An entropy-increasing disturbance-based micro-grid load distribution control method, device and equipment

Abstract

The application discloses a micro-grid load distribution control method, device and equipment based on entropy increase disturbance, which analyzes the micro-grid operation state through chaos measurement, constructs an entropy increase disturbance incentive mechanism to actively stimulate system vitality, extracts inherent frequency characteristics of each power supply after random disturbance is applied, establishes a frequency synchronization coordination mechanism to realize multi-source collaboration, actively identifies distribution scheme loopholes by adopting counterexample search, generates a reinforced distribution strategy to improve robustness, evaluates load bearing capacity of each power supply, constructs a load conduction topology and analyzes stress distribution, acquires each power supply bearing limit parameter through boundary active detection, and fully excavates equipment potential, generates a final control instruction based on a multi-layer time sequence synchronization mechanism, and realizes multi-power supply coordinated output. The chaos theory is introduced into the micro-grid control, local optimum is avoided through active disturbance, a multi-layer coordination mechanism is adopted to adapt to new energy volatility, and an adaptive load distribution control scheme is provided for the micro-grid with high proportion of renewable energy access.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of micro-grid control, in particular to a micro-grid load distribution control method, device and equipment based on entropy-increasing disturbance. BACKGROUND

[0002] As an effective form of distributed energy access to the power grid, micro-grid plays an important role in improving energy utilization efficiency and enhancing power supply reliability. With the increasing penetration of renewable energy such as photovoltaic and wind power, the power balance and load distribution within the micro-grid face new challenges.

[0003] Current micro-grid load distribution technology mainly relies on pre-set optimization algorithms and fixed distribution strategies. This approach is not effective in the face of the randomness and volatility of new energy output. Each distributed power source is often controlled independently, lacking effective coordination mechanisms, resulting in frequent overloading of some power sources and idling of others. At the same time, existing methods lack an understanding of system operating boundaries, and conservative operating strategies limit the potential of equipment. The lack of predictive analysis of potential problems during load distribution often results in passive adjustments after faults occur. In addition, the dynamic coupling characteristics between multiple power sources have not been fully utilized, and the inherent oscillation characteristics and interaction mechanisms of each power source have not been thoroughly explored.

[0004] In the face of the complexity and uncertainty of micro-grid operating environments, traditional deterministic control methods have been unable to meet the needs of system optimization and operation. New control concepts and methods are urgently needed to achieve adaptive optimization and coordinated control of micro-grid load distribution. SUMMARY

[0005] The present application provides a micro-grid load distribution control method, device and equipment based on entropy-increasing disturbance, which comprehensively uses chaos theory to stimulate system vitality, frequency coupling to achieve multi-source collaboration, counterexample search to discover potential defects, active detection to explore equipment potential, and time sequence synchronization to ensure coordinated operation, etc. A complete technical system is constructed from system state perception, collaborative mechanism establishment, strategy optimization and improvement to coordinated control execution, enabling the micro-grid to have intelligent capabilities such as autonomously exploring optimal operating points, dynamically adapting to environmental changes, and coordinating the complementary operation of multiple power sources.

[0006] The present application provides a micro-grid load distribution control method, device and equipment based on entropy-increasing disturbance, which comprehensively uses chaos theory to stimulate system vitality, frequency coupling to achieve multi-source collaboration, counterexample search to discover potential defects, active detection to explore equipment potential, and time sequence synchronization to ensure coordinated operation, etc. A complete technical system is constructed from system state perception, collaborative mechanism establishment, strategy optimization and improvement to coordinated control execution, enabling the micro-grid to have intelligent capabilities such as autonomously exploring optimal operating points, dynamically adapting to environmental changes, and coordinating the complementary operation of multiple power sources.

[0007] Real-time output data and load demand data of power sources within the micro-grid are obtained, and a chaos metric analysis is performed on the real-time output data to generate system entropy value parameters. An entropy-increasing disturbance excitation mechanism is constructed based on the system entropy value parameters.

[0008] generate chaotic state operation data from the chaotic state operation data, and extract inherent frequency characteristics of each power source from the chaotic state operation data to obtain a resonance coupling matrix, and generate a frequency synchronization coordination mechanism between power sources by using the resonance coupling matrix;

[0009] generate a preliminary load distribution scheme based on the frequency synchronization coordination mechanism, generate distribution vulnerability identification data by performing counterexample search on the preliminary load distribution scheme, and obtain a reinforcement distribution strategy according to the distribution vulnerability identification data;

[0010] obtain load bearing capacity evaluation data of each power source based on the reinforcement distribution strategy, perform matching analysis on the load demand data and the load bearing capacity evaluation data to generate a load conduction topology, and obtain pressure distribution data between each power source by using the load conduction topology;

[0011] generate a set of boundary active detection instructions based on the pressure distribution data, drive each power source to gradually approach the operating boundary by using the set of boundary active detection instructions to obtain boundary response data, and generate load bearing limit parameters of each power source by performing extreme value analysis on the boundary response data;

[0012] set a unified timing adjustment reference based on the load bearing limit parameters, synchronize the output adjustment process of each power source by using the timing adjustment reference to obtain coordinated timing data, and generate timing synchronization control parameters according to the coordinated timing data;

[0013] generate a set of final load distribution control instructions based on the timing synchronization control parameters, and each power source executes coordinated output according to the control instruction set to obtain actual operation state data, and completes the micro-grid load distribution control.

[0014] The second aspect of the present application proposes a micro-grid load distribution control device based on entropy increasing disturbance, comprising:

[0015] A data acquisition module is configured to acquire real-time output data and load demand data of power sources in a micro-grid, perform chaotic degree measurement analysis on the real-time output data to generate system entropy value parameters, and construct an entropy increasing disturbance excitation mechanism based on the system entropy value parameters.

[0016] A chaotic coordination module is configured to generate chaotic state operation data from the chaotic state operation data by randomizing output disturbance of each power source based on the entropy increasing disturbance excitation mechanism, extract inherent frequency characteristics of each power source from the chaotic state operation data to obtain a resonance coupling matrix, and generate a frequency synchronization coordination mechanism between power sources by using the resonance coupling matrix.

[0017] A strategy optimization module is configured to generate a preliminary load distribution scheme based on the frequency synchronization coordination mechanism, perform counterexample search on the preliminary load distribution scheme to generate distribution vulnerability identification data, and obtain a reinforced distribution strategy based on the distribution vulnerability identification data.

[0018] A pressure analysis module is configured to obtain load-bearing capacity evaluation data of each power supply based on the reinforced distribution strategy, perform matching analysis on the load demand data and the load-bearing capacity evaluation data to generate a load conduction topology, and obtain pressure distribution data among the power supplies by using the load conduction topology.

[0019] A boundary detection module is configured to generate a set of boundary active detection instructions based on the pressure distribution data, drive each power supply to gradually approach an operating boundary by using the set of boundary active detection instructions to obtain boundary response data, and perform extreme value analysis on the boundary response data to generate a load-bearing limit parameter of each power supply.

[0020] A timing control module is configured to set a unified timing adjustment reference based on the load-bearing limit parameter, perform synchronization processing on an output adjustment process of each power supply by using the timing adjustment reference to obtain coordination timing data, and generate timing synchronization control parameters based on the coordination timing data.

[0021] An instruction execution module is configured to generate a final load distribution control instruction set based on the timing synchronization control parameters, and each power supply performs coordinated output by using the control instruction set to obtain actual operating state data, thereby completing the micro-grid load distribution control.

[0022] A third aspect of the present application provides a computer device, which comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the micro-grid load distribution control method based on entropy increase disturbance disclosed in the first aspect when executing the program.

[0023] The beneficial effects of the present application are embodied in the following aspects: first, the chaos theory is applied to micro-grid control, the chaotic state of the system is actively excited through the entropy increasing disturbance incentive mechanism, the inherent frequency characteristics and coupling relationship of each power supply are extracted from the chaotic response, an analysis approach different from the traditional model is provided, hidden system characteristics and optimization potential can be found, and the internal collaboration of multi-power supply is realized through the frequency synchronization coordination mechanism, thereby enhancing the coordinated operation capability of each power supply. Secondly, the counterexample search technology is used to comprehensively identify the loopholes of the preliminary allocation scheme, the potential problems are actively exposed by constructing the counter mode and conflict scene, and the traditional passive response to faults is changed. Combined with the boundary active detection technology, the operation limit of each power supply can be safely explored, and the equipment potential can be fully tapped. Such active optimization strategy improves the operation efficiency and reliability of the system. Finally, a time sequence synchronization control mechanism based on the bearing limit is established, the coordinated action of multi-power supply is realized through the hierarchical time scale and phase difference configuration, and the system impact caused by simultaneous adjustment is avoided. The whole control process is from disorder to order, from independence to collaboration, which improves the intelligent level of load distribution, and provides a new solution for the stable operation of micro-grid under the condition of high proportion of renewable energy access.

[0024] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present application. BRIEF DESCRIPTION OF DRAWINGS

[0025] The drawings herein show specific examples of the technical solutions described in the present application, and constitute part of the specification together with the specific embodiments, for explaining the technical solutions, principles and effects of the present application.

[0026] Unless specifically stated or defined otherwise, the same reference signs in different drawings represent the same or similar technical features, and different reference signs may also be used to represent the same or similar technical features.

[0027] Figure 1 is a flowchart of a micro-grid load distribution control method based on entropy increasing disturbance according to the present application.

[0028] Figure 2 is a structural block diagram of a micro-grid load distribution control device based on entropy increasing disturbance according to the present application.

[0029] Figure 3 is a structural diagram of a computer device according to the present application. DETAILED DESCRIPTION

[0030] In the following description, for purposes of explanation and not limitation, specific details are set forth such as particular architectures, technologies, techniques, etc. in order to provide a thorough understanding of the embodiments of the present application. However, it will be apparent to those skilled in the art that the present application can be practiced in other embodiments that depart from these specific details. In other instances, detailed descriptions of well-known systems, devices, circuits, and methods are omitted so as not to obscure the description of the present application with unnecessary detail.

[0031] It is to be understood that the terminology "includes", "has", "holds", "contains" and / or "comprising", "including", "having" and their conjugates, as used herein, means "including but not limited to", and not to the exclusion of any other term or aspect.

[0032] It is also to be understood that the terminology "and / or" as used herein refers to and encompasses any and all possible combinations of one or more of the associated listed items, as well as the lack of any associated listed items.

[0033] As used in the description of the application and the appended claims, the term "if' can be interpreted to mean "when" or "upon" or "in response to determining" or "in response to detecting" depending on the context. Similarly, the phrase "if it is determined" or "if [a described condition or event] is detected" can be interpreted to mean "upon determining" or "in response to determining" or "upon [the described condition or event] being detected" or "in response to [the described condition or event] being detected", depending on the context.

[0034] In addition, the terms "first", "second", "third", etc. as used in the description of embodiments herein and throughout the claims, are used for distinguishing between similar elements and do not necessarily have an actual reference to specific, quantitative values. Numerical descriptions should be considered in a flexible manner.

