipment conditions and communication status. In the adaptive wireless grid connection mode, the status of the electric circuit breaker is monitored in real time, and the adjustment of the proximity criterion and closing control are completed. Closing commands are sent in real time through the wireless network, realizing unattended grid connection throughout the entire process. The core lies in the intelligent operation and maintenance terminal driving the electric operating mechanism of the mains-side circuit breaker through the wireless network; the human-machine interactive visual... The auxiliary grid connection criterion mode is suitable for non-electric circuit breakers and extremely simple wiring applications. It performs real-time synchronization adjustment and approximation value criterion calculation and displays the closing window on the terminal interface. The operator manually completes the circuit breaker closing according to the parameter instructions. In the early stage, speed and voltage adjustment are performed. In the closing stage, the synchronization table is dynamically displayed on the terminal interface to guide the operator's manual operation, which not only ensures synchronization accuracy but also eliminates the complex wiring of connecting the circuit breaker control lines. The extreme scenario wired redundancy guarantee mode is used for wireless communication failures and extreme scenarios. It switches to wired control and completes the safety closed loop of the closing action through the physical wired link. The traditional wired backup synchronization mode ensures the grid connection operation is foolproof by restoring the traditional two-circuit cable connection method.

[0037] This implementation plan establishes a complete multimodal grid connection switching mechanism. The core value of this mechanism lies in its superior environmental adaptability and operational flexibility: it can intelligently select and seamlessly switch to the optimal grid connection scheme based on the actual equipment configuration and real-time communication link status on-site. The adaptive wireless mode achieves fully automated closed-loop control, maximizing operational efficiency; the human-machine interaction mode, while maintaining the accurate judgment of the core algorithm, reduces the operational threshold and wiring complexity through visual guidance; and the wired redundancy mode serves as the ultimate safety guarantee, ensuring reliable grid connection operation even in extreme conditions. This layered, progressive, and multi-layered design significantly broadens the scope of application, constructing a defense-in-depth system at the architectural level, thus providing a solid guarantee for the safety, reliability, and ultimate success rate of the entire grid connection process under various complex operating conditions and abnormal scenarios.

[0038] Specifically, the process of distributed in-situ multimodal sensing and heterogeneous signal sliding statistics for in-situ electrical data is as follows: First, acquire the AC voltage signals of the mains and generators to provide the core raw electrical signals for subsequent multimodal sensing. Second, use a Discrete Fourier Transform (DFT) algorithm to extract the effective value of the fundamental wave of the AC voltage signals from the mains and generators. Within the health sliding window, obtain the mains voltage sequence and generator voltage sequence, effectively filtering out grid harmonic interference and thus accurately extracting the basic waveform amplitude information characterizing power quality. Third, use zero-crossing detection and linear interpolation to obtain the mains-side frequency sequence and generator-side frequency sequence within the health sliding window. Estimate the signal period by calculating the time interval between consecutive zero-crossings, and combine linear interpolation to improve accuracy in non-integer cycles. The frequency measurement accuracy under periodic sampling conditions is improved. The mains AC voltage signal and the generator AC voltage signal are aligned using the Fast Hilbert Transform and synchronous sampling clock. Within the health sliding window length, the mains-side phase angle sequence and the generator-side phase angle sequence are obtained. An analytical signal can be constructed to instantaneously track the phase. Synchronous sampling clock alignment is a key prerequisite for ensuring the accuracy of the phase difference calculation between the two sides. The mutual information value of the sliding window signal is obtained from the mains voltage sequence and the generator voltage sequence using the k-nearest neighbor mutual information estimation algorithm, quantifying the statistical nonlinear dependence of the two voltage sequences from an information theory perspective. The sliding window variance of the mains voltage sequence and the generator voltage sequence is obtained from the health sliding window variance statistical algorithm, characterizing the degree of fluctuation of the mains-side and generator-side voltages within the observation window.

[0039] The square root of the sum of squares of the differences between the mains-side and generator-side frequency sequences is calculated and added to 1 to obtain the frequency offset adjustment term. This frequency offset adjustment term is multiplied by the mutual information value of the sliding window signal to obtain the numerator of the cooperative enhancement term. The numerator integrates frequency differences and signal cooperativeness; the smaller the frequency difference and the higher the mutual information value, the larger the numerator, and the more ideal the synchronization state. The root mean square difference (RMS) of the phase angle sequences of the mains-side and generator-side sequences is calculated, and the sum of the mains-side and generator-side voltage sliding window variances is superimposed and squared to obtain the voltage variance composite term. The RMS and voltage variance composite terms are multiplied and added to 1 to obtain the denominator correction term, which integrates the adverse factors of phase misalignment and voltage fluctuations; a larger value indicates more synchronization obstacles. Finally, the cooperative enhancement numerator is divided by the denominator correction term to obtain the signal cooperative stability value. The specific calculation formula is as follows: ;