[0035] Reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present application. Thus, the appearances of the phrases "in one embodiment" or "in an embodiment" or "in a various embodiment" or "in at least one embodiment" in various places throughout this specification are not necessarily referring to the same embodiment, unless otherwise specified. Furthermore, the terms "comprises", "comprising", "includes", "including", "has", "having", and the like, as used herein, are specifically intended to be construed as "including but not limited to".

[0036] The technical solutions of the embodiments of the present application are described as follows.

[0037] As Figure 1As shown, the embodiment of the present application provides a micro-grid load distribution control method based on entropy increase disturbance, which comprises the following steps S110-S170:

[0038] In step S110, real-time output data of power sources in the micro-grid and load demand data are acquired, chaotic metric analysis is performed on the real-time output data to generate system entropy value parameters, and an entropy increase disturbance incentive mechanism is constructed based on the system entropy value parameters.

[0039] Specifically, the real-time output data of the power sources in the micro-grid is acquired by using a distributed data acquisition architecture, and high-precision power measurement sensors are deployed at the output ends of photovoltaic power generation units, wind turbine generators, energy storage systems, diesel generators and other types of distributed power sources to realize real-time monitoring. The power measurement sensors use a combined measurement method of current transformers and voltage transformers based on the Hall effect principle, with a current measurement accuracy of 0.2 level and a voltage measurement accuracy of 0.1 level, and a sampling frequency of 1 kHz, ensuring the capture of the instantaneous fluctuation characteristics of the power output. The output data acquisition of the photovoltaic power generation unit is combined with the irradiance sensor and the temperature sensor to monitor the influence of the change of light intensity on the power generation power in real time. The irradiance measurement range is 0-1500 W / m², the measurement accuracy is ±2%, the temperature measurement range is -40℃ to 85℃, and the accuracy is ±0.5℃. The output data acquisition of the wind turbine generator is matched with the wind speed and direction sensor to monitor the corresponding relationship between the wind speed change and the power generation power. The wind speed measurement range is 0-60 m / s, the measurement accuracy is ±0.2 m / s, and the wind direction measurement accuracy is ±3°. The output data of the energy storage system includes multi-dimensional parameters such as charging and discharging power, state of charge SOC, battery temperature, etc., and the SOC estimation accuracy is controlled within ±2%. Data transmission uses IEC61850 communication protocol, and a high-speed data transmission network is built through optical fiber Ethernet, with a transmission delay controlled within 10 milliseconds. The load demand data is acquired through smart meters and load monitoring terminals, covering different types of electricity demand such as residential load, commercial load and industrial load. The sampling interval of the smart meter is set to 15 minutes, and the sampling interval of the key load node is shortened to 1 minute. The load prediction module uses the LSTM long short-term memory neural network algorithm to perform short-term load prediction based on historical load data, weather forecast information, holiday factors and other multi-inputs. The prediction time scale includes 15 minutes, 1 hour and 24 hours. High-frequency sampling is used on the power side to capture the output fluctuation characteristics and chaos analysis, and low-frequency sampling combined with the prediction algorithm can meet the load distribution demand on the load side. The time scale matching is realized through the prediction model. Through the distributed data acquisition architecture and the intelligent prediction algorithm, the power output data and the load demand data reflecting the current operating state of the micro-grid are acquired in real time.

[0040] The real-time output data is analyzed by chaos measurement to generate system entropy parameters. The chaos measurement analysis adopts a nonlinear dynamics theory framework, and considers the output fluctuation of the microgrid system as a chaotic behavior of a complex dynamic system. The analysis process first reconstructs the phase space of the collected multi-source output data, and uses the delay coordinate embedding method to expand the one-dimensional time series to a high-dimensional phase space. The delay time τ is determined by the minimum value of the mutual information function, and the embedding dimension m is calculated by the false nearest neighbor point method. In a typical photovoltaic-wind power-storage microgrid system, the optimal delay time τ of the photovoltaic output data is 3 sampling periods, and the embedding dimension m is 5; the delay time τ of the wind power output data is 5 sampling periods, and the embedding dimension m is 7, which reflects the higher chaotic characteristics of the wind power output. The Wolf algorithm is used to calculate the Lyapunov exponent, which quantifies the chaotic degree of the system by tracking the separation rate of adjacent orbits in the phase space. A positive Lyapunov exponent indicates that the system has chaotic behavior, and a larger exponent indicates that the system is more sensitive to initial conditions. The G-P algorithm is used to calculate the correlation dimension, which reflects the complexity of the system by analyzing the distribution density of points in the phase space. The fractional value of the correlation dimension is an important indicator of chaotic systems. The Kolmogorov entropy calculation quantifies the information generation rate of the system and reflects the unpredictability of the system motion. Under the typical working condition of alternating sunny and cloudy days, the Lyapunov exponent of the photovoltaic system reaches 0.15, showing obvious chaotic characteristics. Through the nonlinear dynamics analysis method, the chaotic characteristics hidden in the microgrid output data are deeply excavated.

[0041] In some embodiments, the chaos measurement analysis of the real-time output data generates system entropy parameters, including: performing disorder measurement analysis on the real-time output data to obtain a chaotic feature vector; performing entropy value integral processing based on the chaotic feature vector to generate initial entropy value data; and performing time series standardization processing on the initial entropy value data to obtain system entropy parameters.

[0042] Firstly, the chaotic feature vector is obtained by the disorder quantification analysis of real-time output data. The multi-scale entropy analysis method is used for disorder quantification analysis. The original output time series is coarse-grained at different time scales, and the sample entropy value at each scale is calculated. The sample entropy calculation uses the template matching algorithm, and the template length m = 2 and the similar tolerance r = 0.2 times the standard deviation are set. The regularity of the sequence is quantified by calculating the conditional probability of similar templates. For photovoltaic output data, the sample entropy is calculated at 1 minute, 5 minutes, 15 minutes and 60 minutes, respectively, to obtain the multi-scale entropy spectrum. The permutation entropy analysis converts continuous data into ordinal patterns by symbolization processing, and calculates the probability distribution entropy value of different permutation patterns. The embedding dimension is set to 3-7, and the delay time is set to 1-5. The approximate entropy calculation uses the sliding window technique, and the window length is set to 100 data points, and the sliding step is 10 data points. The dynamic change of the disorder degree of the system is tracked in real time. The fuzzy entropy introduces a fuzzy membership function to replace the hard threshold judgment, and uses a Gaussian membership function. The fuzzy factor is set to 0.1-0.3 for adaptive adjustment. During the period of intense fluctuation of wind power output, the permutation entropy value can reach 0.95 or more, indicating that the system is in a highly disordered state. The fluctuation complexity analysis uses the Lempel-Ziv algorithm to reflect the complexity of the sequence by calculating its compressibility. The Lyapunov exponent, correlation dimension, Kolmogorov entropy, multi-scale entropy, permutation entropy, approximate entropy, fuzzy entropy and LZ complexity are normalized to form the chaotic feature vector.

[0043] Then, the entropy value integration processing uses a weighted fusion algorithm to assign weight coefficients according to the contribution of each chaotic index to the system disorder. The weight determination uses the entropy weight method combined with the analytic hierarchy process. First, the information entropy of each index is calculated to determine the objective weight, and then the subjective weight is determined by constructing a judgment matrix combined with expert experience. Finally, the comprehensive weight is obtained by linear combination. In a typical microgrid system, the Lyapunov exponent weight is 0.25, the multi-scale entropy weight is 0.20, the permutation entropy weight is 0.15, and the weights of the remaining indexes are distributed between 0.05-0.10. The integral calculation uses the trapezoidal integration method to numerically integrate the discrete chaotic characteristic values on the time axis, and the integration time window is set to 15 minutes, consistent with the dispatching period of the power system. To avoid the influence of single-time abnormal values, the integral results are smoothed by using a sliding average filter, and the filter window length is 5 data points. The dynamic weight adjustment mechanism adjusts the weights of each index in real time according to the system operating conditions. When the output of new energy is high, the weights of Lyapunov exponent and permutation entropy are increased, and when the load fluctuation is intense, the weight of multi-scale entropy is increased. The cumulative entropy value calculation considers the influence of historical entropy values, and uses an exponential decay factor to weight the historical data. The decay factor a = 0.95 ensures that recent data have a dominant influence on the current entropy value. Through weighted integration, the multi-dimensional chaotic feature vector is fused into a scalar initial entropy value that reflects the overall disorder degree of the system.

[0044] Finally, the initial entropy value data is processed by time series standardization to obtain the system entropy value parameter. The standardization process first detects and corrects outliers, using the 3σ criterion to identify outliers beyond the mean ± 3 times the standard deviation, and correcting them by linear interpolation or spline interpolation. Trend decomposition uses the STL time series decomposition algorithm to decompose the initial entropy value data into trend components, seasonal components and random components, separating out the steady-state entropy value reflecting the long-term evolution trend of the system and the dynamic entropy value reflecting short-term fluctuations. Seasonal adjustment considers the daily and seasonal characteristics of the microgrid operation, and uses the X-12-ARIMA method to extract and eliminate seasonal factors. Normalization uses the maximum-minimum normalization method to map the entropy value to the [0, 1] interval, where 0 represents a completely ordered state and 1 represents the maximum chaotic state. In a typical operating day, the system entropy value can reach 0.7-0.8 in the early morning and evening due to rapid changes in photovoltaic output; the system entropy value drops to 0.2-0.3 at noon when the light is stable. Moving average processing uses the exponential weighted moving average (EWMA) method with a smoothing factor λ = 0.3 to smooth the entropy value while eliminating high-frequency noise. The standardized system entropy value parameter contains four dimensions: instantaneous entropy value, average entropy value, entropy value change rate and entropy value fluctuation amplitude, forming a complete description of the system chaotic state. Through time series standardization, the system entropy value parameter with comparability and stability is generated.

[0045] The entropy-increasing disturbance excitation mechanism is constructed based on the system entropy value parameter. The core idea of the entropy-increasing disturbance excitation mechanism is to improve the chaos degree of the system by actively introducing controlled disturbance, and to stimulate the system to explore a better operating state. The disturbance intensity design adopts an adaptive control strategy, which dynamically adjusts the disturbance amplitude according to the deviation between the current system entropy value and the target entropy value. When the system entropy value is less than 0.3, it indicates that the system may be trapped in a local optimum, and stronger disturbance needs to be applied to stimulate the system vitality; when the system entropy value is higher than 0.8, it indicates that the system is too chaotic, and the disturbance intensity needs to be reduced to maintain the stability of the system. The disturbance signal generation adopts a chaotic mapping method, including Logistic mapping (control parameter μ is adjusted in the interval of 3.57-4.0), Henon mapping (parameters a=1.4, b=0.3) and Lorenz system, etc. classic chaotic generator, the chaotic characteristics of the disturbance are controlled by adjusting the mapping parameters. The disturbance injection point selection is based on the sensitivity analysis results, and the control variables with significant influence on the system entropy value are preferentially selected, such as superimposing 5%-10% disturbance amplitude on the charge and discharge power of the energy storage system, shifting the switching time sequence of controllable loads by ±15 minutes, etc. The disturbance timing design adopts an intermittent excitation mode, and the disturbance duration is set to 5-10 minutes, and the interval time is 15-30 minutes, which can ensure the disturbance effect while avoiding system fatigue. The multi-point collaborative disturbance mechanism determines the coupling relationship between different disturbance points through correlation analysis, and designs orthogonal disturbance sequence to reduce mutual interference. The disturbance frequency is set in the range of 0.1-1 Hz according to the inertia characteristics of the system, which ensures that the disturbance signal can effectively propagate and affect the system state. In an industrial park microgrid application scenario, the system detects that the entropy value drops to 0.25 at 10 am, indicating that the photovoltaic and wind power output tends to be stable but the scheduling mode is solidified, at this time the entropy-increasing disturbance excitation mechanism is automatically triggered, and ±8% amplitude disturbance based on Logistic mapping is applied to the charge and discharge power of the energy storage system, and ±20 minute time shift disturbance generated by Henon mapping is introduced to the controllable load switching time sequence, the disturbance signal frequency is set to 0.5 Hz, after 8 minutes of continuous action, the system entropy value gradually rises to about 0.45, successfully stimulating the system to jump out of the original operating mode and explore a better energy scheduling scheme. Through adaptive adjustment of disturbance intensity, chaotic signal generation and multi-point collaborative injection, an entropy-increasing disturbance excitation mechanism based on system entropy value feedback is constructed.

[0046] In step S120, chaotic state operation data is generated by randomizing the output disturbance of each power supply based on the entropy-increasing disturbance excitation mechanism, the inherent frequency characteristics of each power supply are extracted from the chaotic state operation data to obtain a resonance coupling matrix, and a frequency synchronization and coordination mechanism between power supplies is generated using the resonance coupling matrix.

[0047] Specifically, by applying random disturbance based on the entropy-increasing mechanism to each power supply of the microgrid, the system is stimulated to operate in a chaotic state, the inherent oscillation characteristics of each power supply are identified from the chaotic response, and a resonance coupling relationship is constructed, ultimately forming a multi-power frequency synchronization and coordination control strategy.

[0048] In some embodiments, the randomization of the output of each power supply based on the entropy-increasing disturbance excitation mechanism to generate chaotic state operation data includes: generating a set of random disturbance parameters based on the entropy-increasing disturbance excitation mechanism; inputting the set of random disturbance parameters to each power supply to obtain power supply-specific disturbance instructions; driving each power supply model to execute output disturbance to obtain disturbance response data; and integrating and processing the disturbance response data to generate chaotic state operation data.

[0049] First, the chaotic mapping module in the entropy-increasing disturbance excitation mechanism is used to generate a set of random disturbance parameters that meet different statistical characteristics through parameterized configuration. The Logistic mapping parameter μ is randomly selected between 3.75 and 4.0 to generate disturbance sequences with different degrees of chaos; the Henon mapping parameter a varies in the range of 1.2-1.4, and the parameter b is fixed at 0.3 to generate two-dimensional chaotic attractor projection sequences; the Lorenz system parameters σ=10, ρ=28, and β=8 / 3 are obtained by numerical integration to obtain three-dimensional chaotic trajectories. The set of random disturbance parameters includes parameters of disturbance amplitude, disturbance frequency, disturbance phase, disturbance duration, and disturbance interval time. The disturbance amplitude is determined according to the percentage of the rated capacity of each power supply, with a typical range of 5%-15%; the disturbance frequency covers a wide frequency band of 0.01Hz-10Hz, with a focus on the low-frequency oscillation interval of 0.1Hz-2Hz; the disturbance phase is uniformly distributed between 0 and 2π to ensure the phase dispersion of different power supply disturbances; the disturbance duration is exponentially distributed, with an average duration of 5 minutes; the disturbance interval time is Poisson distributed, with an average interval of 10 minutes. The size of the parameter set is determined according to the number of power supplies and the time length, and a typical configuration is to generate 500-1000 groups of disturbance parameters for each power supply.

[0050] Then, the set of random disturbance parameters is assigned to each power supply to obtain power supply-specific disturbance instructions. The assignment strategy takes into account the technical characteristics and operating constraints of various types of power supplies, and uses a constrained random assignment algorithm to ensure the executability of the disturbance instructions. The disturbance instructions for photovoltaic power supplies mainly act on the duty cycle of the DC / DC converter and the modulation depth of the DC / AC inverter, with the disturbance amplitude limited to a range that does not affect the MPPT efficiency; the disturbance instructions for wind turbines are converted into pitch angle adjustments and generator torque set values, taking into account the response speed limitations of mechanical components; the disturbance instructions for energy storage systems are directly mapped to charge and discharge power instructions, which are checked for constraints according to the current SOC state; the disturbance instructions for diesel generators act on the governor and excitation regulator, with the disturbance rate limited to the device's bearing range. The instruction generation uses a real-time operating system to ensure that the timing accuracy of the disturbance instructions reaches the millisecond level. Through constrained mapping and real-time scheduling, a sequence of executable disturbance instructions specific to each power supply is generated.

[0051] Next, the power sources are driven to perform output disturbance to obtain disturbance response data using power-specific disturbance instructions. The disturbance instructions are issued to the local controllers of the power sources after security verification through the remote control function of the microgrid SCADA system. Real-time monitoring during execution ensures that the disturbance is within the safety boundary. The protection limit is set to a voltage deviation of ±7% and a frequency deviation of ±0.5 Hz. When the limit is exceeded, the disturbance is automatically stopped. The response data collection uses a distributed measurement architecture, with synchronized phasor measurement units (PMUs) configured at the outlets of each power source to achieve synchronous collection of electrical quantities. The collected data includes three-phase voltage, three-phase current, active power, reactive power, power factor, frequency, phase angle, and other electrical parameters, as well as temperature, speed, SOC, and other operating state parameters. The data timestamp accuracy reaches the microsecond level, ensuring time alignment of data from different measurement points. During the disturbance process, the system records the complete dynamic process from steady state to disturbance excitation to new steady state. Typical disturbance response data contains millions of data points.

[0052] Finally, the disturbance response data is integrated and processed to generate chaotic state operation data. First, time alignment processing is performed to unify the scattered data to the same time reference based on the GPS synchronous clock. The data cleaning process identifies and processes abnormal values, missing values, and noise interference. The 3σ criterion is used to remove outliers, and linear interpolation is used to fill short missing data. The data standardization uses the zero-mean normalization method to eliminate the dimension influence of different physical quantities. Feature enhancement processing calculates the change rate, acceleration, and mutual correlation coefficient of each power output to enrich the information dimension of the data. Data segmentation divides the long time series into multiple segments according to the changes in system operating state, with each segment corresponding to a quasi-steady state or transient process. Through data integration and preprocessing, a structured, standardized, and information-rich chaotic state operation data set is generated.

[0053] In some embodiments, the extraction of the inherent frequency characteristics of each power source from the chaotic state operation data to obtain the resonance coupling matrix includes: performing frequency domain transformation on the chaotic state operation data to obtain frequency spectrum distribution data; identifying the dominant frequency of each power source from the frequency spectrum distribution data to generate a frequency feature identifier; and performing coupling strength analysis based on the frequency feature identifier to construct a resonance coupling matrix.

[0054] The frequency spectrum distribution data is obtained by frequency domain transformation of the chaotic state operation data. In the multi-resolution analysis framework, the corresponding processing method is selected according to the physical characteristics and analysis requirements of different frequency bands, and different analysis accuracies are used in different frequency ranges. The low frequency band (0.01-1Hz) contains the main oscillation mode of the microgrid, and the Welch power spectrum estimation with high resolution is adopted, the window length is set to 1 / 8 of the data length, the overlap rate is 75%, the noise influence is suppressed by multi-window average, and the frequency resolution of 0.01Hz is obtained; the middle frequency band (1-10Hz) covers the dynamic response characteristics of the control loop, the standard FFT is adopted with the Hamming window function to reduce the spectrum leakage, and the calculation efficiency and accuracy requirements are balanced; the high frequency band (10-100Hz) mainly reflects the switching frequency and high-order harmonics, and a fast algorithm is adopted to reduce the calculation complexity. Cross spectrum analysis calculates the mutual power spectrum density between different power sources, retains the amplitude and phase information through complex form, and reveals the coupling strength and phase relationship between the power sources at each frequency point. The coherence function calculates and evaluates the linear correlation degree at different frequency components, the value range is 0-1, the frequency point with the coherence coefficient greater than 0.8 is considered to exist strong coupling, 0.5-0.8 is moderate coupling, and less than 0.5 is considered to be weak coupling or no coupling. The de-trend and windowing pretreatment are carried out before frequency domain transformation, and the influence of direct current component and boundary effect is eliminated. Through multi-resolution frequency domain transformation and cross spectrum analysis, the fine frequency spectrum distribution data of the power output of each power source is obtained.

[0055] The frequency characteristic signature is generated by identifying the dominant frequencies of each power source from the frequency spectrum distribution data. The peak detection algorithm with adaptive threshold is adopted. Firstly, the statistical characteristics of power spectrum are calculated, and the frequency points whose power spectral density exceeds the average value by 3 times and are local maximum values are set as candidate peak values. The peak screening process considers the physical meaning of frequency and the characteristics of the system, and eliminates the non- intrinsic frequency components of 50 Hz power frequency and its integer harmonic (100 Hz, 150 Hz, etc.), and retains the characteristic frequencies reflecting the dynamic characteristics of the power source. For photovoltaic power source, the identified 0.35 Hz dominant frequency corresponds to the search oscillation of MPPT control, 0.52 Hz reflects the dynamic response of DC / DC converter, and 0.78 Hz is related to the current loop of inverter; the 0.15 Hz ultra-low frequency oscillation of wind turbine generator set is derived from the mechanical characteristics of wind wheel, 0.28 Hz embodies the regulation process of pitch angle control, and 0.45 Hz is related to the electromagnetic torque control of generator; the 0.65 Hz of energy storage system corresponds to the transition process of charge-discharge switching, and 1.2 Hz and 1.8 Hz reflect the response characteristics of voltage control loop and current control loop respectively. The frequency characteristic signature adopts a structured vector form, which includes four core attributes of frequency value fi, normalized power spectral density amplitude Pi, initial phase φi and quality factor Qi, wherein the quality factor reflects the sharpness of the frequency component, and the higher the Q value, the more stable the oscillation mode. Through systematic spectrum analysis and physical interpretation verification, the frequency characteristic signature accurately describes the oscillation characteristics and control response law of each power source.