[0040] In the formula, This represents the signal coordination stability value, reflecting the overall synchronization quality and health status of the electrical signals between the mains power supply and the generator side; This represents the effective value sequence of the mains voltage sliding window, which quantifies the amplitude level of the mains voltage within the observation window and serves as the basis for evaluating voltage stability and amplitude difference criteria. This represents the sequence of effective values ​​of the sliding window voltage on the generator side, characterizing the amplitude characteristics of the generator output voltage; It represents the mains power side sliding window frequency sequence, which describes the dynamic change process of the mains power frequency. Its stability is the core manifestation of the power quality of the power grid. This represents the frequency sequence of the generator side sliding window, reflecting the regulation performance and tracking capability of the generator set's output frequency; It represents the instantaneous phase angle sequence of the mains power sliding window, accurately depicting the phase evolution trajectory of the mains voltage waveform; It represents the instantaneous phase angle sequence of the generator side sliding window, which dynamically describes the phase state of the generator voltage; This represents the mutual information value of the sliding window signal, which measures the strength of the statistical nonlinear correlation between two voltage signals. A higher value indicates better cooperative variation characteristics of the two voltages. The variance of the sliding window representing the effective value of the mains voltage reveals the degree of fluctuation in the mains voltage; the larger the variance, the more unstable the mains signal. This represents the sliding window variance of the effective value of the generator-side voltage, characterizing the fluctuation of the generator output voltage.

[0041] This implementation scheme achieves deep feature extraction and fusion analysis of electrical signals. It accurately extracts the amplitude, frequency, and phase sequences characterizing power quality from the original voltage signal. By calculating the mutual information and variance of the signals, the signal characteristics are quantified from two dimensions: statistical correlation and volatility. Through the designed signal cooperative stability value calculation formula, multiple factors such as frequency difference, phase deviation, and voltage fluctuation are considered together to generate a comprehensive evaluation index that can fully reflect the synchronization quality and health status of the electrical signals on both sides. The index provides the first key state criterion, which can diagnose the signal synchronization status in real time, identify potential loss of synchronization risks, and provide an objective and reliable decision-making basis for the subsequent intelligent switching of grid connection mode and the activation of safety control strategies. Thus, it lays a solid foundation for the entire grid connection control process from the signal perception level.

[0042] Specifically, the process of constructing a time-series signal fusion and in-situ adaptive health perception system on the intelligent terminal side is as follows: real-time comparison of signal collaborative stability values ​​and signal collaborative stability thresholds. The signal collaborative stability thresholds include a primary stability threshold and a secondary stability threshold. The primary stability threshold is set to distinguish between healthy states and states requiring enhanced monitoring, while the secondary stability threshold is used to define safe operating boundaries and fault states requiring emergency intervention.

[0043] When the signal co-stability value is less than the secondary stability threshold, the local relay is triggered to trip by the intelligent grid-connected operation and maintenance terminal. The remote and local closing control permissions are locked simultaneously. A grid-connected anomaly database is created and the original electrical data and real-time link data are recorded in the grid-connected anomaly database. The system switches to the extreme scenario wired redundancy protection mode, which constitutes a strict safety protection mechanism. This mechanism aims to immediately cut off the risk when the signal quality deteriorates severely and retain fault data for analysis. At the same time, the highest reliability backup mode is activated.

[0044] When the signal co-stabilization value is greater than or equal to the secondary stabilization threshold and less than the primary stabilization threshold, the local sampling frequency is increased, with an upper limit of 2-4 times the original sampling rate, to capture richer signal details without generating excessive data load. Digital filtering and wavelet denoising units are activated, and the number of wavelet denoising decomposition layers is controlled at 3-5 layers to achieve a balance between noise suppression and signal feature preservation. The current sampled signal is filtered, and the system switches to the human-machine interactive visual-assisted grid-connected judgment mode. When the signal quality deteriorates slightly, the data quality is improved by enhancing the signal processing capability, and control is transferred to the operator for final decision-making to achieve safety degradation.

[0045] When the signal coordination stability value is greater than or equal to the first-level stability threshold, the main channel continues to participate in the synchronization criterion closed loop and grid connection decision, and the speed regulation, voltage regulation and closing closed loop continue normally, indicating that the electrical signal is in a good health state and supports the stable operation of the wireless intelligent grid connection mode.

[0046] In this implementation plan, a complete in-situ adaptive health perception and response system is constructed by establishing a real-time comparison between the signal cooperative stability value and two-level thresholds. This system can execute a tiered closed-loop control strategy based on the real-time status of the signal quality: immediately initiating the highest level of safety protection when the signal is severely abnormal, cutting off risks and switching to redundancy mode; achieving safety degradation by enhancing signal processing and data quality improvement when the signal is slightly degraded, transferring control to manual decision-making; and maintaining closed-loop control when the signal is healthy. This hierarchical, adaptive processing mechanism realizes intelligent management across the entire chain from signal perception to control execution, effectively improving adaptability to different operating conditions and overall operational reliability, providing a solid safety guarantee for subsequent grid connection operations.

[0047] Specifically, the process of cross-link information fusion and primary / backup emergency multi-channel criterion selection is as follows: Acquire AC voltage signal, AC current signal, generator AC voltage signal, generator AC current signal, AC heartbeat packet round-trip time, generator link heartbeat packet round-trip time, historical AC communication link delay sequence, historical generator communication link delay sequence, AC equipment temperature, generator equipment temperature, AC link signal-to-noise ratio, and generator link signal-to-noise ratio to provide comprehensive multi-modal input for cross-link information fusion; extract the effective value of the main fundamental frequency from the AC voltage signal, AC current signal, generator AC voltage signal, and generator AC current signal using the DFT algorithm; then obtain the average active power of the AC power and the average active power of the generator through the main fundamental frequency active power calculation method and redundant sliding window mean statistics, effectively suppressing harmonic interference to power calculation; and obtain a stable index characterizing the energy transmission capacity of both power sources by smoothing instantaneous fluctuations through window averaging; obtain the AC communication link delay by collecting the AC heartbeat packet round-trip time and generator link heartbeat packet round-trip time in real time, reflecting the real-time transmission quality of the wireless communication network.