[0056] Based on the frequency characteristics, the coupling strength analysis is performed to construct the resonance coupling matrix. First, the modal analysis method is used to decouple the frequency characteristics and construct the state space model of the multi-source system. The natural frequency, damping ratio and mode shape vector of each order oscillation mode are obtained by solving the system characteristic equation. In the three-source system including photovoltaic, wind power and energy storage, the first order mode is a global in-phase oscillation mode with a frequency of 0.25 Hz, and all sources participate in the oscillation with similar amplitude and phase. The second order mode is a regional oscillation mode with a frequency of 0.48 Hz, and photovoltaic and wind power show anti-phase oscillation characteristics, while energy storage remains relatively stable. The third order mode is a local oscillation mode with a frequency of 0.82 Hz, mainly dominated by the energy storage system. The participation factor analysis quantifies the contribution of each source in different oscillation modes, the observability analysis evaluates the ability to observe each mode from the system output, and the modal damping ratio reflects the damping characteristics of the oscillation. These modal parameters together constitute the standard coupling data that describe the oscillation characteristics of the system. Then, the frequency abnormal signal is injected into the standard coupling data to obtain the abnormal response data. The abnormal signal design covers typical disturbance conditions that may be encountered in actual operation: a step-type frequency offset simulating the frequency jump caused by the sudden switching of a large-capacity load, a step signal with an amplitude of 0.2 Hz and a duration of 5 seconds; a pulse-type disturbance simulating the transient impact caused by line faults or device tripping, a pulse signal with an amplitude of 0.5 Hz and a width of 2 seconds; a random disturbance simulating persistent random interference such as wind speed fluctuations and cloud shading, a Gaussian white noise signal with a standard deviation of 0.1 Hz. The photovoltaic system shows fast but large amplitude frequency fluctuations under disturbance due to its fast control response but small inertia; the wind power system has a relatively slow frequency change but a long duration due to the presence of mechanical inertia; the energy storage system exhibits excellent frequency support characteristics due to its fast power regulation capability. Through comprehensive abnormal injection testing, the abnormal response data set reflecting the response characteristics of each source under disturbance conditions is obtained. Then, the anti-interference analysis is performed on the abnormal response data to generate robust coupling parameters. The anti-interference analysis is based on the modal participation factor and the abnormal response characteristics to calculate the degree of mutual influence of each source under disturbance. By analyzing the correlation of the frequency trajectories of each source in the abnormal response data, the dynamic coupling strength between the sources is extracted. The photovoltaic-wind power system shows a high degree of synchronization in frequency change during the disturbance process, with a correlation coefficient of more than 0.85; the photovoltaic-energy storage system shows complementary frequency response characteristics, with the energy storage quickly compensating for the frequency deviation of the photovoltaic system; the wind power-energy storage coupling shows that the energy storage actively tracks the slow change of the wind power. The stability evaluation is performed by calculating the degree of preservation of the modal characteristics before and after the disturbance, and the recovery capability evaluation is performed by fitting the decay curve after the disturbance to obtain the time constant. Finally, the robust coupling parameters that quantify the interaction strength between the sources under normal and abnormal conditions are generated.The matrix is constructed directly using robust coupling parameters as matrix elements, wherein the non-diagonal element Cij represents the coupling strength between power supply i and power supply j, and the numerical value reflects the degree of mutual influence of the two in frequency oscillation. The coupling parameter between photovoltaic and wind power is the largest, reflecting the strong interaction between new energy power sources; the coupling parameters of energy storage and other power sources are moderate, reflecting the coordination of energy storage in the system. The diagonal element is set as the self-regulating ability parameter of each power supply, which is related to the modal damping characteristics. The symmetry of the matrix is guaranteed by the physical characteristics of the interaction between power sources, and the positive definiteness ensures the stability of the subsequent control design. By directly using the robust coupling parameters verified for anti-interference, the resonance coupling matrix is constructed, which accurately reflects the frequency coupling characteristics of the multi-source microgrid.

[0057] A frequency synchronization coordination mechanism between power sources is generated using the resonance coupling matrix. Based on the dynamic relationship between power sources revealed by the resonance coupling matrix, a distributed cooperative control strategy is designed to achieve frequency synchronization of multiple power sources. The coordination control adopts a distributed algorithm based on consensus theory, and each power source adjusts its frequency control parameters according to the coupling strength with adjacent power sources. The control law design adopts a proportional-integral form, and the proportional gain is determined according to the coupling coefficient, and the integral gain is set according to the system damping requirement. In the strong coupling pair of photovoltaic and wind power, a feedforward compensation of wind power frequency deviation is introduced into the frequency control loop of photovoltaic inverter, and the compensation coefficient is equal to the coupling coefficient; the wind power converter also introduces a feedforward term of photovoltaic frequency to achieve bidirectional frequency support. The control strategy of the energy storage system, as the main actuator for frequency regulation, dynamically allocates frequency modulation power according to the coupling relationship with other power sources. The virtual synchronous generator (VSG) control introduces a coupling matrix to modify the virtual inertia and damping coefficient, enhancing the synchronization stability of the multi-source system. The adaptive adjustment mechanism updates the coupling matrix according to real-time operation data to ensure the dynamic optimization of the coordination control. In an industrial park microgrid application, after implementing the frequency synchronization coordination mechanism based on the resonance coupling matrix, the frequency fluctuation range of the multi-source system is reduced from ±0.3Hz to ±0.1Hz, and the frequency recovery time is shortened from 8 seconds to 3 seconds. Through accurate modeling of the coupling matrix and optimized design of the coordination control, a frequency synchronization coordination mechanism between power sources is generated to achieve frequency synchronization of multiple power sources and improve the stability of the system.

[0058] In step S130, a preliminary load distribution scheme is generated based on the frequency synchronization coordination mechanism, and distribution vulnerability identification data is generated by searching for counterexamples of the preliminary load distribution scheme. The reinforcement distribution strategy is obtained according to the distribution vulnerability identification data.

[0059] Specifically, the frequency synchronization coordination mechanism is used to formulate the preliminary load distribution scheme of the microgrid. The frequency synchronization coordination mechanism reveals the strong coupling characteristics of photovoltaic-wind power (coupling coefficient 0.85), the frequency support capability of energy storage, and the dynamic response characteristics of each power supply, which provides a basis for load distribution. The distribution principle follows the hierarchical goal of frequency stability priority and economy second, and the power supply with fast frequency response and strong regulation capability is preferred to bear the load fluctuation part. According to the coupling strength shown by the resonance coupling matrix, the strong coupling power supply group is regarded as a whole for load distribution to avoid oscillation caused by improper distribution. The energy storage system is designated as the main force of frequency regulation due to its fast response characteristics (response time less than 100 milliseconds) and bears high-frequency load fluctuations; photovoltaic and wind power bear the basic load according to their real-time output and prediction curve; diesel generators are used as backup support and start when new energy output is insufficient. The distribution algorithm uses a multi-objective optimization method, and the objective function includes three dimensions of frequency deviation minimization, power generation cost minimization, and network loss minimization. The constraint conditions cover the upper and lower limits of the output of each power supply, the climbing rate limit, the SOC constraint of energy storage, and other technical limitations. Under typical working conditions, photovoltaic power bears 35% of the basic load, wind power bears 30%, energy storage bears 25% of the regulation load, and diesel generators maintain 10% of the rotating reserve. By determining the output plan and bearing proportion of each power supply at different times, the preliminary load distribution scheme is finally generated.

[0060] In some embodiments, the counterexample search on the preliminary load distribution scheme generates distribution vulnerability identification data, including: constructing an opposite distribution mode of the preliminary load distribution scheme; obtaining a conflict response result based on the conflict scene simulation of the opposite distribution mode; generating a distribution vulnerability label by analyzing the conflict response result; and generating distribution vulnerability identification data according to the distribution vulnerability label.

[0061] An opposite allocation mode is constructed to identify potential defects of the initial allocation scheme by reverse thinking. The design of the opposite mode is based on the "worst-case" principle, which reverses the allocation logic in the initial scheme: in the initial scheme, photovoltaic power is preferentially used to bear daytime load, while in the opposite mode, photovoltaic power is forced to bear high load during low-output periods such as cloudy days or early morning and evening; in the initial scheme, energy storage mainly adjusts high-frequency fluctuations, while in the opposite mode, energy storage is forced to charge and discharge at full power for a long time, quickly depleting the SOC; in the initial scheme, wind power and photovoltaic power are coordinated to output, while in the opposite mode, the two are set to operate in opposite directions. Extreme working condition construction includes: all new energy sources at the lowest output, load surge of 50%, multiple power sources failure, and other small probability but high risk scenarios. The generation of the opposite parameters uses a genetic algorithm to maximize the system instability, searching for parameter combinations that can expose allocation defects. Boundary condition testing pushes various constraints to the limit, such as forcing the energy storage SOC to approach 0% or 100%, and making the diesel generator start and stop frequently. The timing of the confrontation creates conflicts during the transition period of the allocation switch by disrupting the original coordination timing. The opposite mode library ultimately contains 5-10 typical confrontation scenarios, covering capacity shortage, response lag, coordination failure, and other potential problems, forming an opposite allocation mode that comprehensively challenges the initial allocation scheme.

[0062] Based on the opposite allocation mode, the system response under various extreme working conditions is recorded through conflict scenario simulation of the microgrid system. The simulation platform is built using MATLAB / Simulink, including detailed dynamic models of various power sources, control system models, and grid models, with a simulation step size of 50 microseconds to capture fast transient processes. During the execution of the conflict simulation, the load is allocated according to the opposite mode, while monitoring the dynamic changes of key indicators such as frequency, voltage, and power balance. In the "all new energy low output + load surge" scenario, the frequency drops to 49.2 Hz within 2 seconds, triggering the low-frequency protection boundary; in the "energy storage SOC depletion + wind power fluctuation" scenario, the frequency appears ±0.8 Hz sustained oscillation, exceeding the allowed range; in the "photovoltaic rapid fluctuation + wind power reverse" scenario, there is a risk of voltage collapse due to power shortage. The collection of response data includes electrical parameters, control variables, and protection action signals, forming a multi-dimensional system state record. The capture of abnormal events includes protection device action, control saturation, communication delay exceeding limit, etc., which directly reflects the weak links of the allocation scheme. The multi-dimensional system state record and abnormal event data accumulated through conflict scenario simulation ultimately obtain the conflict response results.

[0063] Deep analysis is conducted on the conflict response results to identify and label the vulnerability characteristics of the allocation scheme. The analysis framework includes three dimensions: stability evaluation, robustness evaluation, and recovery capability evaluation. Stability analysis determines the stability margin of the system under disturbance by calculating the Lyapunov index. When the index is positive, it indicates that the system is unstable, and is labeled as "stability vulnerable". Robustness analysis evaluates the sensitivity of the system to parameter changes by identifying high sensitivity parameter combinations through the sensitivity matrix. Parameters with sensitivity exceeding the set threshold are labeled as "robustness vulnerable". Recovery capability analysis measures the recovery time and recovery path after a fault. If the recovery time exceeds 10 seconds or there is a recovery oscillation, it is labeled as "recovery force vulnerable". The quantification of vulnerability uses a risk score mechanism, considering the probability of occurrence and the severity of impact. Risk scores exceeding 0.7 are classified as high-risk vulnerabilities. Correlation analysis identifies the causal relationships between vulnerabilities, such as insufficient energy storage capacity leading to reduced frequency regulation capability, which in turn triggers system oscillation. Time characteristic analysis finds that some vulnerabilities only occur at specific times, such as power balance vulnerability during the morning photovoltaic startup phase. Through multi-dimensional vulnerability analysis and risk quantification evaluation, allocation vulnerability labels are generated.