[0048] The product of the signal coordination stability value and the mutual information value of the sliding window signal is used to obtain the criterion coordination term, which strengthens the contribution of electrical-level synchronization quality. The absolute value of the difference between the average active power of the mains and the average active power of the generator is calculated, plus one, and the natural logarithm is taken, plus one again to obtain the active power deviation term. This amplifies the impact of small power differences while ensuring function smoothness; its value monotonically increases with increasing power deviation. The absolute value of the difference between the mains communication link delay and the generator communication link delay is calculated, multiplied by the link delay sensitivity coefficient, and the negative of the coefficient is used as an exponential function to obtain the link delay attenuation term. This ensures that when the difference in link delay between the two sides increases, the attenuation term decreases sharply, thereby penalizing the connection delay. In cases of signal asymmetry, the link compensation coordination term is obtained by multiplying the criterion coordination term, active power deviation term, and link delay attenuation term. The multiplicative relationship means that degradation of any sub-term directly weakens the overall coordination term. The square root of the absolute value of the temperature difference between the mains equipment and the generator equipment is multiplied by the sum of the reciprocals of the mains link signal-to-noise ratio and the generator link signal-to-noise ratio. This is then multiplied by the temperature deviation term and the signal-to-noise ratio attenuation term, and one is added to obtain the environmental impact term. This quantifies the combined negative impact of environmental and channel noise on stability; the lower the temperature difference and signal-to-noise ratio, the larger the value. The link compensation coordination term is divided by the environmental impact term to obtain the redundancy synchronization criterion value. The specific calculation formula is as follows: ;

[0049] In the formula, The redundancy synchronization criterion value represents the core output of cross-link information fusion, which comprehensively evaluates the overall reliability of synchronous closing after considering communication and environmental factors. This represents the signal coordination stability value, reflecting the synchronization quality and health level of the electrical signal itself; This represents the mutual information value of the sliding window signal, quantifying the statistical dependence between the mains voltage and the generated voltage sequence; It represents the average active power of the mains power, which characterizes the average power transmitted by the mains power side within the observation window and reflects the load and stability of the mains power side; It represents the average active power output of the generator set, which characterizes the average active power output of the generator set; It represents the delay of the power generation communication link and directly measures the response time of the command channel on the power generation side; The link delay sensitivity coefficient is obtained by using a sliding window standard deviation to mean ratio algorithm on historical mains communication link delay sequences and historical generator communication link delay sequences. The value ranges from 0.05 to 1.0. The temperature of the mains-powered equipment directly reflects the operating environment of the mains-side smart terminal. This indicates the temperature of the power generation equipment, reflecting the operating environment of the power generation side controller; The signal-to-noise ratio (SNR) of the mains communication link measures the quality of the wireless communication signal on the mains side. The lower the SNR, the higher the risk of data transmission errors. It represents the signal-to-noise ratio of the power generation communication link, which measures the channel quality of wireless communication on the power generation side.

[0050] In this embodiment, Table 1 is a data table of wireless synchronization grid connection criteria parameters, which records in detail the signal coordination stability value, sliding window signal mutual information value, mains active power, generator active power and redundancy synchronization criteria value under different sampling window numbers, and is used to quantify the grid connection synchronization health and criteria performance in each operating cycle. Among them, the signal coordination stability value of sampling window 1 is 0.90, the sliding window signal mutual information value is 0.93, the mains active power is 0.65, the generator active power is 0.62, and the redundancy synchronization criterion value is 1.12. The key parameters of the window are balanced, and the grid connection status is at a relatively good level. The signal coordination stability value of sampling window 3 is 0.96, the sliding window signal mutual information value is 0.99, the mains active power is 0.72, the generator active power is 0.73, and the redundancy synchronization criterion value is 1.24. The signal health and power matching reach an ideal state within the window period. Further observation shows that the signal coordination stability value of sampling window 5 is 0.97, the sliding window signal mutual information value is 1.02, the mains active power is 0.75, and the generator active power is 0.76, which is the highest level in the window. The corresponding redundancy synchronization criterion value also rises to 1.30, and the total criterion value is 0.98, which represents the arithmetic average level of the redundancy synchronization criterion values ​​of each sampling window. It can be used as a direct basis for evaluating the overall synchronization health of the grid-connected system within the current period.

[0051] Table 1. Data Table of Wireless Synchronous Grid Connection Criteria

[0052]

[0053] like Figure 4 The figure shows a line graph comparing wireless synchronization parameters. Combined with Table 1, it can be seen that the signal coordination stability value, sliding window signal mutual information value, and redundancy synchronization criterion value differ significantly across different operating cycles under each sampling window. Specifically, sampling window 5 has the highest redundancy synchronization criterion value of 1.30, indicating excellent parameter performance and optimal synchronization health within this window. Sampling window 4 has the lowest redundancy synchronization criterion value of only 0.68, indicating a weaker grid connection status and potential synchronization risk in this interval. The normalized mean of the overall criterion is 0.98, which is high, reflecting good synchronization capability under overall operating conditions. Overall, the parameter comparison line graph visually demonstrates the changing trends of the main variables under different window operating conditions and their impact on the redundancy synchronization criterion value, providing data support for optimizing grid connection control strategies and providing early warning of abnormal windows.

[0054] In this implementation plan, a comprehensive evaluation system for synchronization reliability is constructed through cross-link information fusion and multi-channel criterion optimization for primary and backup emergency responses. This step deeply integrates multi-dimensional parameters such as electrical signal quality, power matching degree, communication link status, and equipment operating environment. Through a designed formula for calculating redundant synchronization criterion values, a quantitative assessment of the overall synchronization status is achieved. This assessment can accurately identify potential synchronization risks caused by communication quality fluctuations and changes in environmental factors, providing synchronization reliability criteria based on multi-source information fusion. This step effectively overcomes the limitations of evaluating single electrical parameters, establishes a comprehensive evaluation index that includes communication link quality and equipment operating status, provides a second line of defense, and significantly improves the accuracy and reliability of synchronization judgment.

[0055] Specifically, the process of constructing a self-healing decision link collaboration and entering the adjustment approximation criterion and closing control module is as follows: Figure 5 The diagram shows the redundancy synchronization criterion decision-making flowchart. It compares the redundancy synchronization criterion value and the redundancy synchronization criterion threshold in real time. The redundancy synchronization criterion threshold includes a primary criterion threshold and a secondary criterion threshold. The primary criterion threshold is used to distinguish between a healthy link state and a state requiring optimization, while the secondary criterion threshold is used to define acceptable performance boundaries and fault states requiring urgent intervention.

[0056] When the redundancy synchronization criterion value is less than the secondary criterion threshold, the health and synchronization criterion judgments are mismatched, the closing control authority is locked, the risk link is disconnected, and derated, isolated and soft disconnection are performed. The system switches to the human-machine interactive visual assisted grid connection criterion mode. For the highest level of protection against severe link degradation, the system effectively prevents erroneous operations under unreliable communication conditions by stopping automatic operation, isolating faulty components and switching to manual supervision mode.

[0057] When the redundancy synchronization criterion value is greater than or equal to the secondary criterion threshold but less than the primary criterion threshold, the command communication increases the transmission power and reduces the data transmission rate until the redundancy synchronization criterion value is greater than or equal to the primary criterion threshold. When the link performance slightly degrades, the communication parameters are actively optimized to try to restore it to the healthy level required for operation.

[0058] When the redundancy synchronization criterion value is greater than or equal to the first-level criterion threshold, it indicates that the electrical signal quality and communication link status are both at an excellent level, supporting subsequent accurate closing judgment and execution based on the phase difference convergence trend. The main criterion link and sensing channel are healthy, and the system enters the adjustment approximation criterion and closing control module.

[0059] In this implementation scheme, by comparing the redundant synchronization criterion value with the preset two-level criterion threshold in real time, a hierarchical control strategy can be executed according to the health status of the communication link and its state: when the criterion value is severely abnormal, the highest level of safety protection is immediately activated to cut off the risk and switch to manual mode; when the criterion value is slightly abnormal, the communication parameters are optimized to attempt to restore the healthy state; when the criterion value is good, the next stage of precise closing judgment is smoothly entered. This hierarchical and adaptive processing method ensures that appropriate responses can be made when facing different degrees of communication quality fluctuations and link anomalies, which not only ensures the safety of the grid connection process, but also maximizes the maintenance of automated operation capabilities, and provides reliable link protection for subsequent precise closing control.

[0060] Specifically, based on in-situ electrical data, real-time link data, and historical criterion data, the process of integrating high-order convergence trend analysis is as follows: First, obtain the instantaneous phase angle difference at the center of the historical sliding window and the historical rate of change of the phase angle difference, providing an important reference benchmark for the dynamic trend analysis of the current phase difference. Second, align the phase angle sequences of the mains side and the generator side using synchronous sampling clocks, extract the phase angle difference at the center point of the sliding window to obtain the instantaneous phase angle difference at the center of the sliding window, eliminating the calculation error introduced by asynchronous sampling times, and selecting the window center value as the most representative phase difference state for that period. Third, analyze the instantaneous phase angle difference at the center of the sliding window through... The first-order difference algorithm approximating the sliding window obtains the rate of change of phase angle difference, quantifying the evolution trend of phase difference per unit time. Positive values ​​indicate phase difference expansion, while negative values ​​indicate convergence, making it a key dynamic indicator for predicting the closing window. The average frequency values ​​of the mains-side sliding window and the generator-side sliding window are obtained through the averaging algorithm, effectively smoothing instantaneous fluctuations and better reflecting the steady-state frequency levels of the power sources on both sides. The average effective values ​​of the mains-side sliding window voltage and the generator-side sliding window voltage are obtained through the averaging algorithm, which are used to evaluate the overall matching degree of voltage amplitude on both sides.