[0064] According to the allocation vulnerability labels, systemically organize and classify various allocation defects to generate structured allocation vulnerability identification data. The vulnerability classification uses a hierarchical architecture: the first-level classification includes capacity-type vulnerabilities, time-series-type vulnerabilities, coordination-type vulnerabilities, and extreme-type vulnerabilities; the second-level classification further refines, such as total capacity deficiency, backup capacity deficiency, and regulation capacity deficiency under capacity-type vulnerabilities. Each vulnerability entry contains multi-dimensional information such as vulnerability description, trigger condition, impact assessment, and associated vulnerabilities. Priority is determined based on risk score and repair difficulty, with high-risk and easy-to-repair vulnerabilities being prioritized. The quantitative description of vulnerabilities uses a standardized template to clearly define the numerical boundaries of vulnerabilities, such as "when photovoltaic output is less than 20% of installed capacity and load is higher than average by 30%, a 150kW power gap occurs". The association between vulnerabilities is represented by a directed graph, revealing the propagation path and cumulative effect of vulnerabilities. Statistical analysis shows that time-series-type vulnerabilities account for the highest proportion (35%), reflecting the complexity of coordination control; although capacity-type vulnerabilities account for a lower proportion (20%), they have the most serious impact. The structured storage of vulnerability identification data uses JSON format, facilitating subsequent automated processing and strategy generation.

[0065] According to the various defects exposed in the allocation vulnerability identification data, targeted improvement measures are designed to form the reinforcement allocation strategy. The strategy making adopts the "vulnerability-countermeasure" mapping method, and each type of vulnerability corresponds to a specific reinforcement measure. For capacity-type vulnerabilities, the reinforcement strategy includes: dynamically adjusting the proportion of backup capacity, and adaptively determining the backup demand according to the new energy prediction error; Establish a multi-level backup system, fast backup (energy storage), rotating backup (diesel engine), cold backup (interruptible load) hierarchical configuration; Optimize the energy storage charging and discharging strategy, pre-charge in non-critical periods to ensure sufficient adjustment capacity in critical periods. For time-series vulnerabilities, the reinforcement strategy introduces predictive control to start the backup power source 5-15 minutes in advance to avoid power gaps; Design a smooth switching mechanism to transition through a ramp function when switching between different operating modes; Establish a priority queue to ensure the power supply timing of critical loads. For coordination-type vulnerabilities, the reinforcement strategy uses a distributed consistency algorithm to ensure the synchronization of multiple power source actions; Introduce virtual inertia control to enhance the frequency support capability of the system; Design a conflict resolution mechanism to automatically arbitrate when multiple power source instructions conflict. For extreme vulnerabilities, the reinforcement strategy includes emergency control plans, preset emergency response processes under extreme conditions; Island operation capability building to ensure autonomous operation of local power grids; Black start scheme design, fast recovery path after system collapse. The reinforcement measures for various vulnerabilities are integrated and optimized, and finally the reinforcement allocation strategy is obtained.

[0066] In step S140, based on the reinforcement allocation strategy, obtain the load bearing capacity evaluation data of each power source, match and analyze the load demand data and the load bearing capacity evaluation data to generate a load conduction topology, and use the load conduction topology to obtain the pressure distribution data between each power source.

[0067] Specifically, the load bearing capacity of each power source in the microgrid is comprehensively evaluated by using the reinforcement allocation strategy. The reinforcement allocation strategy clarifies that the energy storage system needs to maintain a rapid standby capacity, the photovoltaic and wind power need to consider the influence of prediction error, and the diesel generator needs to maintain a moderate rotating standby, which directly affects the actual available capacity of each power source. The evaluation framework adopts a multi-dimensional analysis method. The technical dimension evaluation includes hard indicators such as rated capacity, adjustable capacity, climbing rate, response time, etc. The operation dimension evaluation covers real-time factors such as current output level, remaining adjustment margin, cumulative running time, maintenance status, etc. The reliability dimension evaluation considers statistical indicators such as historical failure rate, average failure-free time, recovery time, etc. The bearing capacity evaluation of the photovoltaic system needs to be combined with the irradiance prediction and temperature correction. Based on the rated capacity under standard test conditions, the actual environmental conditions are converted. The evaluation of the wind power system introduces the wind speed-power curve and the influence of turbulence intensity, considering key nodes such as cut-in wind speed, rated wind speed, and cut-out wind speed. The evaluation of the energy storage system focuses on the current SOC state and the charge and discharge power limit, and establishes the mapping relationship between SOC and power bearing capacity. The evaluation of the diesel generator includes cold start time, hot standby state, fuel reserve and other factors. The dynamic evaluation mechanism is updated every 15 minutes to ensure the timeliness of the bearing capacity data. The comprehensive evaluation results are standardized and updated in real time to obtain the load bearing capacity evaluation data of each power source.

[0068] In some embodiments, the matching analysis of the load demand data and the load bearing capacity evaluation data generates a load conduction topology, including: performing difference analysis on the load demand data and the load bearing capacity evaluation data to obtain load gap data; constructing load transfer paths between power sources based on the load gap data; analyzing each transfer path to generate a conduction impedance coefficient; and generating a load conduction topology using the conduction impedance coefficient.

[0069] The real-time load demand data and the load bearing capacity evaluation data of each power supply are compared point by point, and the supply-demand difference is calculated to identify the load gap. The difference analysis adopts a time series alignment method to ensure that the demand data and the bearing capacity data are compared at the same time section. The analysis under normal working conditions shows that during the daytime period with sufficient sunlight, the photovoltaic bearing capacity exceeds the distribution demand, resulting in a positive surplus; during the evening period, the photovoltaic output decreases, while the load demand remains high, resulting in a negative gap. The classification of the gap includes instantaneous gap and continuous gap. The instantaneous gap is caused by short-time load spikes or power output fluctuations, and the duration is usually less than 5 minutes. The continuous gap is caused by insufficient power capacity or sustained high load, and requires the use of backup resources. The quantification of the gap includes not only the power value but also the energy accumulation, which is obtained by time integration of the power gap. Spatial distribution analysis identifies the distribution characteristics of the gap at different nodes. The nodes near the load center have more serious gaps. Time distribution analysis reveals the periodicity of the gap. The gap is most prominent during the morning and evening peak periods of weekdays. The severity of the gap is classified into three levels: a mild gap can be solved by adjusting between power sources, a moderate gap requires the start of backup power, and a severe gap may trigger load control. Through systematic analysis and classification of the supply-demand difference, the load gap data is obtained.

[0070] Based on the identified load gap data, a load transfer path between power sources is constructed to achieve supply-demand rebalancing. The path construction follows the principles of local transfer, minimum loss, and reliability priority. When the photovoltaic power appears a gap, the first choice is to obtain support from the energy storage system, which has fast response speed and high transfer efficiency. The second choice is to adjust from the wind power system, but the output stability of the wind power itself needs to be considered. The third choice is to start the diesel generator, which has slow response but high reliability. The determination of the path considers the electrical distance rather than the physical distance, and the electrical coupling strength between power sources is calculated through the impedance matrix. The multi-path parallel mechanism allows simultaneous load transfer from multiple power sources, improving the reliability and flexibility of the transfer. Dynamic path adjustment optimizes the transfer scheme according to the real-time operating state, and automatically switches to the backup path when a certain path is blocked. Path capacity limitation ensures that the transfer process does not exceed the carrying capacity of the line and equipment. Priority setting ensures that the transfer path of critical loads is always unblocked. The bidirectional design of the path not only supports gap filling, but also supports surplus absorption, forming a flexible energy interconnection network. The topology structure of the transfer path presents a mesh feature, and there is at least one feasible path between any two power sources. Through path planning and network construction driven by gap data, the load transfer path between power sources is formed.

[0071] The impedance characteristics of the load transfer paths between power sources are analyzed to quantify the conduction difficulty of each path. According to the physical characteristics and operating parameters of the transfer paths, the conduction impedance coefficient is calculated using a comprehensive impedance model Z_ij = α·R_ij+ β·X_ij + γ·C_ij, where Z_ij represents the conduction impedance of the path from power source i to power source j, R_ij is the electrical impedance component reflecting the line resistance and transmission loss of the path; X_ij is the control impedance component representing the response delay and adjustment complexity of the path; C_ij is the economic impedance component quantifying the transmission cost of the path; α, β, γ are weight coefficients. For the photovoltaic- energy storage transfer path, due to the short physical distance and simple control, the electrical impedance R_pv-es is small, and the control impedance X_pv-es is also low due to the fast response characteristics of energy storage; the wind power- energy storage path has a relatively large R_wind-es due to the long distance; the control impedance X_diesel of the diesel-related path is generally high, reflecting the complexity of the start-stop process. The weight coefficients are adjusted according to the system operating mode, with α = 0.3, β = 0.5, and γ = 0.2 in normal operation, and β is increased to 0.7 in emergency state to prioritize response speed. The time-varying characteristics of the impedance are reflected through online updating, and the change of energy storage SOC will affect the impedance value of the related path. Standardization processing takes the system average impedance as the benchmark to ensure the comparability of different types of impedance. According to the comprehensive evaluation and parameter calculation of the transfer path, the conduction impedance coefficient is finally generated.

[0072] The conduction impedance coefficient is used to construct a topology structure reflecting the conduction relationship of the load between power sources. The topology construction uses a weighted directed graph model, with nodes representing various types of power sources and loads, and edges representing transfer paths. The edge weight is equal to the inverse of the conduction impedance coefficient, representing the conduction capacity. The hierarchical design of the topology divides the power sources into the main layer, support layer, and standby layer, with the main layer including photovoltaic and wind power, bearing the basic load; the support layer mainly consists of energy storage, providing rapid adjustment; the standby layer includes diesel generators as emergency protection. The adjacency matrix representation method is convenient for computer processing and path search, and the matrix elements directly reflect the connection relationship and conduction strength between nodes. The shortest path algorithm is used to find the optimal conduction path, considering both the minimum impedance and the highest reliability. The dynamic characteristics of the topology are reflected through the time-varying adjacency matrix, and the connection strength between nodes is adjusted accordingly as the system operating state changes. Key node identification is completed by calculating the degree centrality, betweenness centrality, and other indicators, and the energy storage system is usually a key node. Topology optimization improves reliability through measures such as adding redundant paths and reducing the dependence of key nodes. Through the quantitative representation of conduction capacity and network organization, the load conduction topology is generated.

[0073] The load conducting topology is used for power flow analysis and load distribution calculation to obtain the stress state of each power supply. The load conducting topology provides complete network structure and conducting capacity information. Based on the node connection relationship in the topology, a power balance equation set is established, and the Newton-Raphson method is used to solve the voltage and power of each node. The calculation formula of the stress index is P_i = (L_i / C_i) x (1 + σ_i), where P_i is the stress index of power supply i, L_i is the actual load borne according to the topology power flow calculation, C_i is the rated capacity, and σ_i is the dynamic adjustment load coefficient. The pressure conduction in the topology network follows the rule of ΔP_j = Σ(K_ij x (P_i - P_j)), where the conduction coefficient K_ij is directly taken from the edge weight of the load conducting topology. The pressure conduction process is tracked through the topology path. When the photovoltaic pressure is too high, the pressure is preferentially conducted to the energy storage according to the topology connection relationship, and then diffused to the wind power and diesel engine when the energy storage is saturated. The calculation of pressure distribution considers the hierarchical structure of the topology. The main force layer bears the steady-state pressure, the support layer absorbs the dynamic pressure, and the standby layer remains in a low-pressure standby state. The time evolution analysis is based on the dynamic characteristics of the topology to track the migration process of the pressure in the network. Through the power flow calculation and pressure conduction analysis under the constraint of the topology structure, the pressure distribution data among the power supplies are obtained.