[0061] Taking the absolute value of the instantaneous phase angle difference at the center of the sliding window and multiplying it by the phase difference sensitivity coefficient yields the synchronous phase offset suppression term, which directly penalizes any non-zero phase difference. Its value is proportional to the absolute value of the phase difference, and the sensitivity coefficient determines the severity of the penalty. Taking the absolute value of the rate of change of the phase angle difference and multiplying it by the rate of change of the phase difference sensitivity coefficient yields the phase trend change suppression term, which suppresses unfavorable phase divergence trends. Even if the instantaneous phase difference is small, if it is rapidly increasing, it will be significantly suppressed by this term. The synchronous phase offset suppression term and the phase trend change suppression term are added together as the negative exponent of an exponential function, representing the signal cooperative stability value. Using the base term, multiply by a negative exponent to obtain the numerator for enhanced synchronization convergence. The larger the phase difference and its divergence trend, the more drastically the numerator decreases, significantly reducing the final approximation value. Calculate the absolute values ​​of the differences between the average frequencies of the mains and generator sides, and the absolute values ​​of the average effective values ​​of the sliding window voltages of the mains and generator sides. Add these two absolute values ​​together and then add one to obtain the denominator for the synchronization frequency and voltage deviation. This denominator increases with the increase of the voltage and frequency differences between the two sides, thus reducing the final approximation evaluation result. Divide the numerator for enhanced synchronization convergence by the denominator for the synchronization frequency and voltage deviation to obtain the synchronization adjustment approximation value. The specific calculation formula is as follows: ;

[0062] In the formula, This indicates the approximation value of the synchronous adjustment, which is the final quantitative basis for the self-judgment of the closing window; It represents the signal coordination stability value, reflecting the overall health and synchronization quality of the electrical signal; It represents the instantaneous phase angle difference at the center of the sliding window, characterizing the core phase deviation between the mains voltage and generator voltage waveforms at the current observation window center point; This represents the rate of change of the phase angle difference; dynamic parameters reveal the direction and rate of change of the phase difference. It represents the average sliding window frequency of the mains power supply, which indicates the steady-state frequency level of the mains power supply within the observation window and reflects the stability of the power grid operation. This represents the average frequency of the sliding window on the generator side, reflecting the average output frequency of the generator set within the observation window. This represents the mean effective value of the sliding window voltage on the mains side, quantifying the average voltage amplitude on the mains side within the observation window. This represents the average effective value of the sliding window voltage on the generator side, characterizing the average output voltage level on the generator side. The phase difference sensitivity coefficient is calculated by using the ratio of the standard deviation to the mean of the sliding window to the instantaneous phase angle difference at the center of the historical sliding window. The value ranges from 0.2 to 2.0. The sensitivity coefficient for the rate of change of phase difference is obtained by using the ratio of the standard deviation to the mean of the sliding window to the historical rate of change of phase angle difference. The value ranges from 0.1 to 1.5.

[0063] In this implementation scheme, a precise dynamic assessment of the synchronization state is achieved by integrating high-order convergence trend analysis. This step innovatively introduces the key dynamic indicator of phase angle difference change rate, based on the traditional static criteria of phase difference, frequency difference, and voltage difference. Through the designed formula for calculating the synchronization adjustment approximation value, the instantaneous phase deviation, phase change trend, frequency matching degree, and voltage matching degree are comprehensively considered. This assessment method can keenly capture dynamic convergence characteristics, not only reflecting the current synchronization state but also predicting the evolution trend of closing timing. This provides a precise quantitative basis containing dynamic change information for the intelligent judgment of closing window, significantly improving the accuracy of closing timing judgment and grid connection reliability.

[0064] Specifically, the process of self-judgment of the closing window and end-side closed-loop drive is as follows: real-time comparison of the synchronous adjustment approximation value and the synchronous adjustment approximation threshold. The synchronous adjustment approximation threshold includes a first-level approximation threshold and a second-level approximation threshold. The first-level approximation threshold is used to define the ideal closing window, while the second-level approximation threshold is used to distinguish between the adjustable state and the out-of-synchronization state that requires safety intervention.

[0065] When the synchronous adjustment approximation value is greater than the first-level approximation threshold, the drive generator set continuously optimizes speed and excitation control, and reduces the frequency difference in real time. By actively maintaining the generator frequency at a level slightly higher than the mains power by about 0.05Hz, continuous phase convergence conditions are created. The phase difference is monitored synchronously to dynamically determine the closing window. When the phase difference enters -8° and stably approaches 0°, the criterion is met, triggering the closing action. No long-distance transmission of decision-making is required, thus completely eliminating the impact of communication delay on the closing timing and achieving precise closed-loop drive.

[0066] When the synchronous adjustment approximation value is greater than the second-level approximation threshold and less than or equal to the first-level approximation threshold, the generator speed regulation and voltage regulation amplitude are increased. By increasing the control gain, the convergence process of frequency and voltage is accelerated. If the phase difference, frequency difference and trend indicators meet the standards for k consecutive approximation sliding windows, the closing action is triggered. By requiring the stable achievement of multiple consecutive windows, the risk of false closing caused by instantaneous achievement is effectively avoided.

[0067] When the synchronization adjustment approximation value is less than or equal to the secondary approximation threshold, the closing control authority is locked, the original electrical data and real-time link data are recorded to the grid connection anomaly database, and the human-machine interactive visual auxiliary grid connection judgment mode is switched. This is a safety fallback measure taken when the synchronization state deviates significantly from the expectation. By suspending the automatic control cycle and handing it over to the operator for judgment, the ultimate safety of the grid connection operation is ensured.

[0068] In this implementation scheme, intelligent judgment of the closing window and end-side closed-loop drive are achieved by comparing the synchronous adjustment approximation value in real time. This step executes differentiated control strategies according to different levels of synchronization: when the state is ideal, a preset frequency offset strategy is used to create a continuous phase convergence condition, and the real-time performance of local calculation is used to trigger closing in the optimal phase window, completely avoiding the impact of communication delay; when the state is acceptable but not optimal, a conservative strategy of increasing the adjustment force and adopting multi-window continuous criteria is used to ensure the reliability of closing; when the state is abnormal, the operation is stopped and switched to manual mode to ensure final safety. This hierarchical and intelligent processing method not only ensures the speed and accuracy of grid connection under ideal operating conditions, but also ensures safety through conservative strategies under non-ideal operating conditions, ultimately achieving closed-loop control that can reliably complete grid connection operations under complex operating conditions.

[0069] like Figure 2 As shown, the second aspect of the present invention provides a multimodal wireless synchronous grid-connected control system, including an acquisition and preprocessing module for acquiring in-situ electrical data and real-time link data, and obtaining historical criterion data; preprocessing the in-situ electrical data, real-time link data, and historical criterion data; a multimodal grid-connected and switching mechanism module for integrating circuit breaker type identification, operating condition perception, and communication link discrimination to construct a reliable intelligent decision-making system for grid-connected operations; a signal health discrimination and prevention module for performing distributed in-situ multimodal perception and heterogeneous signal sliding statistics on in-situ electrical data to construct a smart terminal-side time-series signal fusion and in-situ adaptive health perception system; a redundancy criterion and synchronous health judgment module for constructing a self-healing judgment link collaboration through cross-link information fusion and primary / backup emergency multi-channel criterion optimization, and entering the adjustment approximation criterion and closing control module; and an adjustment approximation criterion and closing control module for performing closing window self-judgment and end-side closed-loop drive based on in-situ electrical data, real-time link data, and historical criterion data, integrating high-order convergence trend analysis.

[0070] This implementation scheme effectively solves the core problem of excessive reliance on communication links in traditional solutions by deeply integrating distributed in-situ sensing with multi-layer intelligent criteria. By adopting in-situ computing on the terminal side and multi-modal collaborative criteria, it achieves localized and accurate sensing of electrical status and closing decisions, significantly reducing the impact of communication delay on grid connection accuracy. By constructing a multi-mode collaboration that includes wireless, human-machine interaction, and wired redundancy, it can intelligently adapt to different equipment conditions and operating conditions, achieving optimal grid connection results while ensuring operational safety. It innovatively introduces multi-dimensional evaluation indicators such as signal collaborative stability value and synchronous adjustment approximation value, forming a complete closed-loop control from signal quality assessment and link reliability verification to closing timing judgment. This layered and progressive design concept with multiple guarantees enables the system to maintain a high degree of automation while possessing excellent fault tolerance and operating condition adaptability, providing a highly reliable and high-precision complete solution for synchronous grid connection operations.

[0071] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus.

[0072] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to the specific implementations described. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A multimodal wireless synchronous grid-connected control method, characterized in that, Includes the following steps: S1, collect in-situ electrical data and real-time link data, and obtain historical criterion data; preprocess the in-situ electrical data, real-time link data, and historical criterion data; S2 integrates circuit breaker type identification, operating condition perception and communication link discrimination to build a reliable grid connection operation intelligent decision-making system. S3 performs distributed in-situ multimodal sensing and heterogeneous signal sliding statistics on in-situ electrical data to construct a smart terminal-side time-series signal fusion and in-situ adaptive health sensing system. S4, through cross-link information fusion and primary / backup emergency multi-channel criterion optimization, constructs a self-healing decision link collaboration, and enters the adjustment proximity criterion and closing control module; The specific process of constructing a self-healing decision link collaboration and entering the adjustment approximation criterion and closing control module is as follows: Real-time comparison of redundant synchronization criterion values ​​and redundant synchronization criterion thresholds. The redundant synchronization criterion thresholds include primary criterion thresholds and secondary criterion thresholds. When the redundant synchronization criterion value is less than the secondary criterion threshold, the health status and synchronization criterion judgments are mismatched, the closing control authority is locked, the risk link is disconnected, and the rate reduction, isolation and soft disconnection are carried out. The system is then switched to the human-machine interactive visual auxiliary grid connection criterion mode. When the redundancy synchronization criterion value is greater than or equal to the secondary criterion threshold and less than the primary criterion threshold, the command communication increases the transmission power and decreases the data transmission rate until the redundancy synchronization criterion value is greater than or equal to the primary criterion threshold. When the redundant synchronization criterion value is greater than or equal to the first-level criterion threshold, the main criterion link and sensing channel are healthy, and the process enters the adjustment of the approximation criterion and the closing control module. S5, based on in-situ electrical data, real-time link data and historical criterion data, integrates high-order convergence trend analysis to perform self-judgment of closing window and end-side closed-loop drive.

2. The multimodal wireless synchronous grid-connected control method according to claim 1, characterized in that: The in-situ electrical data and real-time link data are collected to obtain historical judgment data; The specific process for preprocessing in-situ electrical data, real-time link data, and historical criterion data is as follows: Collect in-situ electrical data, including: AC voltage signal of mains power, AC voltage signal of generator, AC current signal of mains power, and AC current signal of generator. Collect real-time link data, including: round-trip time of heartbeat packets in the mains link, round-trip time of heartbeat packets in the generator link, temperature of mains equipment, temperature of generator equipment, signal-to-noise ratio of mains link and signal-to-noise ratio of generator link; Acquire historical criteria data and establish a historical criteria database. The historical criteria data includes: historical mains communication link delay sequence, historical power generation communication link delay sequence, historical sliding window center instantaneous phase angle difference, historical phase angle difference change rate, health sliding window, redundancy sliding window and approximation sliding window. The real-time link data is repaired for missing values ​​and outliers is identified by interpolation completion and extreme value removal algorithms; historical criterion data is synchronized and windowed by multi-step sampling alignment and time-series window mapping; and in-situ electrical data, real-time link data and historical criterion data are standardized and normalized by distribution standardization and linear normalization algorithms.