[0074] In step S150, a boundary active detection instruction set is generated based on the pressure distribution data. The boundary active detection instruction set is used to drive each power supply to gradually approach the operating boundary to obtain boundary response data. Extreme value analysis is performed on the boundary response data to generate the load bearing limit parameters of each power supply.

[0075] Specifically, according to the current pressure state and pressure gradient of each power source displayed by the pressure distribution data, a proactive detection strategy is designed to identify the operating boundary of each power source. The pressure distribution data shows that the energy storage system pressure index fluctuates sharply, and its charge-discharge switching boundary needs to be detected; photovoltaic power has a larger detection space when the pressure is low at noon, but needs to be cautious when the pressure is high in the morning and evening; the pressure of wind power is related to wind speed, and detection needs to consider wind speed prediction; the pressure of diesel generator is low, and has a larger boundary detection potential. The design of the detection instruction adopts a gradual strategy to avoid aggressive detection leading to system instability. The instruction parameters include detection direction, step size, duration, and rollback conditions. The detection direction is determined according to the current pressure state, and the low-pressure power source is detected in the direction of increasing output, and the high-pressure power source is detected in the direction of reducing output. The step size design uses an adaptive algorithm, with an initial step size of 2% of the rated capacity, which is dynamically adjusted according to the response sensitivity. The duration is matched with the system time constant to ensure that the next step is detected after reaching steady state. The rollback conditions include safety constraints such as frequency deviation exceeding limit, voltage exceeding limit, and temperature exceeding limit. The arrangement of the detection sequence considers the mutual influence between power sources to avoid the superposition effect caused by simultaneous detection of multiple power sources. The time window is selected during the stable period of load to reduce the interference of external disturbances on detection. Through the pressure state-driven detection strategy development and parameter optimization, a set of boundary proactive detection instructions is generated.

[0076] In some embodiments, the use of the boundary proactive detection instruction set to drive each power source to gradually approach the operating boundary to obtain boundary response data includes: based on the boundary proactive detection instruction set, the output of each power source is increased in steps; during the stepwise output increase, the power source state is monitored to obtain response characteristic data; the response characteristic data is analyzed nonlinearly to identify mutation points, and the mutation point parameters are recorded; and based on the mutation point parameters, boundary response data is generated.

[0077] The stepwise increase instructions in the boundary active probing instruction set drive each power source to gradually increase its output level. The stepwise increase uses a step control method, and each step increment is determined according to the probing instruction set. The energy storage system increases the charging and discharging power by 5 kW each step, the photovoltaic increases by 3% each step by adjusting the MPPT bias, the wind power increases by 2% each step by adjusting the pitch angle, and the diesel generator increases the load by 4% each step by adjusting the governor. The step interval is set to 2 minutes to ensure that the system reaches a new steady state balance. The control of the increase process uses closed-loop feedback, which monitors the frequency and voltage deviation in real time and automatically slows down or pauses the increase when the deviation exceeds the set threshold. When multiple power sources are coordinated for probing, a polling mechanism is used, and after one power source completes a step of probing, the other power sources remain in their current state to avoid mutual interference. The probing range starts from the current operating point and gradually approaches the theoretical limit value. The energy storage probes to the SOC close to the upper and lower limits, the photovoltaic probes to the inverter capacity limit, the wind power probes to the rated power or cut-out wind speed, and the diesel engine probes to the rated power or temperature limit. The abnormal processing mechanism runs continuously during the probing process, and any abnormality immediately stops the probing and reverts to a safe operating point. Finally, through the precise execution and safety guarantee of the stepwise control, the gradual increase of the output of each power source is realized.

[0078] During the entire process of stepwise output increase, a comprehensive monitoring system is deployed to collect the state parameters of each power source in real time. The monitoring content covers multiple dimensions such as electrical parameters, thermal parameters, and mechanical parameters. Electrical parameters include output power, voltage, current, power factor, harmonic content, etc., with a sampling frequency of 1 kHz to capture rapid changes; thermal parameters monitor the temperatures of key components such as battery temperature, inverter radiator temperature, and diesel engine cylinder temperature, with temperature rise rate as an important monitoring indicator; mechanical parameters mainly target rotating equipment, including wind turbine rotor speed, vibration amplitude, and diesel engine speed fluctuation. The synchronization of data acquisition is guaranteed by GPS clock, and all measurement point data have a unified timestamp. Abnormality identification algorithms run in real time, including over-limit detection, change rate detection, trend prediction, and other methods. Data storage uses a circular buffer mechanism to retain the last 30 minutes of high-precision data for analysis. Feature extraction focuses on sensitive parameters related to the boundary, such as efficiency inflection points, temperature rise acceleration points, and vibration sudden increase points. Nonlinear characteristics of response characteristics are identified through phase space reconstruction methods, revealing the dynamic changes of the system when approaching the boundary. Through comprehensive monitoring and feature extraction, rich response characteristic data is obtained.

[0079] The nonlinear analysis of the collected response characteristic data identifies the mutation point where the system behavior changes qualitatively. Multiple methods are used to verify the accuracy of the mutation point identification. The bifurcation analysis identifies the critical point where the system stability changes by plotting the parameter-state bifurcation diagram. When the photovoltaic inverter output approaches the capacity limit, the voltage-power curve exhibits a turning-back phenomenon, indicating that the static stability boundary has been reached. The phase plane analysis projects the multi-dimensional state variables onto a two-dimensional plane to observe the topological structure changes of the system trajectory. When the SOC of the energy storage system approaches the limit, the charging and discharging switching trajectory changes from smooth transition to sharp jump. The Lyapunov exponent calculation assesses the system's sensitivity to initial conditions. When the index changes from negative to positive, it indicates that the system has entered a chaotic state, which is a characteristic of wind turbines near the cut-out wind speed. The catastrophe theory applies the cusp catastrophe model to describe the mutation behavior of the output power with respect to the control parameters, and identifies the power collapse point of the diesel generator when it is overloaded. The recording of the mutation point parameters includes the power level at which the mutation occurs, the corresponding control parameter value, the environmental conditions, and the precursor characteristics. In a multi-source system, it is also necessary to identify the coupling mutation, which triggers a chain reaction in other power sources. Through the system analysis of nonlinear characteristics and the identification of critical points, the key mutation point parameters are recorded.

[0080] Based on the identified and recorded mutation point parameters, the response data set reflecting the boundary characteristics of each power source is organized. The structured organization of boundary response data includes two types: static boundary and dynamic boundary. The static boundary reflects the steady-state operating limit. The static boundary of the energy storage system includes the maximum continuous charging and discharging power, the SOC available range, and the temperature limit boundary. The static boundary of the photovoltaic system is determined by the inverter capacity, the MPPT range, and the voltage limit. The static boundary of the wind power includes the minimum output corresponding to the cut-in wind speed, the rated output corresponding to the rated wind speed, and the shutdown boundary corresponding to the cut-out wind speed. The static boundary of the diesel engine is mainly the rated power and the minimum stable operating power. The dynamic boundary describes the limit in the transient process, including the maximum power change rate, the frequency response limit, and the fault ride-through capability. The time-varying characteristics of the boundary are represented by parameter curves, such as the available power of the energy storage system with respect to the SOC, the power output of the photovoltaic system with respect to the temperature, and so on. The quantification of safety margin is based on the distance from the mutation point to the operating point, providing a reference for subsequent control strategies. The relationship between boundaries is described by the association matrix, and the tightening of one boundary may lead to the relaxation of other boundaries. The influence of environmental factors on the boundary is reflected by the correction coefficient, and the boundaries of each power source generally shrink in high-temperature environments. After the system organization of mutation point parameters and the extraction of boundary characteristics, the boundary response data is finally generated.

[0081] The extreme value analysis of the boundary response data determines the load carrying capacity limit of each power source under different operating conditions. The extreme value analysis uses statistical extreme value theory to fit the boundary data with generalized extreme value distribution (GEV) to obtain the limit values under different confidence levels. For the energy storage system, the analysis shows that the maximum continuous discharge power under standard conditions is 120% of the rated power, but the duration is not more than 15 minutes; the maximum instantaneous power can reach 150% of the rated value, and the duration is not more than 1 second; the available SOC range is 10%-90%, and in extreme cases it can be extended to 5%-95% but will affect the battery life. The load carrying limit of the photovoltaic system is mainly limited by the inverter capacity, and the short-time overload capacity can reach 110% of the rated power; under low irradiance conditions, the MPPT efficiency decreases, resulting in actual available power of only 85% of the theoretical value. The wind turbine maintains rated output above the rated wind speed through pitch angle control, but can output 105% of the rated power for a short time under gust conditions; the minimum stable output under low wind speed is 5% of the rated power. The diesel generator has strong overload capacity and can run continuously at 110% of the rated power for 1 hour and at 120% for 10 minutes, but the increase in fuel consumption and maintenance cost needs to be considered. The probability distribution analysis of the limit parameters reveals the reliability level of the load carrying capacity, and P95 represents the load carrying capacity that can be reached under 95% probability. The time series correlation analysis finds that some limit parameters have memory effect, and high load operation in the early stage will reduce the subsequent load carrying capacity. Through statistical analysis and probability evaluation of the boundary data, the load carrying limit parameters of each power source are obtained.

[0082] Step S160, based on the load carrying limit parameters, set a unified time sequence adjustment reference, use the time sequence adjustment reference to synchronize the power output adjustment process of each power source to obtain coordinated time sequence data, and generate time sequence synchronization control parameters according to the coordinated time sequence data.

[0083] Specifically, a unified timing adjustment reference of the microgrid is established according to the power carrying limit parameters of each power source. The carrying limit parameters show that the energy storage system has a millisecond-level response capability but the duration is limited, the adjustment speed of photovoltaic and wind power is limited by physical processes, and the diesel generator has a long start-up time but good stability. These differentiated characteristics need to be coordinated through a unified reference. The design of the timing reference adopts a hierarchical time scale, the fast regulation layer takes 100 milliseconds as the reference period, mainly by the energy storage system to respond to instantaneous power fluctuations; the medium-speed regulation layer takes 1 second as the reference period, photovoltaic MPPT and wind power pitch angle adjustment respond at this level; the slow regulation layer takes 10 seconds as the reference period, diesel generator start-stop and power ramping are completed at this level. The reference clock uses a GPS synchronous clock source, with an accuracy of microseconds, ensuring the time synchronization of all power controllers. The adjustment margin is set based on the carrying limit parameters, the energy storage reserves 20% of the power margin for fast response, the photovoltaic and wind power maintains 5% of the adjustment margin, and the diesel engine maintains 15% of the standby capacity. Timing constraints include minimum action interval, maximum adjustment rate, coordination waiting time, etc., to prevent equipment wear caused by frequent adjustment. The priority mechanism is determined according to the response speed and adjustment cost, the energy storage responds to high-frequency components first, the new energy undertakes medium-frequency adjustment, and the diesel engine is used as low-frequency support. Through time scale layering and priority setting under limit parameter constraints, a unified timing adjustment reference is established.