3. The multimodal wireless synchronous grid-connected control method according to claim 1, characterized in that: The specific process of integrating circuit breaker type identification, operating condition perception, and communication link discrimination to construct a reliable grid connection operation intelligent decision-making system is as follows: It supports three working modes: adaptive wireless intelligent grid connection, human-machine interaction visual assisted grid connection criterion, and wired redundancy protection for extreme scenarios. In the adaptive wireless grid connection mode, the status of the electric circuit breaker is monitored in real time, and the adjustment of the proximity criterion and closing control are completed. The closing command is sent in real time through the wireless network to realize the whole process of unattended grid connection. The human-machine interaction visual assisted grid connection criterion mode is suitable for non-electric circuit breakers and extremely simple wiring. It performs synchronous adjustment of the proximity value criterion calculation in real time and displays the closing window on the terminal interface. The operator manually completes the circuit breaker closing according to the parameter instructions. The extreme scenario wired redundancy protection mode is used for wireless communication anomalies and extreme scenarios, switching to wired control and completing the safe closed loop of closing action through physical wired links.

4. The multimodal wireless synchronous grid-connected control method according to claim 1, characterized in that: The specific process of performing distributed in-situ multimodal sensing and heterogeneous signal sliding statistics on in-situ electrical data is as follows: The system acquires AC mains voltage and AC generator voltage signals. It then uses a Discrete Fourier Transform (DFT) algorithm to extract the fundamental effective value of the AC mains voltage and AC generator voltage signals, obtaining the AC mains voltage sequence and AC generator voltage sequence within the health sliding window length. Finally, it uses zero-crossing detection and linear interpolation to obtain the AC mains voltage and AC generator voltage sequences within the health sliding window length. The system also uses a Fast Hilbert Transform and synchronous sampling clock alignment to obtain the AC mains voltage and AC generator voltage sequences within the health sliding window length. Finally, it uses a k-nearest neighbor mutual information estimation algorithm to obtain the mutual information values ​​of the sliding window signals. Finally, it uses a health sliding window variance statistical algorithm to obtain the AC mains voltage sliding window variance and the AC generator voltage sliding window variance. The square root of the sum of squares of the differences between the mains-side frequency sequence and the generator-side frequency sequence is calculated and added to 1 to obtain the frequency offset adjustment term. The frequency offset adjustment term is multiplied by the mutual information value of the sliding window signal to obtain the numerator of the cooperative enhancement term. The root mean square difference term between the mains-side phase angle sequence and the generator-side phase angle sequence is calculated, and the root root of the mains-side voltage sliding window variance and the generator-side voltage sliding window variance is superimposed to obtain the voltage variance composite term. The root mean square difference term and the voltage variance composite term are multiplied and added to 1 to obtain the denominator correction term. Finally, the cooperative enhancement numerator is divided by the denominator correction term to obtain the signal cooperative stability value.

5. The multimodal wireless synchronous grid-connected control method according to claim 1, characterized in that: The specific process for constructing the intelligent terminal-side time-series signal fusion and in-situ adaptive health sensing system is as follows: Real-time comparison of signal co-stability values ​​and signal co-stability thresholds, including primary stability thresholds and secondary stability thresholds: When the signal coordination stability value is less than the secondary stability threshold, the local relay is triggered to trip through the intelligent grid-connected operation and maintenance terminal, the remote and local closing control permissions are locked simultaneously, a grid-connected anomaly database is created and the original electrical data and real-time link data are recorded to the grid-connected anomaly database, and the system switches to the extreme scenario wired redundancy protection mode. When the signal co-stabilization value is greater than or equal to the secondary stabilization threshold and less than the primary stabilization threshold, the local sampling frequency is increased, the digital filtering and wavelet denoising units are activated, the current sampled signal is filtered, and the system switches to the human-machine interactive visual-assisted grid connection criterion mode. When the signal coordination stability value is greater than or equal to the first-level stability threshold, the main channel continues to participate in the synchronization criterion closed loop and grid connection decision, and the speed regulation, voltage regulation and closing closed loop continue normally.

6. The multimodal wireless synchronous grid-connected control method according to claim 1, characterized in that: The specific process of cross-link information fusion and primary / backup emergency multi-channel criterion selection is as follows: The system acquires AC voltage and current signals from the mains power grid, AC voltage and current signals from the generator, round-trip time of the mains power link heartbeat, round-trip time of the generator's heartbeat, historical mains communication link delay sequences, historical generator communication link delay sequences, mains equipment temperature, generator equipment temperature, mains link signal-to-noise ratio (SNR), and generator link SNR. It extracts the fundamental RMS values ​​from the AC voltage, current, voltage, and current signals using the DFT algorithm, and then obtains the average active power of the mains power grid and the average active power of the generator through the fundamental active power calculation method and redundant sliding window mean statistics. The mains communication link delay is obtained by real-time packet transmission and reception acquisition of the mains link heartbeat round-trip time and the generator link heartbeat round-trip time. The product of the cooperative stability value of the signal and the mutual information value of the sliding window signal is calculated to obtain the criterion cooperative term. Calculate the absolute value of the difference between the average active power of the mains and the average active power of the generator, add one, take the natural logarithm, and add one again to obtain the active power deviation term; calculate the absolute value of the difference between the mains communication link delay and the generator communication link delay, multiply by the link delay sensitivity coefficient, take the opposite number as an exponential function, and obtain the link delay attenuation term; multiply the criterion coordination term, the active power deviation term, and the link delay attenuation term together to obtain the link compensation coordination term; calculate the square root of the absolute value of the difference between the temperature of the mains equipment and the temperature of the generator equipment, multiply by the sum of the reciprocals of the mains link signal-to-noise ratio and the generator link signal-to-noise ratio, multiply by the temperature deviation term and the signal-to-noise ratio attenuation term, and add one to obtain the environmental impact term; divide the link compensation coordination term by the environmental impact term to obtain the redundancy synchronization criterion value.

7. The multimodal wireless synchronous grid-connected control method according to claim 1, characterized in that: The specific process of fusing high-order convergence trend analysis based on in-situ electrical data, real-time link data, and historical criterion data is as follows: The process involves obtaining the instantaneous phase angle difference at the center of the historical sliding window and the rate of change of the historical phase angle difference; aligning the phase angle sequences of the mains side and the generator side using a synchronous sampling clock, and extracting the phase angle difference at the center of the sliding window to obtain the instantaneous phase angle difference at the center of the sliding window; obtaining the rate of change of the phase angle difference at the center of the sliding window using a first-order difference algorithm approximating the sliding window; obtaining the mean frequency of the mains side and the mean frequency of the generator side using an averaging algorithm for the mains side and the generator side; and obtaining the mean effective value of the mains side sliding window voltage and the mean effective value of the generator side sliding window voltage using an averaging algorithm for the mains voltage sequence and the generator voltage sequence. The absolute value of the instantaneous phase angle difference at the center of the sliding window is multiplied by the phase difference sensitivity coefficient to obtain the synchronization phase offset suppression term; the absolute value of the rate of change of the phase angle difference is multiplied by the rate of change of the phase difference sensitivity coefficient to obtain the phase trend change suppression term; the synchronization phase offset suppression term and the phase trend change suppression term are added together as the negative exponent of the exponential function, and multiplied by the negative exponent with the signal cooperative stability value as the base to obtain the synchronization convergence enhancement numerator; the absolute value of the difference between the average frequency values ​​of the mains side and the generator side, and the absolute value of the difference between the average effective values ​​of the sliding window voltages of the mains side and the generator side are calculated, and the two absolute values ​​are added together and one is added to obtain the denominator of the synchronization frequency and voltage deviation; the synchronization convergence enhancement numerator is divided by the denominator of the synchronization frequency and voltage deviation to obtain the synchronization adjustment approximation value.

8. The multimodal wireless synchronous grid-connected control method according to claim 1, characterized in that: The specific process of performing self-judgment of the closing window and end-side closed-loop drive is as follows: Real-time comparison and synchronous adjustment of the approximation value and the synchronous adjustment of the approximation threshold, which includes a first-level approximation threshold and a second-level approximation threshold: When the synchronous adjustment approximation value is greater than the first-level approximation threshold, the drive generator set continuously optimizes speed and excitation control, reduces frequency difference in real time, synchronously monitors phase difference to dynamically determine the closing window, and triggers closing action after the criteria are met. When the synchronous adjustment approximation value is greater than the secondary approximation threshold and less than or equal to the primary approximation threshold, the generator speed regulation and voltage regulation amplitude are increased. If the phase difference, frequency difference and trend indicators meet the standards for k consecutive approximation sliding windows, the closing action is triggered. When the synchronous adjustment approximation value is less than or equal to the secondary approximation threshold, the closing control authority is locked, the original electrical data and real-time link data are recorded to the grid connection anomaly database, and the system is switched to the human-machine interactive visual auxiliary grid connection criterion mode.

9. A multimodal wireless synchronous grid-connected control system, employing a multimodal wireless synchronous grid-connected control method as described in any one of claims 1-8, comprising: The data acquisition and preprocessing module is used to acquire in-situ electrical data and real-time link data, and obtain historical criterion data. Preprocess the in-situ electrical data, real-time link data, and historical criterion data; The multimodal grid connection and switching mechanism module is used to integrate circuit breaker type identification, operating condition perception and communication link discrimination to build a reliable grid connection operation intelligent decision-making system. The signal health judgment and control module is used to perform distributed in-situ multimodal sensing and heterogeneous signal sliding statistics on in-situ electrical data, and to build a smart terminal side time-series signal fusion and in-situ adaptive health sensing system. The redundancy criterion and synchronous health decision module is used to construct a self-healing decision link collaboration through cross-link information fusion and primary / backup emergency multi-channel criterion optimization, and then enter the adjustment proximity criterion and closing control module. The approximation criterion and closing control module is used to perform self-judgment of the closing window and end-side closed-loop drive based on in-situ electrical data, real-time link data and historical criterion data, and by integrating high-order convergence trend analysis.

Images