[0084] In some embodiments, the synchronization processing of the power output adjustment process of each power source based on the timing adjustment reference to obtain coordinated timing data includes: generating power output adjustment instructions for each power source based on the timing adjustment reference; time labeling the power output adjustment instructions to obtain timing identification data; synchronously arranging the timing identification data to generate a coordinated execution sequence; and time peak-shaving optimization of the coordinated execution sequence to obtain coordinated timing data.

[0085] Based on the hierarchical period and priority rules defined in the unified timing adjustment reference, the corresponding output adjustment instructions are generated for each power source. The instruction generation process first decomposes the total regulation demand, and through spectral analysis, the demand is decomposed into different frequency components. High-frequency components above 100 Hz are allocated to the energy storage system, medium-frequency components between 1-100 Hz are allocated to photovoltaic and wind power, and low-frequency components below 1 Hz are allocated to diesel generators. The adjustment instruction of the energy storage system adopts a power direct control mode, and the instruction format includes target power value, execution time, duration, etc. The photovoltaic adjustment instruction is implemented through MPPT reference voltage biasing, and the instruction includes voltage adjustment amount and adjustment rate limit. The wind power adjustment instruction is converted into pitch angle control parameters, considering mechanical response delay. The diesel engine adjustment instruction includes start-stop command and power set value, with preheating and synchronization requirements. The generation time of the instruction is determined according to the prediction algorithm, and the energy storage instruction is generated in real time, the new energy instruction is generated 1-2 seconds in advance, and the diesel engine instruction is generated 30 seconds in advance to leave the starting time. The calculation of the instruction amplitude considers the distance from the current operating point to the limit, ensuring that it does not exceed the carrying capacity. The redundant instruction mechanism generates two sets of instructions for critical regulation tasks to improve reliability. After demand decomposition and parameter calculation under the timing reference framework, the output adjustment instructions of each power source are generated.

[0086] The generated output adjustment instructions of each power source are uniformly time-stamped to ensure accurate positioning in the time dimension. The time-stamping uses a dual identification method of absolute timestamp and relative timing. The absolute timestamp is based on the GPS synchronous clock, accurate to the microsecond level, and the format is "year-month-day hour:minute:second.microsecond". The relative timing takes the start point of the scheduling period as the reference to identify the execution position of the instruction within the period. In the marking process, the response delay characteristics of each power source are considered. The energy storage system has a response delay of about 50 milliseconds, which is pre-compensated based on the instruction timestamp. The photovoltaic MPPT response delay is 200-500 milliseconds, which is dynamically compensated according to real-time measurement. The wind power pitch angle adjustment delay is 1-3 seconds, and the diesel engine startup delay is 30-60 seconds. These delays are reflected in the time-stamping. Priority marking is synchronized with time-stamping. High-priority instructions have priority for execution, and low-priority instructions are automatically delayed when there is a conflict at the same time. Consistency checks ensure that there are no timing conflicts and logical contradictions, such as opposite adjustment instructions received by the same power source at the same time. A version control mechanism assigns a unique identifier to each instruction, supporting instruction updates and revocations. Structured storage of marked data uses a timing database to support fast time range queries and sorting operations. Through accurate time positioning and attribute labeling of instructions, time-stamped data is obtained.

[0087] The power adjustment instructions with time sequence identifiers are uniformly arranged to form a coordinated execution sequence on the time axis. The synchronized arrangement first sorts all instructions in ascending order of execution timestamp to form a global instruction queue. The time interval analysis of adjacent instructions in the queue identifies potential execution conflicts. When the execution times of multiple instructions overlap, the execution order is adjusted according to priority rules. The high-frequency regulation instructions of the energy storage system maintain the original timing as the first line of defense for fast response; the medium-frequency regulation instructions of photovoltaic and wind power are inserted without affecting the energy storage regulation; the low-frequency support instructions of the diesel engine are arranged for execution after the load change trend is clear. The evaluation of parallel execution capability determines which instructions can be executed simultaneously. Electrically independent power instructions can be executed in parallel, and instructions with power coupling need to be executed in series. A synchronization point is set every 1 second, and all power sources are aligned at the synchronization point to facilitate coordinated control. The integrity check of the execution sequence ensures that all necessary regulation requirements are covered and there is no regulation blank period. The executability verification of the sequence is completed through simulation rehearsal to check whether there are resource conflicts or timing lock. The generation of backup sequence considers the case of failure of the main sequence execution and provides a degradation operation scheme. According to the ordered organization and conflict resolution of timing identifier data, the coordinated execution sequence is finally obtained.

[0088] The coordinated execution sequence is optimized for timing peak shifting to avoid system impact caused by simultaneous large adjustment of multiple power sources. The objective function of peak shifting optimization is set to minimize the peak value of system power change rate, and the constraint conditions include that the adjustment task must be completed, not exceeding the carrying limit of each power source, and meeting the real-time balance requirement. The optimization algorithm uses a dynamic programming method to discretize continuous time into 100 millisecond time slices and optimize the execution time of instructions in each time slice. The fast regulation instructions of the energy storage system remain in priority, but the number of consecutive large adjustments is limited, and at least 500 milliseconds is required between two large adjustments. The regulation instructions of photovoltaic and wind power are staggered in execution. When photovoltaic output increases, wind power remains stable or adjusts slightly. The start and stop of the diesel engine is arranged during the load valley period to reduce the disturbance to the system. The smoothing of power change rate is achieved by inserting transition instructions to convert step changes to ramp changes. The determination of peak shifting interval is based on the time constant of the system to ensure that the next adjustment is executed after the previous adjustment is basically stable. Multi-objective optimization also considers economic cost, and preferentially uses low-cost regulation resources under the premise of meeting technical constraints. After peak shifting adjustment and smoothing processing of the execution sequence, the coordinated timing data is generated.

[0089] The time sequence synchronization parameters directly used for the control system are generated according to the coordinated time sequence data. Firstly, the key time nodes in the coordinated time sequence data are extracted, including the starting time of each power source action, the peak time, the stable time, etc., to form a time node sequence. The synchronization control cycle parameters are determined according to the greatest common divisor principle, and the cycle settings of 100 milliseconds for the fast control layer, 1 second for the medium-speed control layer, and 10 seconds for the slow control layer ensure the coordination of each level. The phase relationship parameters describe the relative time sequence of each power source action, and the energy storage leading response phase is set to 0°, the photovoltaic lagging phase is 30°, the wind power lagging phase is 45°, and the diesel engine lagging phase is 90°, and the natural peak shifting is realized through the phase difference. The trigger condition parameters define the threshold values for starting the adjustment of each power source, including the frequency deviation threshold, the power shortage threshold, the SOC threshold, etc., and the differentiated threshold values for different power sources realize the hierarchical response. The coordination waiting time parameters regulate the interlocking relationship between the power sources, such as waiting for the energy storage response to stabilize before starting the diesel engine, and the waiting time is set to 2 seconds. The communication synchronization parameters include data refresh rate, transmission delay compensation, and packet retransmission strategy, etc., to ensure the reliable transmission of control instructions. The coding of the parameters adopts a standardized format, which supports the analysis of control systems of different manufacturers. The integrity of the parameter set contains two sets of schemes for normal operation and fault handling. After the parameterization conversion and standardized packaging of the coordinated time sequence data, the time sequence synchronization control parameters are obtained.

[0090] In step S170, the final load distribution control instruction set is generated based on the time sequence synchronization control parameters, and each power source executes the coordinated output according to the control instruction set to obtain the actual operation state data, and completes the microgrid load distribution control.

[0091] Specifically, the final control instructions of each power source are generated by using the time-synchronous control parameters. The time-synchronous control parameters define the hierarchical control periods of the fast layer 100 milliseconds, the medium layer 1 second, and the slow layer 10 seconds, the phase difference configurations of the energy storage 0°, the photovoltaic 30°, the wind power 45°, and the diesel engine 90°, and the differentiated trigger thresholds of each power source, which are directly converted into the time arrangement of the control instructions. The instruction generation adopts a templating method, and each type of power source corresponds to a specific instruction template. The energy storage system instruction template includes fields such as the charge and discharge power set value, the power change rate limit, and the SOC protection boundary. The photovoltaic instruction template covers the MPPT operating point adjustment, the active and reactive power ratio, and the inverter protection parameters. The wind power instruction template includes the pitch angle setting, the speed control target, and the power limitation curve. The diesel engine instruction template contains the start-stop control, the load rate setting, and the synchronous grid connection parameters. The time stamp of the instruction is calculated according to the phase relationship of the synchronous control parameters, ensuring that each power source performs actions according to the designed time sequence. The trigger logic converts the threshold parameters into conditional judgment statements and embeds them in the instructions to achieve autonomous triggering. The encoding of the instructions adopts the IEC 61850 standard format, ensuring compatibility with devices from various manufacturers. The verification mechanism checks the integrity and consistency of the generated instructions to prevent logical conflicts or parameter overruns. The instruction set is organized into a queue structure according to the execution time sequence, supporting real-time scheduling and dynamic updating. The analysis and conversion of the synchronization parameters and the templating packaging generate the final load distribution control instruction set.

[0092] Each power controller receives and parses the final load distribution control command set and executes coordinated output actions according to a predetermined sequence. The execution process strictly follows the time arrangement in the command set. The energy storage system responds first, adjusting its charging and discharging power within milliseconds to absorb or release power to maintain system balance. The photovoltaic system begins adjustment after a 30° phase delay, smoothly increasing or decreasing output by changing the MPPT operating point. The wind power system operates after a 45° phase delay, with the blade angle controller adjusting the blade angle according to the command. The diesel generator starts with a 90° phase delay when necessary and connects to the grid after a synchronization process. The execution monitoring system tracks the operational status of each power source in real time, recording the deviation between the actual execution time and the planned time. Closed-loop control ensures accurate command execution, automatically compensating for deviations when detected. The coordination mechanism handles mutual influences during execution, such as voltage fluctuations caused by rapid energy storage response interfering with the photovoltaic MPPT, which is eliminated through feedforward compensation. The anomaly handling process handles unexpected situations such as equipment failure or communication interruption, automatically switching to standby control mode. Data acquisition during execution includes electrical parameters such as power, voltage, current, and frequency, as well as operating parameters such as temperature, speed, and SOC. The data acquisition frequency is set according to the variation characteristics of different parameters, with fast variables sampled at 1kHz and slow variables sampled at 1Hz. Based on chaotic exploration driven by entropy increase disturbances, coupled control with frequency synchronization coordination, vulnerability repair through counterexample search, boundary identification through active detection, and coordinated execution with time synchronization, intelligent load distribution control of multiple power sources in the microgrid is achieved, moving from disorder to order and from independence to collaboration. Each power source operates stably according to the optimized control strategy and generates actual operating status data, thus completing the microgrid load distribution control.

[0093] To implement the microgrid load distribution control method based on entropy increase disturbance corresponding to the above method embodiments, and to achieve the corresponding functions and technical effects. See Figure 2 , Figure 2 This diagram illustrates a structural block diagram of a microgrid load distribution control device 200 based on entropy increase disturbance provided in an embodiment of this application. For ease of explanation, only the parts relevant to this embodiment are shown. The microgrid load distribution control device 200 based on entropy increase disturbance provided in this embodiment includes:

[0094] The data acquisition module 201 is used to acquire real-time power output data and load demand data of the power sources in the microgrid, perform chaotic measurement analysis on the real-time power output data to generate system entropy parameters, and construct an entropy increase disturbance excitation mechanism based on the system entropy parameters.

[0095] Chaotic coordination module 202 is used to generate chaotic state operation data by randomly perturbing the output of each power source based on the entropy increase perturbation excitation mechanism, extract the inherent frequency characteristics of each power source from the chaotic state operation data to obtain the resonance coupling matrix, and use the resonance coupling matrix to generate a frequency synchronization coordination mechanism between power sources.

[0096] The strategy optimization module 203 is used to generate a preliminary load allocation scheme based on the frequency synchronization coordination mechanism, perform counterexample search on the preliminary load allocation scheme to generate allocation vulnerability identification data, and obtain an enhanced allocation strategy based on the allocation vulnerability identification data.

[0097] The pressure analysis module 204 is used to obtain load-bearing capacity assessment data of each power source based on the enhanced allocation strategy, match and analyze the load demand data with the load-bearing capacity assessment data to generate a load transmission topology, and use the load transmission topology to obtain pressure distribution data among each power source.

[0098] Boundary detection module 205 is used to generate a boundary active detection instruction set based on the pressure distribution data, use the boundary active detection instruction set to drive each power supply to gradually approach the operating boundary to obtain boundary response data, and perform extreme value analysis on the boundary response data to generate the bearing limit parameters of each power supply.

[0099] The timing control module 206 is used to set a unified timing adjustment benchmark based on the load limit parameter, use the timing adjustment benchmark to synchronize the power output adjustment process of each power source to obtain coordinated timing data, and generate timing synchronization control parameters based on the coordinated timing data.

[0100] The instruction execution module 207 is used to generate a final load distribution control instruction set based on the timing synchronization control parameters. Each power source executes the control instruction set to coordinate output and obtain actual operating status data, thereby completing the microgrid load distribution control.

[0101] The aforementioned microgrid load distribution control device 200 based on entropy increase disturbance can implement the microgrid load distribution control method based on entropy increase disturbance described in the above method embodiments. The options in the above method embodiments are also applicable to this embodiment, and will not be detailed here. The remaining content of this application embodiment can be referred to the content of the above method embodiments, and will not be repeated in this embodiment.

[0102] like Figure 3 As shown, the third embodiment of the present invention also provides a computer device, including a memory 301, a processor 302, and a computer program stored in the memory 301 and executable on the processor 302, characterized in that the processor 302 executes the program to implement the steps of the microgrid load distribution control method based on entropy increase disturbance described in the first embodiment of the present invention.

[0103] The above examples are intended to illustrate and deduce the technical solutions of the present application, and to completely describe the technical solutions, objects and effects of the present application. The purpose is to make the public more thoroughly and comprehensively understand the disclosure of the present application, and not to limit the protection scope of the present application.

[0104] The above examples are not based on an exhaustive enumeration of the present application, and there can be many other unlisted embodiments. Any substitution and improvement made without violating the concept of the present application shall fall within the protection scope of the present application.

Claims

1. A microgrid load distribution control method based on entropy-increasing perturbation, characterized in that, The method comprises the following steps: acquiring real-time output data and load demand data of power sources in a micro-grid, performing chaotic metric analysis on the real-time output data to generate system entropy value parameters, and constructing an entropy increase disturbance incentive mechanism based on the system entropy value parameters; generating chaotic state operation data by performing random output disturbance on each power source based on the entropy increase disturbance incentive mechanism, extracting inherent frequency characteristics of each power source from the chaotic state operation data to obtain a resonance coupling matrix, and generating a frequency synchronization coordination mechanism between power sources by using the resonance coupling matrix; generating a preliminary load distribution scheme based on the frequency synchronization coordination mechanism, generating distribution vulnerability identification data by performing counterexample search on the preliminary load distribution scheme, and acquiring a reinforcement distribution strategy according to the distribution vulnerability identification data; acquiring load bearing capacity evaluation data of each power source based on the reinforcement distribution strategy, matching and analyzing the load demand data and the load bearing capacity evaluation data to generate a load conduction topology, and acquiring pressure distribution data between each power source by using the load conduction topology; generating a boundary active detection instruction set based on the pressure distribution data, driving each power source to gradually approach the operating boundary by using the boundary active detection instruction set to acquire boundary response data, and generating bearing limit parameters of each power source by performing extreme value analysis on the boundary response data; setting a unified time sequence adjustment benchmark based on the bearing limit parameters, the unified time sequence adjustment benchmark being a hierarchical time scale system including a fast adjustment layer, a medium-speed adjustment layer and a slow adjustment layer, which is used for time synchronization of power source controllers, synchronizing the output adjustment process of each power source by using the unified time sequence adjustment benchmark to acquire coordinated time sequence data, including: generating power output adjustment instructions of each power source based on the unified time sequence adjustment benchmark; acquiring time sequence identification data by marking the time sequence of the power output adjustment instructions; generating a coordinated execution sequence by synchronously arranging the time sequence identification data; performing time sequence peak shifting optimization on the coordinated execution sequence to acquire coordinated time sequence data, and generating time sequence synchronization control parameters according to the coordinated time sequence data; generating a final load distribution control instruction set based on the time sequence synchronization control parameters, and each power source executing coordinated output according to the control instruction set to acquire actual operation state data, thereby completing the load distribution control of the micro-grid.

2. The method of claim 1, wherein, The chaotic metric analysis on the real-time output data to generate system entropy value parameters comprises the following steps: performing disorder metric analysis on the real-time output data to acquire a chaotic characteristic vector; performing entropy value integral processing based on the chaotic characteristic vector to generate initial entropy value data; performing time sequence standardization processing on the initial entropy value data to acquire system entropy value parameters.

3. The method of claim 1, wherein, The random output disturbance on each power source based on the entropy increase disturbance incentive mechanism to generate chaotic state operation data comprises the following steps: generating a random disturbance parameter set based on the entropy increase disturbance incentive mechanism; inputting the random disturbance parameter set into each power source to acquire power source exclusive disturbance instructions; driving each power source model to execute output disturbance by using the power source exclusive disturbance instructions to acquire disturbance response data; integrating the disturbance response data to generate chaotic state operation data.

4. The method of claim 1, wherein, The extracting of the inherent frequency characteristics of each power supply from the chaotic state operation data to obtain the resonance coupling matrix comprises: Performing frequency domain transformation on the chaotic state operation data to obtain frequency spectrum distribution data; Identifying dominant frequencies of each power supply from the frequency spectrum distribution data to generate frequency characteristic identifiers; Performing coupling strength analysis based on the frequency characteristic identifiers to construct the resonance coupling matrix.

5. The method of claim 1, wherein, The anti-example search on the preliminary load distribution scheme to generate distribution vulnerability identification data comprises: Constructing an opposite distribution mode opposite to the preliminary load distribution scheme; Performing conflict scenario simulation based on the opposite distribution mode to obtain conflict response results; Analyzing the conflict response results to generate distribution vulnerability markers; Generating distribution vulnerability identification data according to the distribution vulnerability markers.

6. The method of claim 1, wherein, The matching analysis of the load demand data and the load bearing capacity evaluation data to generate a load conduction topology comprises: Performing difference analysis on the load demand data and the load bearing capacity evaluation data to obtain load gap data; Constructing load transfer paths between power supplies based on the load gap data; Analyzing each transfer path to generate conduction impedance coefficients; Generating a load conduction topology using the conduction impedance coefficients.

7. The method of claim 1, wherein, The driving of each power supply by the boundary active detection instruction set to gradually approach the operation boundary to obtain boundary response data comprises: Performing stepwise output increase on each power supply based on the boundary active detection instruction set; Monitoring power supply states in the stepwise output increase process to obtain response characteristic data; Performing nonlinear analysis on the response characteristic data to identify mutation points and record mutation point parameters; Generating boundary response data based on the mutation point parameters.

8. A microgrid load distribution control device based on entropy increasing perturbation, characterized by, Comprise: A data acquisition module configured to acquire real-time output data of power supplies in a microgrid and load demand data, perform chaotic degree measurement analysis on the real-time output data to generate system entropy value parameters, and construct an entropy increase disturbance excitation mechanism based on the system entropy value parameters; A chaotic coordination module configured to generate chaotic state operation data by randomizing output disturbance of each power supply based on the entropy increase disturbance excitation mechanism, extract inherent frequency characteristics of each power supply from the chaotic state operation data to obtain a resonance coupling matrix, and generate a frequency synchronization coordination mechanism between power supplies by using the resonance coupling matrix; A strategy optimization module configured to generate a preliminary load distribution scheme based on the frequency synchronization coordination mechanism, generate distribution vulnerability identification data by anti-example search on the preliminary load distribution scheme, and obtain a strengthened distribution strategy according to the distribution vulnerability identification data; A stress analysis module configured to obtain load bearing capacity evaluation data of each power supply based on the strengthened distribution strategy, perform matching analysis of the load demand data and the load bearing capacity evaluation data to generate a load conduction topology, and obtain stress distribution data between each power supply by using the load conduction topology; A boundary detection module configured to generate a boundary active detection instruction set based on the stress distribution data, drive each power supply to gradually approach the operation boundary by using the boundary active detection instruction set to obtain boundary response data, and generate load bearing limit parameters of each power supply by performing extreme value analysis on the boundary response data. The timing control module is configured to set a unified timing adjustment reference based on the bearing limit parameter, the unified timing adjustment reference is a hierarchical time scale system including a fast adjustment layer, a medium adjustment layer and a slow adjustment layer, and is used for time synchronization of each power supply controller; the unified timing adjustment reference is used to synchronize each power output adjustment process to obtain coordinated timing data, including: generating each power output adjustment instruction based on the unified timing adjustment reference; performing timing marking on the each power output adjustment instruction to obtain timing identification data; synchronously arranging the timing identification data to generate a coordinated execution sequence; performing timing peak-shaving optimization on the coordinated execution sequence to obtain coordinated timing data, and generating timing synchronization control parameters according to the coordinated timing data; The instruction execution module is configured to generate a final load distribution control instruction set based on the timing synchronization control parameters, each power supply executes coordinated output according to the control instruction set to obtain actual running state data, and micro-grid load distribution control is completed.

9. A computer device, comprising: The computer program is stored in the memory and executed by the processor, and the computer program is used to implement the method in any one of claims 1 to 7.

Images